Online commerce, once a luxury, is now central to people’s lives. The Internet is more than just a convenient place to shop for electronics or to book a vacation. Increasingly, it’s also where people go to find a loan, evaluate college degree programs, and seek financial advice. In order to ensure that vulnerable consumers are appropriately protected, consumer and civil rights advocates, regulators, journalists, and others need to understand the realities of new online marketplaces.
Lead generation is the business of selling leads — pieces of evidence that a consumer is interested in a product or service. Businesses have long relied upon leads to find new customers. However, the Internet ushered in sophisticated new lead generation practices, including highly-targeted online advertisements and automated, real-time auction houses for consumer data. These powerful techniques deserve special scrutiny when they are employed to promote potentially exploitative goods and services, such as payday loans and costly for-profit degree programs.
This report focuses on lead generators that encourage consumers to provide information about themselves in order to learn more about a product or service. These lead generators are middlemen. Many of them sell consumers’ data to businesses that offer risky financial products and other controversial services. Their practices can at times be reckless, unethical, or even illegal.
These lead generators are central to the market for online payday loans. Payday loans — small-dollar, short-term credit products with very high interest rates — are harmful to most borrowers’ financial health, and they are illegal or restricted in many states. Some states even restrict solicitations for payday loans. Nevertheless, today, payday lead generators pursue borrowers across the United States. They gather sensitive financial information from vulnerable and often desperate consumers. They can sell this information widely: not only to payday lenders, but also to peddlers of other fringe financial products and sometimes (wittingly or not) to outright thieves.
Some states have sued payday lead generators, alleging violation of their laws. Federal regulators have uncovered large-scale fraud operations fueled by payday leads. Nevertheless, payday lead generators continue to target consumers across the web, even consumers who reside in states where payday lending is illegal.
Lead generators do not operate in a vacuum. They rely extensively on online advertising platforms and commercial data providers. These partnerships allow them to target their desired audience, including, for example, by restricting the geographic scope of their ads. Nonetheless, payday lead generators advertise payday loans nationwide.
In the pages that follow, we explain how online lead generation works; describe the risks and legal complexities specific to lead generation for online payday loans; document the widespread use of search ads by payday lead generators; and recommend interventions.
In preparing this report, we spoke with payday lead generation firms, major online advertising platforms, consumer and civil rights advocates, and federal and state regulators. We reviewed company policies, industry white papers, research reports, and a variety of publicly-available forums and Internet relay chat (IRC) channels. We also ran tests to learn how online payday lead generators are using search engine ads to target consumers online.
This report is organized in three sections:
Understanding Online Lead Generation looks behind the scenes at how each step in the lead generation process works. We begin with a short, illustrative story of what a borrower might experience when seeking a loan online — a process where lead generators play an important but largely invisible role. We then describe how lead generators fit within the broader ecosystem of online marketing.
The Online Payday Lending Ecosystem explores the risks introduced by online payday loans, and the lead generators that promote them. Online payday lenders are often more harmful than their storefront counterparts, and they often operate on dubious legal ground. These online lenders rely heavily, in turn, on lead generators to find new borrowers. Payday lead generators can aggressively resell consumers’ sensitive data, creating significant risks of fraud and identity theft. We then explain that payday lead generators are using online search ads to advertise payday loans nationwide, including to consumers in states where payday lending is illegal.
Interventions: What Can Be Done outlines steps that ad platforms, industry participants, and regulators could take to help mitigate the harms associated with online payday lead generation. We explain that major online advertising platforms have an opportunity to adopt a more practical and effective approach to handling payday ads, point toward stronger best practices for the payday lead generation industry, and highlight oversight options available to the Federal Trade Commission and the Consumer Financial Protection Bureau.
Lead generation, the production and sale of evidence of consumer interest, is marketing in its most concrete and individualized form. It is the practice of “getting people to ‘raise their hands’ and say they are interested in buying, or learning more about, [a] product or service.”1 A lead is evidence of interest by a consumer that can be tracked and monetized.2 A lead generator is an entity that sells leads. Lead generators help a wide variety of businesses find new customers.
Lead generation differs from traditional “brand awareness” marketing (like a billboard along a highway) in two main respects. First, lead generation marketing is measured in concrete actions taken by a consumer. Second, lead generation marketing tends to be highly targeted. (These also happen to be trends in digital marketing generally.3)
Today, the marketing industry lacks a shared understanding of what counts as a “lead.” On one hand, a lead is certainly created when a consumer chooses to fill out an online form for the purpose of receiving information about a product or service. However, some marketers use the term lead to refer a consumer’s click on an online ad: a trackable, monetizable event that directs a visitor to their website.4 Different definitions of “lead” make it difficult to determine what types of companies count as lead generators. For clarity, in the remainder of this report, we use the term “lead generator” to refer only to companies that sell “leads,” defined as information provided by consumers for the purpose of learning more about a product or service.
Leads are often sold to businesses offering suspect products and services.5 For example, lead generators played a central role in the mortgage crisis, connecting consumers with predatory lenders.6 Today, major lead markets include online payday loans, for-profit education, and various debt-relief products: Online payday lenders rely on lead generators to supply as many as 75 percent of their borrowers;7 the already-sizable market for educational leads is poised for further growth;8 and the Consumer Financial Protection Bureau (CFPB) recently reported a spike in student debt relief marketers targeting distressed borrowers online.9
Lead generators can be gateways to fraud and abuse. For example, more than one-third of online payday borrowers interviewed by Pew claimed that their personal financial information was sold without their knowledge, and many had to close their bank accounts in order to stop unauthorized withdrawals.10 Consumers can unknowingly trigger a barrage of high-pressure phone calls just by sharing their contact information through a web form. In an undercover investigation, the Government Accountability Office (GAO) reported that one “student” received 182 phone calls within a month of giving personal information to for-profit education lead generation websites.11
Lead generators come in many different shapes and sizes, and they are highly dependant upon a broader online marketing ecosystem. There are many specialized, brand-name lead generators that focus on particular markets (such as payday loans and mortgages). However, there are also many small, amateur lead generators, known as “affiliates,” that gather leads and sell them to larger, more established lead generation firms. Many lead generators rely extensively on online advertising platforms, which allow them to reach consumers as they search the web, share on social media, or read the news. These marketing companies work in concert, creating complex layers of sales and commissions.
An Illustration: Becky’s Search For a Loan
The following is an illustration of what a consumer might encounter when she seeks a loan online:
Becky, a resident of Philadelphia, lives in a rented apartment that she shares with her daughter. She is 27 years old, has an associate’s degree, and works as a receptionist at a local clinic, earning $25,000 a year. Becky recently separated from her partner, and quickly found herself struggling with the loss of a second income. This month, she doesn’t have enough money to pay all of her bills — including cable, groceries, utilities, childcare, and rent — before her next paycheck.
Becky opens her laptop and types “need money to pay bills” into a search engine. An advertisement next to the search results catches her eye: “Fast Cash! $100-$1000! Approved in 2 minutes, direct to your bank. Bad credit OK!” Becky clicks on the ad and lands on the website of SpeedyLoans. The site features a picture of a smiling couple and the assurance that “sometimes everyone needs help making it to their next payday.” Becky enters her name, email address, and zip code, and clicks the “Get Cash!” button. She is greeted by a second form, which asks more information, including for her bank account numbers. After entering this data, Becky is redirected to another website, LenderCo, where she agrees to loan terms. The next day, LenderCo deposits $500 into Becky’s bank account.
In the weeks following, Becky is unable to repay the full amount of the loan. She repeatedly pays fee after fee to push the due date forward. Three months later, by the time she pays off the loan, Becky’s has repaid $1,200 — $700 in interest and fees on top of the $500 amount she initially borrowed.
In the meantime, Becky begins receiving unsolicited phone calls and text messages. She is offered new loans, “debt relief” services, and expensive online classes. Becky asks to be taken off these callers’ lists, but is unable to stop the calls completely.
This story, though fictional, mirrors the experience of thousands of American consumers who deal with online payday lead generators. Becky suffered through several problems: the $700 she paid in interest and fees to cover a smaller loan; unsolicited calls from other businesses who targeted her financial vulnerability; and she may be at risk of fraudulent withdrawals from her bank account. All this occurred despite that fact that that Becky’s home state, Pennsylvania, has some of the strongest usury laws in the nation and has worked hard to keep payday lenders and lead generators from targeting its residents.
Becky’s initial click on the search engine ad triggered a complex set of transactions: First, SpeedyLoans owed the search engine $10. SpeedyLoans, an affiliate website run by self-employed marketer, collected Becky’s loan application data and sold it to a company Becky never saw, called “Lightning Leads,” for $75. Lightning Leads resold Becky’s data through an instant auction to its network of lenders. The winning bidder in that auction was a lender called LenderCo; LenderCo paid $150 to have Becky redirected to its website. But LenderCo wasn’t the only buyer of Becky’s information: both SpeedyLoans and Lightning Leads continued to sell her data to other businesses (at much lower prices), leading to the unsolicited phone calls.
The remainder of this section explains each of these steps in more depth.
Step by Step
Online lead generation involves a long chain of different actors, including online advertising platforms, affiliates, lead aggregators, and end-buyers. This section outlines, at a high level, how leads are created, enriched, and sold.
Targeted Online Advertising
Online lead generation often begins with online ads.1212 Lead generators pay large online advertising platforms to show ads to that platform’s users. These online ads — whether shown by a search engine, a social network, or on a blog — are targeted with increasing sophistication and insight into people’s lives. However, different types of online ad platforms target users in different ways. This subsection briefly outlines how consumers see ads across the web, and the company policies that govern those ads.
Most search engines show ads alongside the search results that they deliver to their users. Today, Google and Microsoft (through its search engine, Bing) handle more than 80 percent of all web search queries in the United States, and sell the lion’s share of search advertising.13 These companies show ads based on a user’s search term — a strong indicator of what that user is interested in at that particular moment in time. However, advertisers can specify additional targeting criteria, including the user’s current location and rough estimates of that user’s household income.
To show an ad next to Google or Bing search results, an advertiser starts by bidding on specific words or phrases.14 For example, a florist might choose “fresh flowers” in the hopes of having his ads appear alongside search results for that query. Advertisers can further target consumers by geographic location, language, and, with Google, by estimated average household income.15 Both Google and Microsoft make several geographic targeting techniques readily available to advertisers.16 For example, in addition to targeting specific zip/postal codes, advertisers can target arbitrary areas, such as a five mile radius around a shopping mall.17 Ads are ultimately shown based on the amount an advertiser bids, the search provider’s judgment of how well the ad relates to what a person is searching for, and the targeting filters selected by the advertiser.18
Google and Microsoft have company policies that restrict or prohibit certain types of search ads. These policies not only implement legal requirements, but also go further to cover ads in trouble-prone categories. Some ads, such as those for adult-oriented content,19 alcoholic beverages,20 and healthcare-related content,21 cannot be displayed until the advertiser meets special requirements, such as providing a copy of a relevant business license. Other ads are prohibited outright, including those for counterfeit goods and dangerous products or services.22
Social Media Ads
Social media sites typically allow advertisers to show ads next to social content and to more prominently feature their own social content (such as product pages or tweets). Today, Facebook and Twitter dominate the social media market, accounting for the majority of all U.S. social media site visits.23 Unlike search ads, which are targeted primarily based off a user’s search term and geographic location, social media advertising relies more heavily on data supplied by users and third-party data providers.
Facebook and Twitter allow advertisers to target ads based on data they collect from users, data they collect from others, and inferences that they make.24 A user may provide these companies with their location (country, state, city, or zip code), age (or age range), gender, and language preferences.25 Facebook might also collect a user’s relationship status, educational status, employment status, familial relationships, interests, “page likes,” and political and religious affiliations.26 Twitter can collect what users tweet about or the terms or hashtags that they search for.27 Both companies also allow advertisers to target users based on various assumptions that they make.28 For example, Facebook makes educated guesses regarding a user’s financial status (income and net worth), home status (home type, home ownership, home value, or household composition), ethnic affinity, and parental status.29 Twitter also infers a user’s interests and behaviors.30
Facebook and Twitter also allow marketers to leverage data held by third-party commercial data providers, including Acxiom, Datalogix, and Epsilon.31 These partners allow advertisers to use consumers’ purchase history, as well as other online and offline behavior to target ads.32 For example, an advertiser could target “children’s cereal buyers” (relying on data collected and analyzed by third-party data providers) who live in Washington, D.C. (relying data that a user has provided directly to Facebook or Twitter). Also, using both on-site and off-site data, Facebook and Twitter help marketers create “lookalike audiences,” which allow marketers to show ads to people who are similar to their current customers.33
Facebook and Twitter voluntarily restrict and prohibit certain types of ads. These policies not only implement legal requirements, but also go further to cover ads in trouble-prone categories. For example, both companies restrict ads for certain products or services, including alcohol, online real money gambling, state lotteries, online pharmacies, and supplements.34 And both prohibit ads that promote the sale or use of weapons, explosives, tobacco products, and adult products or services.35
Web and Mobile Ads
Thousands of websites and mobile apps sell virtual real estate for advertisements. Sometimes, they negotiate directly with advertisers. But usually, individual sites and apps delegate the task of choosing and displaying ads to an online ad network. Ad networks work with advertisers and data providers to show consumers targeted ads across the web.
To target ads, ad networks build segments — groups of users who share common features or interests. Segments are developed using many different types techniques and data. For example, an ad network might track a consumer as they browse the web, inferring that, based on their recent browsing behavior, they might be a middle-aged man interested in sailing. (Because an ad network displays ads across many different websites, they can simultaneously observe consumers’ behavior and serve ads.) Ad networks also work with commercial data providers to use offline data for targeting. For example, a consumer could be targeted for ads based on their offline purchasing habits, or a rough approximation of their credit score. A consumer could even be targeted because they look like a group of customers that a marketer has enjoyed success with in the past (based on both online and offline data).36
These segments, however they are built, dramatically alter the ads a consumer will see as they browse the web. However, it is often impossible for a consumer to know how they’ve been classified, or why. It is even infeasible for outside researchers to know why particular ads are shown.37
Landing Pages and Affiliates
Online ads are often doorways to landing pages — the websites through which consumers’ information enters the lead generation marketplace. Landing pages usually feature a “call to action” (such as “Get Cash Now!”) that entices consumers to enter information about themselves into a form on the page. In some cases, landing pages are run by large, brand-name lead generation companies like MoneyMutual and LowerMyBills. However, in many cases, “affiliates” — individuals and small businesses looking to make money by generating leads — form the front lines, hosting landing pages and drawing consumers in.
Affiliates (sometimes called publishers) are independent actors that generate leads for a commission. Most affiliates are lead generators themselves, but they typically serve other lead generators. Some affiliates post links to landing pages across the web — in online forums, in blog posts, and elsewhere — and collect a small commission for each click. Others embed another lead generator’s application form on their own landing page, and try to convince consumers to fill it out. And some collect consumers’ information directly, and then sell it onward. Complicating matters further, many affiliates contract with their own sub-affiliates, creating a complex scheme of sales and commissions. Affiliates are sometimes paid immediately upon handing over a qualified lead, and sometimes once a lead results in a sale.
Affiliates are conscripts of larger, more sophisticated lead generation firms. These firms typically make it easy to join their affiliate network.38 Some provide catalogs of pre-designed landing page templates and other creative materials. (“You don’t need to think about anything but driving traffic to your site,” boasts one lead generator.39) Successful affiliates invest heavily in online advertising, ensuring that their websites rank highly in search results, and designing their websites appear trustworthy.40
Lead generation firms sometimes struggle to police their affiliates. “You have to keep in mind that there are monster affiliate networks made of 12-year-olds that have no sense of ethics or morals,” observed one lead generation specialist.42 In fact, many affiliates have a financial incentive to misbehave. Affiliates will frequently submit data that is old or fraudulent, and try to inflate their statistics. Some try to inflate their profits by selling a single lead to multiple buyers. Lead generators that rely on affiliate networks are sometimes forced to play a constant game of “whack-a-mole” to shut down bad actors. Leveraging affiliates requires lead generators to strike a delicate balance between the desire for a high volume of leads on one hand, and the desire for high quality leads on the other.43
Aggregation, Scoring, and Sale
Once a consumer submits their information through a landing page, it becomes a lead and enters a hidden, digital marketplace. Leads are often aggregated by a class of large, professional lead generators that act as clearinghouses for end-buyers.44 These lead generators sell leads to the highest bidders using automated auction systems. Before or after sale, a lead can be validated, enriched, and scored, adding much more detail about the consumer in question.
Lead validation is the process of verifying and “scrubbing” leads. Large lead generators are typically responsible for weeding out leads that contain invalid data, are duplicative, or originate from fraudulent sources. They might verify that the name, address, phone number, and bank information in a lead appear to be legitimate.45 (“Only leads that pass our rigorous validation process get ready distribution,” promises one lead generation firm.46 “[Our] [s]tringent lead validation system ensures you spend time contacting real, interested consumers, not calling wrong numbers,” reports another.47) These validation procedures are often sold as a service by commercial data providers, which maintain large dossiers of information about millions of consumers.
Next, a lead can be enriched with additional data. For example, a “short-form” lead — a lead that contains only a consumer’s name and address — can be enhanced to yield a far more detailed picture.48 A commercial data provider can cross-reference a short-form lead against consumer profiles already in its databases, “filling in the blanks” by adding information about a person’s gender, age, household income, household demographic information, educational level, and more.49 Commercial data providers offer a range of other services too. For example: a zip code can yield a surprising amount of detail, perhaps indicating that a person lives in a low-income area that is more likely to use subprime financial products;50 a person’s name can be automatically scrutinized for clues about their ethnicity;51 and credit bureaus can append information that approximate a person’s credit score.52
Lead scoring, another process typically outsourced to a commercial data provider, can help determine whether a consumer is likely to be a good customer. In some cases, lead scoring might include pulling a consumer’s credit score for underwriting purposes. However, lead scoring also includes other, less regulated scoring contexts.53 For example, a lead scoring model could determine that Latino households in low-income neighborhoods are the most common customers for a particular kind of mortgage refinancing. This insight could be used by a lead generator to price and prioritize its leads. Lead scoring is rarely explained in public documents, and sometimes not even to those who purchase and use the scores. “These complex predictive scoring models and algorithms are ‘under the hood’ items . . . . They do not need to be explained to users,” remarks one industry white paper.54
Eventually, a lead generator will auction its leads to the highest bidders, including both end-buyers and other lead generators.55 In many verticals, leads are sold through real-time online auction systems, which allow buyers to filter available leads based on price and demographic information.56 For example, a buyer might configure its filters so it only bids when lead generators have a new lead on 40-year-olds who live in Georgia and earn less than $30,000 annually.57 Leads are offered first to preferred buyers, and then to others.58 The cycle will continue until the lead is purchased a set number of times.59 After a sale, commissions can sometimes be automatically paid back through the chain of lead generators and affiliates.
Speed is critical in many lead markets. All of the tasks described above — validation, enrichment, scoring, and sale — might be completed within seconds of a user submitting her data through a landing page. The chances that an end-buyer will make a sale can decrease dramatically as time passes. According to one oft-cited study, a company’s chance of contacting a web-generated lead is 100 times higher if a call is made within five minutes after an lead submission is made than if a call is made within 30 minutes.60 Accordingly, some companies specialize in helping end-buyers reach out to leads quickly. One call center service guarantees that leads will be called by a human operator within two minutes, but claims that it “usually dial[s] in less than 30 seconds.”61
For many leads, the story does not end after the race for initial contact. Some lead generators will retain aged leads for sale at continually-dwindling prices. Old leads are often compiled into marketing lists and resold for year to come. For example, one publicly-available list purports to contain Hispanic mortgage holders who are good targets for payday loans.62 “[D]ebt is also on the rise for Hispanic families,” claims the listing. “You can target known mortgage holders needing cash to pay their bills.” These marketing lists can be used to target a new set of online ads, starting the lead generation cycle all over again.
David T. Scott, The New Rules of Lead Generation, (AMACOM), March 20, 2013, 17. ↩
Leads can also refer to interest by a business or other entity. However, this report is focused on exclusively on leads related to consumers. ↩
See, e.g., Claudia Perlich, Creatives are from Venus, Targeting is from Mars, July 22, 2015, available athttps://www.linkedin.com/pulse/creatives-from-venus-targeting-mars-claudia-perlich (Describes a convergence of lead generation marketing techniques and advertising: “This brave new world of reaching consumers has gone from having a feeling consumers are going to engage with a brand’s ad to accurately predicting which consumers will engage with a brand’s ad.”). ↩
See, e.g., Interactive Advertising Bureau, “Glossary of Interactive Advertising Terms,” 19, available athttp://www.iab.net/media/file/GlossaryofInteractivAdvertisingTerms.pdf (“Pay-per-lead — an advertising pricing model in which advertisers pay for each ‘sales lead’ generated. For example, an advertiser might pay for every visitor that clicked on an ad….”). ↩
One lead generation expert observes that lead generators often serve businesses with a common set of features: “some sort of (a) government intervention into the market that creates and sustains the margin, (b) a third-party funding source, (c) a purely IP product, (d) a scam product, or some combination thereof,” observes Rich McIver, an attorney familiar with these industries. Andy Hagans, Interview with Rich McIver, Lead Gen Entrepreneur & Texas Attorney, Monetize Pros, Sept. 4, 2013, available athttp://monetizepros.com/interviews/interview-with-rich-mciver-lead-gen-entrepreneur-texas-attorney/. ↩
Gregory D. Kutz, Testimony Before the Committee on Health, Education, Labor, and Pensions, United States Senate, “For-Profit Colleges: Undercover Testing Finds Colleges Encouraged Fraud and Engaged in Deceptive and Questionable Marketing Practices,” August 4, 2010, 16, available athttp://www.help.senate.gov/imo/media/doc/KutzTestimony.pdf. ↩
Online ads are not the only way that lead generators find consumers. Although this report focuses on online marketing, some lead generators successfully use offline marketing to draw consumers to their websites. For example, Money Mutual, a powerful payday loan lead generator, has invested heavily in both online and offline ads. The company was responsible for most of the $277 million spent on television and radio ads for payday loans nationally between June 2012 and May 2013. See generally The Pew Charitable Trusts, Fraud and Abuse Online: Harmful Practices in Internet Payday Lending, October 2014, 6, available athttp://www.pewtrusts.org/~/media/Assets/2014/10/Payday-Lending-Report/Fraud_and_Abuse_Online_Harmful_Practices_in_Internet_Payday_Lending.pdf. ↩
Google, AdWords Help, “Target ads to geographic locations,” available athttps://support.google.com/adwords/answer/1722043?hl=en. (Google allows marketers to target five different income tiers, ranging from lower 50 percent to top 10 percent. Google does not offer more precise targeting of groups within the bottom 50th income percentile. Even though the IRS provides more granular data sets, Google apparently chooses not to target more precisely.) See, e.g., Google AdWords Benefits, available athttps://www.google.com/adwords/benefits. ↩
Facebook Help Center, “How do I create an audience for my ad?” available athttps://www.facebook.com/help/1381741108746415. (“People are included in different targeting segments based on data they share and assumptions Facebook makes.”). ↩
For example, an insurance broker who wanted to find new customers could share a list of their existing customers with a commercial data provider, which might in turn discover that the list suggests that low-income individuals living in urban areas are highly desirable targets. It would then draw new targets from its dossier of consumers, and help the broker target them with the help of ad networks. See, e.g., TruSignal, Specialized Audience Buyer’s Guide, 2015, available athttp://www.tru-signal.com/wp-content/uploads/2014/11/TruSignal_interactive_audience_guide.040615.pdf. (How Lookalike and Act-Alike Audiences Can Help You Find New Customers. Start with a consumer sample that “exemplifies the type of new prospects you want to find.”). ↩
Upturn, Knowing the Score: New Data, Underwriting, and Marketing in the Consumer Credit Marketplace, October 2014, 20, available at https://www.teamupturn.org/static/files/Knowing_the_Score_Oct_2014_v1_1.pdf. ↩
David T. Scott, The New Rules of Lead Generation, (AMACOM), March 20, 2013, 70 (“If you give them a postal or e-mail address list for your [leads], these companies will research customer information and share that data with you. (Usually, they match your customer contact information against profiles they’ve already established in their consumer databases.)”). ↩
Gerry Brown, “How Real-Time Online Sales Lead Scoring Drives A Competitive Edge,” Bloor Research, February 2009, available athttp://www.ebureau.com/sites/all/files/file/eBureau-Bloor.pdf. (“These complex predictive scoring models and algorithms are ‘under the hood’ items … They do not need to be explained to users.”). ↩
Lead Capsule, “Ping Tree Software,” available athttps://www.leadcapsule.com/ping-tree.aspx (“Potential lenders are preconfigured to participate in the ping tree on certain price tiers. Depending on the quality of the lead, it will be initially offered to the top tier members of the ping tree.”). ↩
The lead generation process described above is central to the market for online payday loans. Online payday lenders rely extensively on lead generators to attract customers.63 Payday leads are expensive, a fact that ripples across the online marketing ecosystem.64 At the outset, affiliates can pay more than $10 per click to display ads alongside Google search terms like “payday loans.”65 These clicks might result in payday leads, which can sold for as much as $200 at auction to other lead generators and online payday lenders, and then resold to other buyers.
This section first explains that online payday loans are often worse for consumers than their storefront counterparts: They are associated with higher fees, longer-term indebtedness, higher rates of borrower abuse, and startling rates of fraud.66 Next, we explore the diverse backdrop of state lending laws. Finally, we show that generators help lenders skirt state laws by advertising payday loans nationwide, including to consumers in states where payday lending is illegal.
The Risks of Online Payday Lending
Payday loans are small-dollar, short-term credit products with high interest rates. A longstanding body of research shows that payday loans are harmful to most borrowers’ financial health.6767 Payday loans are seldom short-term solutions: more than 80 percent of payday loans are rolled over or renewed within two weeks, and the average payday loan borrower is indebted to a payday lender for five months per year.68 Most borrowers end up renewing their loans so many times that they pay more in fees than the amount of money they originally borrowed.69 A 2006 Department of Defense study found that payday loans and other “[p]redatory lending undermines military readiness, harms the morale of troops and their families, and adds to the cost of fielding an all volunteer fighting force,” prompting Congress to legislate to protect members of the armed forces fro high-interest loans.70
Payday borrowers disproportionately come from poor and minority communities. The groups with the highest odds of having used a payday loan include “those without a four-year college degree; home renters; African Americans; those earning below $40,000 annually; and those who are separated or divorced,” reports Pew.71 Of these characteristics, being African American is the single strongest predictor: African Americans are 105 percent more likely to use a payday loan than other ethnic groups.72
Online payday loans appear to account for a meaningful portion of the payday market, and they are often riskier than their offline counterparts.73 90 percent of Better Business Bureau complaints about payday lenders relate to online, not storefront, lenders.74 They are associated with higher fees and longer term indebtedness.75 They often come with complex terms and repayment structures and can be especially confusing for consumers.76 And online borrowers report high rates of abusive phone calls.77
Online payday loans can also be a gateway to fraud. Because online lenders typically rely on electronic access to borrowers’ bank accounts (as opposed to a postdated check), payday lead generators almost invariably collect consumers’ bank account information. This data is sometimes shared recklessly. Almost a third of online payday borrowers surveyed by Pew reported that their personal or financial data was sold without their consent.78 Nearly as many reported unauthorized bank withdrawals in connection with an online payday loan.79
Federal regulators have repeatedly discovered payday lead generators at the center of sweeping financial fraud operations. In 2014, the Federal Trade Commission (FTC) sued LeapLab, a company that “collected hundreds of thousands of consumer payday loan applications” from lead generators, and then “used [the leads] to make millions of dollars in unauthorized debits and charges.”80 The same year, it also sued CWB Services LLC, which made unauthorized withdrawals from consumers’ bank accounts using data purchased from lead generators.”81 In 2015, it sued Sequoia One, LLC and Gen X Marketing, two companies who purchased (or collected) payday loan leads from lead generators and sold those leads to non-lenders who fraudulently withdrew funds from consumers’ bank accounts.82 Similarly, the CFPB sued Hydra Group, which made repeated unauthorized withdrawals from consumers’ bank accounts using data purchased from lead generators.83
Our own survey of payday lead generation websites revealed alarmingly weak privacy policies.84 For example, Money Mutual reserves a virtually unlimited right to “share, rent, sell or otherwise disclose” leads to other businesses and also reserves the right to contact users in any way, “even if [their] number is found on a do-not-call registry or similar registry.85 Another company contemplates selling consumers’ data to a wide array of non-lenders, including “financial service providers, such as mortgage and life insurance agencies; title service companies; debt & credit services companies; and auto-finance companies.”86 For entities entrusted with consumers’ sensitive financial details, these are incredibly permissive policies.
We also observed some Internet forums and chat rooms that were rife with evidence of misbehavior by lead generators, especially by smaller affiliates.87 We saw affiliates sharing tips for monetizing “unqualified leads” — leads that the major lead generators don’t want to buy. One forum poster advised that new affiliates should “[find] lead buyers willing to take a chance on a ‘warm body’ with a high accept rate for somewhere in the $0.50 - $2.50 range.”88 Another reported that they were passing unqualified leads on to debt consolidation and credit monitoring companies. It was common to see affiliates selling “legacy” payday leads (leads that had already been sold to lenders) at a steep discounts, and in large quantities.
A Complex and Controversial Legal Landscape
Many states restrict payday lending. According to a Pew study of state laws, payday lending is limited in twenty-four states — it is somewhat restricted in nine and severely restricted in fifteen.89 Approximately 70 percent of online payday lenders fail to obtain a required license in one or more of the states in which they make loans, resorting to offshore incorporation, sovereign nation partnerships, or arguments that the less restrictive laws of the lender’s home state should apply.90 A growing number of legal judgments weigh against online lenders who disregard state usury laws.9191 These jurisdictional strategies put online lenders on “increasingly tenuous legal ground,” says Nick Bourke of Pew.92 Similarly, New York’s Department of Financial Services (DFS) claims that “Internet payday lending is just as unlawful as payday lending made in person in New York.”93
In addition to regulating lenders themselves, a growing number of states appear to require that payday lead generators also be licensed and comply with usury laws.9494 For example, Pennsylvania requires that anyone who “hold[s] himself out as willing or able to arrange for” certain loans be licensed.95 Citing this provision, a Pennsylvania regulator prevailed in obtaining a commitment from MoneyMutual, a prominent payday lead generator, to stop accepting applications from and targeting advertisements toward Pennsylvania residents.96
Some states have also pursued payday lead generators under more general purpose laws. For example, New York’s Department of Financial Services (DFS) sued MoneyMutual under a state law that prohibits fraud and misrepresentation associated with financial products.97 DFS alleged that Money Mutual lied to consumers by claiming that loans provided by its network were suitable for “emergency, one-time, affordable and efficient use,” when in fact those loans “contained terms that often led consumers to roll over their debt and obtain additional high-interest loans to pay off their prior loans.”98
Other states have gone tried stop online payday lenders and lead generators that target their residents with ads. Most prominently, Vermont, as part of a larger operation against illegal online payday lending, requested that several major online advertising platforms — including Google and Microsoft — disable advertising for unlicensed lenders that they had identified in violation of state law.99 Google and Microsoft agreed, and prohibited a number lenders from advertising.100 Vermont launched list of “Unlicensed Lenders,” in cooperation with several other states.101 However, some entities in the “Unlicensed Lenders” list continue to advertise on major platforms, despite a state claim of non-compliance.102 And, as we explain below, many payday lead generators continue to target ads to Vermont residents, and residents of other states where payday lending is illegal.
Using Online Ads, Payday Lead Generators Target Consumers Nationwide
In a series of tests, we saw payday lead generators targeting ads to, and solicit sensitive financial information from, consumers nationwide. In many cases, these lead generators were violating company policies and state laws.
To test how payday lead generators were using major ad platforms to advertise, we ran a series search queries on Google and Bing (including, for example, “payday loan,” “need a loan fast,” and “need money to pay rent”) from internet protocol (IP) addresses originating in states with strong payday lending laws (including Pennsylvania, New York, and Vermont). In each jurisdiction, we saw many payday loan ads commissioned by lead generators.
We clicked on many of these ads, and entered test data into these lead generators’ landing pages — including address information consistent with the apparent jurisdiction of the initial search and test bank account data. The lead generators almost always collected this test data, failing to filter their form submission processes. Some even claimed that they had matched our test data with lenders. And one falsely reported that Pennsylvania “permits payday lenders to operate and charge any interest rate or fees which the borrower agrees to pay.”103
Nearly every ad that we saw during this testing came from a lead generator, not a lender. This was not surprising. Even payday affiliates themselves might not have direct contact with online lenders. “[Y]ou can’t find 90% of these lenders. Most want to be secretive, most use [‘doing business as’ names] that are different then the real name and do not provide contact info anywhere on the internet,” observed one payday affiliate on a message board.104 And as described above, leads can travel through multiple entities — from one lead generator to the next — before they are purchased by lenders.
Our testing had limits. We did not submit valid bank account information to the lead generators, and thus we did not formally complete a loan application process. Nonetheless, the testing that we were able to complete strongly suggests that lead generators (and the lenders that they serve) continue to operate in states where payday lending is illegal.
One former lead generation employee, speaking about his previous company, said that “[t]hey sell college leads, they sell auto lads, they sell all types of leads. … But payday loans are where they make most of their money.” Danny Bradbury, “Scammed,” Matter Archive, available at https://medium.com/matter-archive/scammed-f4a5d98a4f51. ↩
See, e.g., Uriah King and Leslie Parrish, Center for Responsible Lending, Springing the Debt Trap: Rate caps are only proven payday lending reform, 3-7, December 13, 2007, available athttp://www.responsiblelending.org/payday-lending/research-analysis/springing-the-debt-trap.pdf (“The vast majority of families taking out payday loans are ensnared in long-term debt, making them worse off than they would be without high-cost payday lending … State regulator data demonstrates that only one to two percent of transactions are made to borrowers who take out one loan, pay it off on time, and do not need to borrow again that year.”); The Pew Charitable Trusts, How Borrowers Choose and Repay Payday Loans, 6, February 2013, available athttp://www.pewtrusts.org/~/media/assets/2013/02/20/pew_choosing_borrowing_payday_feb2013-(1).pdf (“The average borrower can afford to pay $50 per two weeks to a payday lender — similar to the fee for renewing a typical payday or bank deposit advance loan — but only 14 percent can afford the more than $400 needed to pay off the full amount of these non-amortizing loans.”); Jean Ann Fox, Consumer Federation of America, The Growth of Legal Loan Sharking: A Report on the Payday Loan Industry, 6, November 1998, available athttp://www.consumerfed.org/pdfs/The_Growth_of_Legal_Loan_Sharking_1998.pdf (“It is not unusual for borrowers to become mired in debt and renew cash advance loans every week or two. Payday loans are structured to make it difficult for consumers to pay in full at the end of the loan period without needing to borrow again before the next payday. A consumer paying off a loan of $100 to $300 plus the $15 to $45 fee within a few days often finds it difficult to make it to the next payday without having to borrow again.”); Consumer Financial Protection Bureau, “CFPB Considers Proposal to End Payday Debt Traps,” CFPB News Room, March 26, 2015, available athttp://www.consumerfinance.gov/newsroom/cfpb-considers-proposal-to-end-payday-debt-traps/ (“The CFPB … is concerned that the practices often associated with these products — such as failure to underwrite for affordable payments, repeatedly rolling over or refinancing loans, holding a security interest in a vehicle as collateral, accessing the consumer’s account for repayment, and performing costly withdrawal attempts — can trap consumers in debt.”). ↩
FTC V. CWB Services, et al., available athttps://www.ftc.gov/system/files/documents/cases/140917cwbcmpt.pdf (Noting that “Defendants generally purchase two general categories of consumer leads: (a) data from consumers who submitted applications for online payday loans through third-party lead generator websites, but whose application was denied or who never consented to Defendants’ loan terms; and (b) data from consumers who never applied for an online payday loan, but may have submitted personal information to a nonpayday-related website.”). ↩
Courts have not looked favorably upon online lenders who attempt to avoid state usury laws or regulations by employing choice-of-law provisions in payday loan contracts. See, e.g., Jackson v. Payday Financial, LLC, where the Northern District Court of Illinois (East Division), upon remand from the Court of Appeals for the Seventh Circuit, found an online lender’s tribal choice-of-law provision unenforceable because the lender’s underlying business activity was contrary to Illinois’ public policy against usury. Accordingly, the online lender could not avoid potential liability under Illinois’ usury laws.; Otoe-Missouria Tribe of Indians, et al., v. New York Department of Financial Services, where the United States District Court for the Southern District of New York held that the New York Department of Financial Services could regulate the activities of sovereign tribal nations offering online payday loan services, even if the lenders claimed sovereign immunity, because the tribe’s online payday lending constituted regulable off-reservation activity.; Quik Payday, Inc. v. Stork, where the court found “[t]he discrete nature of the regulated transactions make the internet payday loan industry similar to the insurance industry or any other industry in which a company must tailor its business to conform to the laws of its customer’s state of residence.” (emphasis added).; Bankwest, Inc. et al., v. Oxendine, where the Court of Appeals of Georgia found that “parties to a private contract who admittedly make loans to George residents cannot, by virtue of a choice of law provision, exempt themselves from investigation for potential violations of Georgia’s usury laws.” Federal regulators have also found certain acts by unlicensed online payday lenders in violation of a state’s usury laws to be unfair, deceptive, and/or abusive. See, e.g., Consumer Financial Protection Bureau v. CashCall, Inc., where the CFPB alleged that because payday loans made by unlicensed lenders in contravention of state usury laws limited or voided consumers’ obligation to repay, online lenders’ “servicing, extracting payments for, and collecting” on those loans constituted an unfair practice not reasonably avoidable by the consumer. Though the CFPB did not argue that unlicensed payday lending in contravention of state usury law is de jure an unfair, deceptive, or abusive act or practice, the CFPB’s pleadings in CashCall establish a de facto regulatory regime where unlicensed payday lending in contravention of state law could be subject to UDAAP liability, as “servicing, extracting payments for, and collecting,” are core functions of an online payday lender. ↩
See, e.g., Richard Eckman, “Lead Generators: The law of Internet loan lead generators,” November 5, 2009, 9, available athttp://pdlba.com/coursematerials2009.html (“Eleven states require brokers of payday loans that ‘broker’, ‘arrange’ or ‘facilitate’ the origination of a payday loan to be licensed or registered with a state.”). In some states, payday loan lead generators appear to be regulated under existing law. See, e.g., Colorado Deferred Deposit Loan Act, 5-3.1-102 (5)(a), available athttps://www.coloradoattorneygeneral.gov/sites/default/files/uploads/uccc/DDLA.pdf (“‘Lender’ means any person who offers, or makes a deferred deposit loan, who arranges a deferred deposit loan for a third party, or who acts as an agent for a third party, regardless of whether the third party is exempt from licensing under this article or whether approval, acceptance, or ratification by the third party is necessary to create a legal obligation for the third party, through any method including mail, telephone, internet, or any electronic means.”); see also Idaho Code 28-46-402(1), available athttp://www.legislature.idaho.gov/idstat/Title28/T28CH46SECT28-46-402.htm. (“No person shall engage in the business of payday loans, offer or make a payday loan, or arrange a payday loan for a third party lender in a payday loan transaction without having first obtained a license under this chapter.”); see also Illinois Payday Loan Reform Act, 815 ILCS 122/1-10, 3-3, available athttp://www.ilga.gov/legislation/ilcs/ilcs4.asp?DocName=081501220HArt%2E+3&ActID=2697&ChapterID=67&SeqStart=2100000&SeqEnd=2500000. (“‘Lender’ and ‘licensee’ mean any person or entity, including any affiliate or subsidiary of a lender or licensee, that offers or makes a payday loan, buys a whole or partial interest in a payday loan, arranges a payday loan for a third party, or acts as an agent for a third party in making a payday loan, regardless of whether approval, acceptance, or ratification by the third party is necessary to create a legal obligation for the third party, and includes any other person or entity if the Department determines that the person or entity is engaged in a transaction that is in substance a disguised payday loan or a subterfuge for the purpose of avoiding this Act.”); see also Montana Deferred Deposit Loan Act 31-1-703 - 705, available athttp://leg.mt.gov/bills/mca_toc/31_1_7.htm. (“This part applies to deferred deposit lenders and to persons who facilitate, enable, or act as a conduit for persons making deferred deposit loans. A person may not engage in or offer to engage in the business of making deferred deposit loans unless licensed by the department.”); see also Rhode Island General Laws 19-14-1, 19-14-2, available athttp://webserver.rilin.state.ri.us/Statutes/TITLE19/19-14/INDEX.HTM. (“‘Loan broker’ means any person who, for compensation or gain, or in the expectation of compensation or gain, either directly or indirectly, solicits, processes, negotiates, places or sells a loan within this state for others in the primary market, or offers to do so … No person shall engage within this state in the business of: (1) making or funding loans or acting as a lender or small loan lender; (2) brokering loans or acting as a loan broker … without first obtaining a license or registration from the director or the director’s designee..”); see also Vermont Statutes Annotated 9-63-2481w, available athttp://legislature.vermont.gov/statutes/section/09/063/02481w. (“It is an unfair and deceptive act and practice in commerce for a lender directly or through an agent to solicit or make a loan to a consumer by any means unless the lender is in compliance with all provisions of 8 V.S.A. chapter 73 [No person shall without first obtaining a license under this chapter … engage in the business of making loans of money, credit, goods, or things in action and charge, contract for, or receive on any such loan interest, a finance charge, discount, or consideration therefor.])” State regulators have also taken action to regulate payday loan lead generators. See, e.g., Washington’s Department of Financial Institutions WSR 14-24-048, available athttp://lawfilesext.leg.wa.gov/law/wsr/2014/24/14-24-048.htm. (“‘Small loan agent services’ include, but are not limited to: (a) Marketing and advertising small loans; (b) Collecting nonpublic personal information from consumers in anticipation of selling the information to potential licensed lenders or other entities providing small loan agent services.”); see also California’s Department of Business Oversight, “Invitation for Comments Proposed Changes Under the California Deferred Depoist Transaction Law,” Section 2023, available athttp://www.dbo.ca.gov/Licensees/Payday_Lenders/pdfs/PRO0408-CDDTL_2ndInviteProposedText.pdf. (“A deferred deposit originator means any person who … conducts deferred deposit transaction business, or arranges a deferred deposit transaction for a deferred deposit originator, acts as an agent for a deferred deposit originator, or assists or helps facilitate a deferred deposit originator in the origination of a deferred deposit transaction … and collects personal information from prospective customers, in this state by originating business from or directing business to this state.”). ↩
In the Matter of: Selling Source, LLC, et al., “Consent Order,” New York State Department of Financial Services, Financial Frauds and Consumer Protection Division, March 10, 2015, available athttp://www.dfs.ny.gov/about/ea/ea150310.pdf. ↩
California has pursued similar action. See, e.g., California Department of Business Oversight Press Release, “DBO Announces Effort to Fight Search Engine Advertising by Unlicensed Payday Lenders,” April 7, 2015, available athttp://www.dbo.ca.gov/Press/press_releases/2015/Search_Engine_initiative_04-07-15.asp. (“When the DBO identifies unlicensed online payday lenders, it issues cease and desist orders against them. Under the protocol, when those orders become final, the DBO will notify designated individuals at Microsoft and Google. The firms then will take quick action to block the lenders’ ads, if they are advertising on the search engine pages.”). ↩
Options for ad platforms, payday lead generators, and regulators
Payday lead generators expose consumers to two types of risk: First, they connect consumers with an especially hazardous breed of payday loan. Second, they can share consumers’ sensitive financial data widely, increasing the chance that it will fall into the hands of bad actors. These risks fall disproportionately on poor and minority communities. Today, payday lead generators are using ad platforms like Google and Bing to show payday loan ads nationwide, even in states that outlaw both payday lending and payday lead generation.
Stronger federal and state restrictions on payday lending are likely to help solve these problems. The CFPB is considering a nationwide rule that would require payday lenders to take steps to ensure that borrowers can repay loans.105 And state lawmakers will continue to consider whether their laws appropriately protect their residents (to date, approximately twenty-four states have some limits on payday lending, as reported by the Pew Charitable Trusts).106 New rules on the federal and state level will not only help to limit irresponsible lending, but also narrow the demand for payday leads.
However, in the short term, it will fall to ad platforms, lead generators, trade groups, and state and federal regulators to protect consumers from harmful payday lead generation practices. Today, there is no overarching federal law that governs the collection and sale of personal data by commercial actors.107
Below, we describe three areas of intervention. First, Google, Bing, and similar online ad platforms have an opportunity to adopt a more practical and effective approach to regulating payday loan ads. Second, lead generators and their trade groups could develop stronger best practices to limit dissemination of sensitive consumer data, and clarify where payday lead generators should operate. Third, federal regulators, including the CFPB and the FTC, could exercise additional oversight over lead generators and their affiliates.
Online Advertising Platforms
Google, Bing, and similar online ad platforms have an opportunity to adopt a more practical and effective approach to regulating payday loan ads. These companies already have relevant policies with good aspirations: For example, both Google and Bing require that advertisers comply with applicable laws. But, in practice, these policies are hard to enforce effectively. Payday lead generators are currently taking advantage of this enforcement gap.
Below, we present several different approaches that major online ad platforms could take to payday loans ads. We urge ad platforms to engage with other stakeholders — including civil rights and financial advocates — in considering these choices. We begin by describing the wide range of circumstances in which online ad platforms have adopted voluntary policies that protect their users. We then explain that major advertising platforms have technical tools to identify and label different types of ads in an effective and automated fashion. We also explain that platforms can automatically restrict how ads are shown, for example, based on location. We conclude that new approaches to policy and oversight by ad platforms could have a positive impact on consumers and help states more effectively enforce their laws.
A Spectrum of Company Policies
Today, online ad platforms have a range of policies regarding ads for payday loans. Facebook, for example, recently decided to flatly prohibit ads for “[p]ayday loans, paycheck advances or any other short-term loan intended to cover someone’s expenses until their next payday,” a policy it adopted in August of 2015.108 (Previously, Facebook required that any such ads be authorized by the company.109) Microsoft and Twitter prohibit the advertisement of illegal products or services,110 but neither appear to have restrictions specific to payday loan ads.111
Google currently has two sets of payday loan-specific ad policies. The first requires that payday loan advertisers provide certain disclosures on their websites, such as a physical address and information about interest rates.112 It also requires advertisers to comply with state and local regulations.113 However, we observed many payday lead generators advertising on Google in violation of this policy, either by neglecting to include the necessary disclosures, or by serving their advertisements into geographic markets where it is illegal for lead generators to operate. Google’s second policy states that, for ads tied to a search, Google will “will only serve payday loan ads if the phrase ‘payday loan’ (or similar terms) are included in the user’s query,” and that for ads the company places on other web sites, payday loan ads “will be shown only on sites related to payday loans.”114 Here again, however, there is an enforcement gap. We saw payday loan ads appear in response to searches that do not use a term similar to “payday loan” including, for example, a search for “i need money to pay my rent.”
These rules are part of a larger body of ad platform policies that restrict ads in a variety of trouble-prone and sensitive categories. Many of these policies are discretionary choices, made at a human level, and go well beyond compliance with minimum legal requirements.
Google — the dominant player in web searches and associated ads115 — goes beyond its minimum legal obligations and imposes additional rules in order to “help keep people safe both online and offline,” and to ensure that its users can “trust that information about them will be respected and handled with appropriate care” by advertisers.116 For example, Google prohibits all advertisements for fireworks, tobacco products, and weapons.117 In other potentially concerning areas, Google allows some ads, but subjects them to special restrictions. For example, ads for alcoholic beverages are allowed in the United States, but the company prohibits any alcohol ads that “imply that drinking alcohol can improve social, sexual, professional, intellectual, or athletic standing,” or that “feature binge or competition drinking.”118 Advertisers are permitted to promote some gambling-related content, but only after Google checks the advertiser’s license.119
Similarly, for ads appearing alongside searches on its Bing search engine, Microsoft bars ads in “[a]reas of questionable legality,” including those that are “considered sensitive, legal, dangerous, harmful and/or potentially unethical in nature.”120 Ads for dating services and peer-to-peer file sharing are prohibited.121 Some gambling ads are allowed, but ads that “imply or suggest that gambling is a viable alternative to employment or financial investments, [or] a way to recover from financial losses” are prohibited.122
Ad Platforms Can Automatically Classify and Geotarget Ads
Major online ad platforms have a suite of powerful technologies at their disposal. They can control when, where, and in what context each ad is displayed. On the one hand, these abilities are precisely what makes online advertising attractive to many marketers, including payday loan advertisers. On the other hand, ad platforms can use these capabilities to better enforce their policies.
Ad platforms can “geotarget” ads to particular countries, states, cities, and neighborhoods. Geographic limits are a key to some ad restrictions on both Google and Bing. On Google, ads for alcoholic beverages may not “violate applicable laws and industry standards for any location that your campaign targets.”123 And gambling related ads must “[m]eet local licensing requirements for all gambling-related products and services that you’re promoting.”124 On Bing, online pharmacies “must be certified in the market they are serving to advertise prescription drugs in that market,”125 paralleling a similar restriction at Google.126
Internally, Google and Bing can automatically categorize different types of ads with a reasonably high degree of accuracy. For example, Google has sophisticated software that helps it sift through its enormous haystack of ads, flagging those that are likely to be subject to policy restrictions.127 This automated pipeline employs machine learning models and a rules engine to examine each ad, and the website behind each ad.128 Having automatically established that an ad likely belongs to a certain category, Google can then automatically limit the range of circumstances in which the ad appears, allowing it to “show only in certain regions, only to certain ages, or only on certain devices.”129 Microsoft has indicated that it has similar capabilities.130
Looking Ahead: Three Approaches to Payday Loan Ads
These many policies and technical capabilities point to a range of options for restricting online ads for payday loans. These approaches vary significantly in their costs, efficiencies, and effects. Ad platforms like Google and Bing could:
Maintain the status quo of broad policies and limited enforcement. Ad platforms could choose to continue requiring that payday advertisers both comply with state law and (as Google requires) provide consumers with important disclosures. However, due in part to the complexity of state laws and debates over how these laws apply, these policies cannot be automatically and efficiently enforced at scale. There is simply too much human judgment required.
Under this approach, enforcement falls primarily to advertisers themselves, and to state enforcement agencies. Unfortunately, payday advertisers have shown a willingness to disregard platform policies. State enforcers are not equipped to efficiently deal with an ever-shifting array of payday ads: they have no efficient, automated way of flagging ads for review by an ad platform. Moreover, they must divide their time between dealing with online ads, and payday lenders and lead generators themselves.
The result is widespread violation of both the letter and spirit of ad platform policies by payday lead generators. Consumers see ads for payday loans nationwide, even consumers residing in states with protective lending laws. These ads are doorways to debt traps and fraud.
Commit more resources to enforcing existing, judgment-intensive policies. Ad platforms could choose to devote more resources to manually reviewing ads submitted by payday lenders and lead generators. There is some precedent for a more resource-intensive ad review process. For example, since 2009, Google has required that online pharmacy advertisers be certified by the National Association of Boards of Pharmacy before showing ads.131 Google reports that this requirement, along with other review steps, has reduced the number of ads placed by unlicensed pharmacies by 99.9 percent.132
Ad platforms could take a similar approach to payday ads by, for example, requiring that advertisers to demonstrate compliance with state licensure requirements, including requirements for loan arrangers, before targeting any ads in states that require such licensure. Such an approach could be highly effective at preventing payday lending activities that violate state law. However, this approach would likely come at a significant cost, requiring the ad platform to create a human review team, or outsource review to another entity.
Adopt a new, streamlined policy for payday loan ads that can consistently and automatically be enforced at scale. Ad platforms could choose to adopt a new policy that would be easier to apply in an automated and consistent way. For example, they could ban all payday-related advertising, as Facebook has done. Alternatively, they could adopt a policy that prohibits payday loans ads in states that the platform (or another suitable arbiter) has identified as substantially restricting payday lending. For example, the Pew Charitable Trusts has classified state payday loan regulations into three categories, as follows:
Ad platforms could automatically prevent the delivery of payday loan ads into the 24 “restrictive” and “hybrid” states, or merely prevent delivery of such ads into the 15 “restrictive” states. In either case, the ad platform would protect many consumers from seeing ads for potentially harmful loans that their states have chosen to prohibit. These policies would, to varying extents, curtail some activity that is clearly or arguably lawful. (Ad platforms have repeatedly made such judgments before, in a variety of other contexts.) The ad platform applying such a policy would also lose revenue that it might otherwise earn from showing newly-restricted payday loan ads.
Importantly, a streamlined approach would allow for effective, automated, and relatively low-cost enforcement. This enforcement need not be perfect to be highly effective.133 Under this approach, state law enforcement officials could focus on reporting the occasional bad actors who slip through the cracks of the platform’s automated review, rather than trying to combat the entire field of lead generators and their affiliates. The likely result would be more effective consumer protection, more meaningful company policies, and fewer users following ads to debt traps and financial fraud.
OECD, The Role of Internet Intermediaries in Advancing Public Policy Objectives, OECD Publishing, 2011, 139.↩
The challenge of dealing with ads for fringe financial products is still evolving, as evidenced by the fact that Facebook only recently revisited its own payday ad policies. Google, Bing, and other platforms have an opportunity to consider new approaches themselves. We urge ad platforms to engage with other stakeholders — including civil rights and financial advocates — in considering their options. In our view, meaningful new limits on payday loan ads are feasible, and are consistent with the values already reflected in the policies of major online advertising platforms.
Payday Lead Generators and Trade Groups
Large payday lead generators could make and enforce stronger commitments to limit the sharing and use of consumers’ data. Today, the Online Lenders Alliance (OLA) maintains the most visible set of best practices for the entire online payday ecosystem, including payday lead generators.134 These guidelines have some strengths, such as barring false or misleading statements and requiring certain disclosures.
However, the guidelines are notably permissive when it comes to the handling and resale of consumers’ data. They offer no concrete limits on the number of times a lead may be sold, and no prohibitions on sharing with non-lenders, unlicensed lenders, or third parties that have no legitimate interest in the data. And the guidelines recommend, but do not require, contractual limitations to protect leads as they move through the industry.135
Further, although the OLA requires companies to comply with federal and state laws to qualify for membership,136 it has not issued best practices clarifying when payday lead generators should, if ever, market payday loans in states where such loans are severely restricted or prohibited.
These guidelines could be revised to better protect the subjects of leads.
The FTC and the CFPB could exercise direct oversight over large lead generation companies. Both regulators have already sued fraudsters empowered by payday leads. However, these enforcement actions might demonstrate a need for closer attention to the payday lead generation industry’s handling of sensitive financial data more broadly.
The FTC has a broad and flexible grant of authority to police “unfair or deceptive acts or practices in or affecting commerce.”137 The Commission has already pursued lead generators and their affiliates for misrepresentations. In the future, it could consider using its authority to prevent widespread sale of sensitive data without reasonable safeguards.138 In its complaint against LeapLab, the Commission alleged that the unfettered sale of payday loan applications to non-lender third parties was an unfair practice when those purchasers actually resulted in fraud.139 However, in other contexts, the Commission has alleged that “failure to employ reasonable and appropriate security measures to protect consumers’ personal information” is itself an unfair practice (even when that personal information does not include sensitive financial data).140 Looking ahead, the Commission could consider when the widespread sale of sensitive leads triggers a similar standard, even if the lead generator did not have advance knowledge of a buyer’s intent to commit fraud.
Payday lead generators may also be subject to CFPB jurisdiction as “service providers” to lenders.141 The CFPB is charged with protecting consumers from harmful practices in the financial industry. Its jurisdiction includes lead generation companies that act as “service providers” to companies that offer consumer financial products or services.142 A service provider is an entity that provides a “material service . . . in connection with the offering or provision by such covered person of a consumer financial product or service.”143 The CFPB has already treated lead generators as service providers over which it has supervisory and enforcement authority.144
The CFPB is empowered to regulate ex ante — to prevent problems from occurring, rather than being limited to fixing problems that have already occurred.145 The Bureau is thus well-positioned to examine the lead generation industry and help spur the creation of new best practices. The CFPB could pay careful attention to the lead industry’s contractual weaknesses, and scrutinize whether lead sale practices could constitute an unfair or abusive act or practice. Eventually, the Bureau could also consider issuing rules governing the collection and resale of consumers’ financial data by service providers.
United States Government Accountability Office, “Information Resellers: Consumer Privacy Framework Needs to Reflect Changes in Technology and the Marketplace,” September 2013, available athttp://www.gao.gov/assets/660/658151.pdf. ↩
Facebook Advertising Policy, “Prohibited Content,” available athttps://www.facebook.com/policies/ads/#prohibited_content. (“Prohibited Content: Payday loans, paycheck advances or any other short-term loan intended to cover someone’s expenses until their next payday.”) Facebook adopted this policy in August of 2015. ↩
Microsoft requires that all ads for financial services “comply with all applicable local laws and regulatory requirements.” Similarly, Twitter requires ads “be compatible with all applicable law; provide necessary disclosures, balanced information of risks and benefits, and all information that must be provided to the investor,” be clearly identified as a financial service advertisement, and must also “indicate the nature and specific type of financial service.” See, e.g. Microsoft, Bing Ads, “Financial products and services policies,” available athttp://advertise.bingads.microsoft.com/en-us/financial-products-and-services-policies; see also Twitter Ads policies, “Financial services,” available athttps://support.twitter.com/articles/20170445. ↩
Google, Advertising Policies Help, “Approval (limited),” available athttps://support.google.com/adwordspolicy/answer/2684542?hl=en. (“An ad can be marked ‘Approved (limited)’ based on the specifications of our advertising policies and local legal requirements, such as those allowing certain types of ads to show only in certain regions, only to certain ages, or only on certain devices.”). ↩
An act of commerce or practice is unfair when it causes or is likely to cause substantial injury to consumers, cannot be reasonably avoided by consumers, and is not outweighed by countervailing benefits to consumers or to competition. See, e.g. Federal Trade Commission, “Policy Statement on Unfairness,” December 17, 1980, available athttps://www.ftc.gov/public-statements/1980/12/ftc-policy-statement-unfairness. ↩
The CFPB is granted its regulatory authority under Title X of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010. Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (Dodd-Frank) (12 USC § 5481), available athttps://www.law.cornell.edu/uscode/text/12/5481. ↩
This includes providers that design, operate or maintain the product or service, or that process transactions. It does not include ministerial or non-material support services offered to businesses generally and those who provide advertising space. Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (Dodd-Frank) (12 USC § 5481), available athttps://www.law.cornell.edu/uscode/text/12/5481. ↩
Lead generation is an old practice that has become more powerful and sophisticated in the digital age. The online lead generation ecosystem includes a variety of different actors, including online advertising platforms, commercial data providers, lead generation firms, and small affiliate marketers. Lead generation practices deserves special scrutiny when they are employed to promote potentially exploitative goods and services.
Payday lead generators promote risky online payday loans, help lenders skirt state laws, and can expose consumers to fraud. Today, payday lead generators are targeting consumers across the web, even consumers who reside in states where payday lending is illegal.
More can be done. We recommend that major online advertising platforms consider new approaches to payday loan ads, that the payday lead generation industry consider stronger best practices, and that the Federal Trade Commission and Consumer Financial Protection Bureau consider enhanced oversight of the payday lead generation industry.
In preparing this report, we spoke with payday lead generation firms, major advertising platforms, consumer advocates, and federal and state regulators. We also reviewed federal laws, state laws, state regulations and rulemakings, state and federal court documents, company policies, industry white papers, presentations, videos, research reports, and a variety of publicly-available forums and Internet relay chat (IRC) channels.
To explore how online payday lead generators were showing search ads, we ran a series search queries on Google and Bing (including, for example, “payday loan,” “need a loan fast,” and “need money to pay rent”) from internet protocol (IP) addresses originating in Pennsylvania, New York, and Vermont using a commercial virtual private network (VPN) service between July and October of 2015. We clicked on many of the payday loan-related advertisements that we saw during this process. On each attendant website, we reviewed policies and submitted test data, including address information consistent with the jurisdiction of the initial search.
This report was made possible through the support of the Ford Foundation.
We are grateful to the many organizations that provided valuable contributions and feedback on this report, especially The Leadership Conference on Civil and Human Rights, The Pew Charitable Trusts, Americans for Financial Reform, the Center for Responsible Lending, and the Center on Privacy & Technology at Georgetown Law. Thanks also to the public servants, company staff, and other stakeholders who took the time to engage thoughtfully with us throughout this process.