Want to read my step-by-step guide to automate the validation and anti-fraud measures in intake form submissions and affiliate sign-ups? Read this post.
In 2020, as the Special Projects Lead of FairShake, I took the initiative to develop and grow an in-house affiliate program for a company operating in the consumer legal services industry. I managed all the company's marketing, business operations, and advocacy channels in this role, reporting directly to the two co-founders. Working closely with a cross-functional team, I successfully established the affiliate program, which rapidly grew to become the second-largest acquisition channel for the company, only surpassed by organic search.
The company sought to diversify its acquisition channels and broaden the top of the sales funnel to reach more potential customers. Although we had a growing base of satisfied customers, we needed a structured system to turn them into active advocates for our product. Our product involved a complex claims submission and approval process, and we aimed to prevent the incentivization of low-quality claims that could burden the system.
To tackle these challenges, we devised a strategy to offer referral bonuses exclusively for claims approved and mailed to the respective companies. This approach encouraged affiliates to focus on high-quality referrals, mitigating the risk of effectively overwhelming our system with low-quality claims while promoting our product.
Due to the complexity of our claim approval process and the high 30% commission typically charged by affiliate marketing software providers, I built our affiliate program in-house using off-the-shelf tools and closely integrated it with our product database.
Initially, I offered a $15 referral bonus for approved claims against the 60+ companies we processed claims for. However, I quickly noticed that we received a disproportionate number of claims against common payment apps like PayPal and Venmo instead of against our target companies, which tend to be more responsive to arbitration claims, such as telecommunications companies and banks. To address this, I announced an increased bonus of $20 per approved claim against our target companies and reduced the bonus to $10 for claims against other companies in our portfolio. This change was well-received, prompting affiliates to optimize their outreach and expand into complaint groups, forums, and emerging channels like TikTok and Instagram Reels to target high-value claimants they previously missed.
I collaborated closely with the Head of Customer Success and their team, who reviewed each claim. We established a feedback loop to collect data on claim quality submitted through referral links. This data is piped into a dashboard in Amazon QuickSight, enabling us to advise individual affiliates and program participants on improving claim quality.
I initiated the project by researching affiliate programs in related consumer financial services verticals and examining promotional activities in Facebook Groups and complaint forums. This research informed our decision to build an in-house affiliate program using off-the-shelf tools and closely integrate it with our product database. As part of our overall strategy, we leveraged owned channels like in-app messaging, email, and social media to drive the adoption of the referral program and amplify its reach.
The success criteria we established for the affiliate program were based on several key performance indicators, including increased approved claims, diversification of acquisition channels, and controlled acquisition costs. We monitored the performance of our owned channels, such as the effectiveness of in-app messaging for user engagement, email open and click-through rates, and social media reach and engagement, to ensure they contributed positively to the program's overall success. By setting up success criteria encompassing these owned channels, we could continuously assess and optimize our strategies to achieve the desired results.
Throughout the execution phase, I issued payments to affiliates every other week using a dashboard in Amazon QuickSight. I exported the tally of approved and mailed claims from the preceding two-week period, integrated these values into a Google Sheet, matched affiliate codes with payment emails, and filtered out banned affiliates. The sheet, formatted for PayPal's bulk payouts system, was uploaded directly to PayPal for payouts. I also used the mail merge tool YAMM to send confirmation emails from the spreadsheet detailing the number of invalid and valid referral payments for each referrer.
By implementing these owned channels and continually optimizing their performance, I successfully drove the affiliate program across the finish line, achieving significant growth and impact.
I conducted strategic outreach and applied data-driven enhancements to expedite the program's growth. I targeted complaint groups on Facebook, such as "Optimum Complaints" and "AT&T Complaints," informing them that my company offered financial rewards for generating claims through our system. This tactic motivated people to share legal recourse options with others experiencing issues with these companies.
I also engaged influential Twitter accounts amplifying consumer complaints, such as @mrcomplaintbox, persuading them to join the program. Additionally, I collaborated with nonprofits like the Fair Internet Coalition, which collected consumer complaints against our target companies, to promote their participation in the affiliate program.
To refine the program, I developed a dashboard displaying claim success metrics for each affiliate. The data showed that about 30% of our "super-affiliates" consistently produced high-quality referrals with a greater likelihood of approval and claimants who remained engaged long enough to receive and accept settlement offers. To reward and encourage this behavior, I introduced the "Trusted Affiliate Program."
Upon generating 15 valid claims, affiliates could apply to join the Trusted Affiliate Program and gain immediate insight into their referral metrics. Trusted affiliates received notifications each time a claim was submitted or approved, including the name of the company involved and the claimant's first name. We limited this transparency to trusted affiliates to prevent raising new affiliates' hopes—and potentially encouraging unnecessary inquiries—when they did not receive a bonus for every claim submitted via their link. This data-driven method fostered continuous program improvement and heightened the overall efficiency of claim submissions and resolutions.
Four significant challenges arose during the program: referral fraud, a technical error in how Amazon QuickSight pulled data from our product, inefficiencies due to a lack of automation, and maintaining compliance on social media.
1. Referral fraud: Monitoring social discourse, I discovered that some affiliates incentivize people to submit claims through their referral links, leading to low-quality claims and program rule violations. Additionally, many affiliates submitted claims using their referral links.
To address these issues, I employed automated checks and social discourse monitoring tools to detect and suspend affiliates violating our terms. After multiple violations, we banned these affiliates from participating in the program. I also included email and IP address checks to prevent affiliates from signing up for the program multiple times. To combat self-referrals, I established automated reviews comparing the contact information of the referrer and the claimant, sending warning emails, and withholding payment for policy violations. These measures improved claim quality and expanded to detect duplicate claims and claims from outside the United States beyond our representation scope.
2. Technical error in Amazon QuickSight: I noticed that the number of referral bonuses exceeded the number of claims received through the referral channel due to a glitch in Amazon QuickSight that double-counted manually reapproved claims. This error led to overpayment to affiliates.
To resolve this issue, the CEO and I forgave overpayments under $300, covering 95% of our affiliates. For our "super-affiliates" with more significant overpayments, we communicated our mistake and implemented a repayment program, deducting 50% of their future earnings to recoup the overpaid funds gradually. While some affiliates were dissatisfied, most affiliates continued generating referrals until their balances were paid off.
3. Inefficiencies due to a lack of automation: Not automating more of my workflow early on led to increased manual workload, delays in processing payouts, slower response times to fraud incidents, and limited scalability. As the program grew, these inefficiencies consumed more of my time, making it challenging to focus on higher-level strategic tasks and improvements to the program.
To resolve this issue, I implemented several automation solutions to resolve this issue to streamline operations and enhance overall performance. First, I configured the dashboards and lookup spreadsheets to easily export payouts for bulk processing and send mail merge emails to affiliates, confirming their statistics for each period. This automation significantly reduced the time spent on manual tasks and facilitated faster payouts.
Additionally, I automated communications to guide affiliates in soliciting better claims. When a claimant submitted a short or non-US-based claim, the affiliate received an automated email to coach them on our claim quality guidelines and rules necessitated by the American Arbitration Association. This system ensured that affiliates received timely feedback, improving claim quality and reducing the need for manual intervention.
4. Maintaining compliance on social media: Ensuring affiliates adhered to guidelines and best practices on social media proved challenging. Monitoring social media conversations, I identified instances where affiliates deviated from our guidelines or shared misleading information about the program.
To overcome this challenge, I actively monitored social media channels and set up alerts to track conversations related to our program. When necessary, I contacted affiliates who did not comply with our guidelines, providing them with clear instructions on adjusting their promotional efforts. By maintaining an active presence on social media and promptly addressing compliance issues, I ensured that the program held a positive reputation and aligned with our company's values.
These solutions improved the program’s efficiency and allowed me to focus on strategic improvements, enhancing the program's overall performance and scalability.
The Affiliate Program played a significant role in the company's growth, becoming the second-largest acquisition channel and leading to a considerable increase in claims processed. This success broadened our user base and extended our reach into untapped networks of potential customers.
Our project success criteria included diversifying acquisition channels, maintaining high-quality claims, and keeping the program's cost within acceptable acquisition cost ranges. The Affiliate Program exceeded expectations in all these aspects, with ongoing adjustments like differentiated referral bonuses and the Trusted Affiliate Program contributing to its continued success.
During 2020 and 2021, the affiliate program consistently ranked as our second-highest acquisition channel, closely following organic search. The program's ability to rival this crucial acquisition channel demonstrated its effectiveness, even becoming the company's top acquisition channel during periods of heightened interest.
The affiliate program's customer acquisition cost consistently outperformed other paid channels, such as search and remarketing ads, while generating valuable backlinks and promoting conversations about our company's dispute resolution successes. Affiliate feedback further reinforced the program's positive impact, revealing high satisfaction levels and highlighting the benefits of empowering individuals and helping victimized consumers.
If I were to approach this project again, I would:
This work is licensed under a Creative Commons Attribution 4.0 License.