If you're still auto-approving affiliate applications or letting every partner type into your program without a clear strategy, this episode explains exactly why that approach will cost you this year. Ishtvan Torpoi, known industry-wide as “that affiliate guy,” joins Lee-Ann to unpack the MegaLag controversy, distinguish between coupon browser extensions and cashback models, and reveal why the vast majority of affiliate management work happens after launch, not before. This conversation delivers the strategic framework you need to build programs that actually drive incremental value rather than just revenue on paper.
The MegaLag video exposed how Honey's coupon browser extension operates, but Ishtvan learned the hard way that you can't trust sales pitches without testing. Five years ago, he evaluated Honey's claim that they improved cart abandonment rates. After analysing data before and during their program participation, the improvement was less than 1%. When factored against higher discount costs and commission payments, the overall business impact was negative.
The critical distinction: cashback browser extensions operate completely differently than coupon extensions. Cashback platforms maintain active databases of millions of users with verified personal information, home addresses, and payment details. They start customer journeys more frequently because users actively search their platforms for offers before shopping. One recent US client set lower commissions specifically for Rakuten's browser extension triggers versus website-initiated purchases because the browser extension represented lower-intent impulse clicks while website traffic showed genuine product research.
Auto-approving affiliate applications in 2026 invites disaster. With AI generating fake websites within days, brand bidding and cookie stuffing attempts flood every program. One wellness app launch attracted an affiliate who generated 800 app installs in the first day but only three purchases. That conversion ratio had never occurred in any marketing channel for that brand. By watching data like a hawk during new launches, they caught the bot farm before significant damage occurred.
Media buyers often get declined by automated systems because they register with affiliate links rather than owned websites. The algorithm sees insufficient traffic and rejects valuable partners who could drive significantly higher volumes than individual content creators. Winter sports affiliates get flagged for inactivity in July when dormancy is expected for seasonal inventory. These contextual failures cost programs their best partnerships because nobody questions the algorithm's recommendations.
Program terms need explicit guidelines about AI usage. If you allow voice replication for language localisation, specify that factual claims must remain accurate. If you permit AI-generated creative assets, require disclosure. If you work in regulated industries, make explicit that AI tools don't exempt partners from compliance requirements.
For new e-commerce programs, the question isn't which partners to include but rather what gaps exist in your current marketing. If you're not running retargeting internally, a retargeting affiliate removes that risk from your budget while generating fast results. If you're already running retargeting, adding an affiliate creates double-dipping that increases costs without incremental value.
Cashback platforms almost guarantee sales for brands with strong product-market fit and brand recognition. CSS affiliates with product feeds deliver value despite hesitation from paid teams who don't understand that Google's own documentation confirms you're not bidding against yourself. The typical e-commerce program can leverage 18-19 different affiliate types, but only after understanding which channels you're already covering internally.
For B2B and SaaS brands, the sales cycle demands different models. One client transitioning from B2C to B2B needed to implement cost-per-lead payments because longer sales cycles with sales team involvement made cost-per-sale models unattractive for quality affiliates. US brands generally understand customer lifetime value better than European brands, making them more willing to pay higher initial commissions knowing the reactivation and nurturing campaigns will generate repeat purchases.
Brands doing £30 million annually still lack proper tooling to understand customer journeys. They're not spending the £1,000 yearly investment that would provide insights eliminating hours of manual work. Smaller brands see the £100 monthly tool cost as unnecessary rather than foundational. The mindset problem isn't budget, it's understanding that data becomes sexy when you're trying to achieve incremental margin increases.
The brands succeeding have five analysts and data scientists building proper data structures. They're achieving high accuracy rates between network data and backend data because they're not depending on black box systems. One client recently realised they needed backend data infrastructure improvements before making any program changes, and an affiliate manager had to push that conversation rather than internal management driving it forward.
When using data analysis tools, start with specific problem statements. Why do conversions drop on Wednesdays? Which partner types perform best with discount codes versus content features? This focused approach produces actionable insights rather than overwhelming pattern recognition that may not matter to business objectives.
Affiliate managers at general digital marketing events still get treated like support agents rather than skilled strategists. The reality: a single day might require negotiating deals (sales skills), optimising CSS performance (understanding Google algorithms), fixing tracking issues (Google Tag Manager and Shopify technical knowledge), and analysing eight different partner types simultaneously.
Read affiliate network documentation thoroughly. Test new affiliates as actual users, not just by reading their sales pitches. Learn developer tools including application, console, and network tabs. Understand how tracking implementations work across different platforms. Talk to paid specialists in other channels to comprehend how display, SEO, Google, and Meta teams operate because affiliate marketing encompasses all of those tactics within one channel.
The managers who succeed in 2026 won't be relationship builders alone. They'll combine partnership skills with technical knowledge, data analysis capabilities, and strategic thinking that connects affiliate performance to overall business profitability rather than just revenue numbers.
[12:10] How to determine which of the 18-19 affiliate types make sense for new program launches based on internal marketing capabilities and desired speed to revenue
[22:00] Why brands at every size from startups to £30 million operations struggle with data aggregation and proper tooling investments
[27:45] The three non-negotiables for 2026: watching data like a hawk, changing default payout models, and implementing compliance monitoring
[30:50] Quality control processes that prevent fraud, including conversion validation checks and why auto-approval creates cleanup work rather than preventing problems
Huge thanks to Ishtvan Torpoi for sharing the frameworks he implements across diverse client programs at every stage of maturity. If this episode clarified where your program strategy needs adjustment, subscribe to the Affiliate Marketing Podcast so you catch every tactical insight that helps you build programs delivering genuine incremental value. Join us at the ELEVATE Summit in July 2026 in London, where strategic conversations happen between sessions and partnerships form that transform programs. Early bird tickets launch in January.
“I love the Affiliate Marketing Podcast.” If that sounds like you, please give us a 5 Star rating here! Taking the time to do that helps us support more people in our community to access affiliate marketing insights, expert-led learnings, and allows us to share the latest tactics that help affiliate programs and businesses grow. Click here, scroll to the bottom, tap to rate with five stars, and select “Write a Review.”