Always Be Testing Podcast – Episode #101

Smarter Affiliate Marketing Starts Here | Kayla Castro of Zenni Optical
Host: Tye DeGrange
Guest: Kayla Castro, Affiliate Partnerships Lead, Zenni Optical
Date: September 23, 2025

Episode Overview

This episode offers a deep dive into the realities, challenges, and modern best practices of affiliate marketing. Host Tye DeGrange is joined by Kayla Castro of Zenni Optical, a seasoned affiliate leader. Together, they explore the pitfalls many brands fall into, evolving measurement techniques (MMM, MTA, incrementality), the influence of AI, the importance of authentic partnerships, and the often-misunderstood nuances of effective affiliate programs.

Key Discussion Points & Insights

1. Evaluating and Auditing Affiliate Programs

• Early Steps:

  • Kayla’s first move in a new affiliate role is a thorough audit of the program—analyzing contracts, existing partners, and performance distribution.
  • “Most brands really think they have a healthy affiliate program because the overall numbers look great. But really there’s like three or four partners driving 99% of the sales.” — Kayla [01:46]

• Common Pitfalls:

  • Overconcentration: Overreliance on a few “deal” or “coupon” partners.
  • Missed Opportunities: Gaps where low-hanging, high-potential partners are missing.
  • Lack of Engagement: Smaller partners may not be interacting with the brand at all (e.g., not being included in newsletters).

2. Internal Communication & Executive Buy-In

• On Speaking the Language:

  • Success stems from internal advocacy—translating affiliate’s value to finance, execs, and other teams.
  • “Not taking that extra effort to learn those departmental languages inside a company can be a massive mistake down the road.” — Kayla [02:54]

• Perception Problems:

  • Affiliates are often mischaracterized as “last-click grabbers” and not incremental contributors.
  • Changing perceptions requires cross-team relationship-building and education.

3. Measurement: MMM, MTA, and Incrementality

• Misunderstandings Abound:

  • There’s confusion—even among senior marketing and data professionals—about MMM (Media Mix Modeling), MTA (Multi-Touch Attribution), and incrementality.
  • “People don’t understand the difference between MMM, MTA, incrementality as a whole, but they’re really different measurement tools.” — Kayla [06:48]

• Unique Affiliate Challenges:

  • Affiliate doesn’t always fit standard measurement tools, especially for content-heavy programs.
  • Ability to AB test or use geo-targeting may be limited with large content partners.

• In-House Solution:

  • After interviewing third-party vendors, Zenni built its own MMM solution, prioritizing close collaboration between affiliate and data teams.
  • “Having an actual like another three people on the data team…weekly calls…I will take the time and do affiliate 101 with you.” — Kayla [10:06]

• Defining Incrementality:

  • Sometimes, blunt tests (e.g., cutting non-incremental partners) are used:
    • “We cut the top three partners who were all deal sites and they, everyone freaked out… Then they went to the data team, looked internally…actually, no, we didn’t see sales drop for the day.” — Kayla [13:41]

4. Attribution Models & Partner Rewards

• Going Beyond Last Click:

  • Kayla advocates for flexible attribution—rewarding creators and content partners, not just those who get the last click.
  • “We will repurpose [creator] content into ads… that’s a value that’s not necessarily tied to a sale.” — Kayla [18:09]

• Custom Models:

  • Experiments with weighted attribution, contract-level tweaks, and seeking smarter automation.
  • “I think there’s still a tool to be developed that’s a little bit more of an in-between, where it does require some human logic and placing value with the partners as each brand sees fit.” — Kayla [19:30]

5. The Slow and Steady Nature of the Affiliate Channel

• Efficiency vs. Volume:
  • Affiliate is efficient, but not as instant as channels like search or paid social. Onboarding major partners or launching big initiatives is relationship-driven and takes time.
  • “[People say] if we give you five more million dollars this year, what can we do? ... I can’t just immediately…turn around a partner that’s going to generate a high volume of sales.” — Kayla [23:46]

6. Artificial Intelligence & Workflow Optimization

• Current State:

  • AI isn’t fully integrated into affiliate networks’ day-to-day, but excels at assisting with templates, competitive research, and aligning affiliate goals to company KPIs.
  • “I use AI to help me…draft templates, competitor research, fix workflow efficiencies…using AI to effectively figure out how to translate affiliate to the company’s KPIs…” — Kayla [25:42]

• The Human Element:

  • “The authentic trust isn’t coming from AIs and algorithms… it’s coming from humans… third parties are what we are leaning into completely…” — Tye [28:03]

7. Building Trust & Authenticity with Third Parties

• Why Trust Matters:
  • Authentic relationships—internally and externally—drive results. If trust is broken with affiliates or with the brand’s audience, recovery is difficult.
  • “There are affiliates that have really burned me and…that trust is really hard to build back, especially…they’re virtual relationships.” — Kayla [29:28]

Notable Quotes & Memorable Moments

• On Executive Narratives:
  “You could be doing all the right things to a table. But if that message isn’t getting heard throughout the organization then really doesn’t matter.” — Tye [03:56]

• On Defining Incrementality:
  “Blunt force proving incremental value of a partner. If these were incremental sales, we would see the site suffer.” — Kayla [13:41]

• On Attribution Tools:
  “Now we can…tweak the system in the contracts such that [creators] get higher priority for getting attribution for that sale.” — Kayla [18:09]

• On Program Management:
  “Once you’ve got a healthy affiliate program, it is more of a slow and steady growth channel. Not every partner … is going to be in a top three immediately.” — Kayla [23:46]

• On AI's Role:
  “[AI is] not an end all be all necessarily yet, which may not be a bad thing.” — Tye [27:57]

• Book Recommendation:
  “Dungeon Crawler Carl is like Dungeons and Dragons meets an action movie, and it’s wonderful.” — Kayla [31:32]

Key Timestamps for Important Segments

• Audit Process & First Steps: [01:46]
• Common Mistakes & Internal Buy-In: [02:54–05:09]
• Measurement Misconceptions (MMM, MTA, Incrementality): [06:48–09:00]
• Building Custom Measurement Tools: [10:06–12:34]
• Proving Incrementality by Dropping Partners: [13:41–15:59]
• Attribution Techniques Beyond Last Click: [18:09–20:24]
• Affiliate Channel Pace/Efficiency Discussion: [23:46–25:22]
• How Kayla Uses AI: [25:42–26:56]
• Human Trust and Authenticity in Affiliates: [28:03–30:46]
• Book Picks & Personal Questions: [31:15–32:53]

Final Insights

• Affiliate is not “set it and forget it”—constant education, advocacy, and measurement refinement are essential.
• Measurement is nuanced—no one-size-fits-all tools, and sometimes internal builds work best.
• Attribution models must evolve to match diverse partners and objectives, not just reward the last click.
• AI is a valuable tool to augment work, but the affiliate ecosystem remains deeply human and relationship-driven.
• Trust and authenticity—both in brand/audience and brand/partner relations—are vital for sustained success.

Connect with Kayla Castro:

• LinkedIn: Kayla Castro
• Email: kayla@zenni.com