Annie Gateva has joined Reward Gateway | Edenred as Head of Partnerships – Emerging as part of the business' investment in its UK affiliate team.
Having worked both on the client-side and earlier in her career at Reward Gateway | Edenred, Gateva brings a view of the industry from both sides of the table.
As Q4 plans get finalised and brands look for ways to prove ROI while tracking-based attribution gets harder, we asked Gateva what that experience has taught her and how brands and publishers can work better together.
What are the big trends you're seeing directing spend in the affiliate space right now?
Advertisers want to know the same three things: show me your real reach and numbers, where you sit in the purchase journey and what spend is genuinely incremental. What's changed is how confident brands now are in answering them. Success has moved well past last-click conversions and more advertisers are putting budget behind non-last-click models. The reason they give is usually pretty practical. They want access, and guaranteed visibility at the moments that matter to them. That feels like a much healthier conversation than the one we were having a few years ago.
Having sat on both sides, what do brands get wrong about how publishers actually work, and what do publishers get wrong about brands?
When I joined a brand to build their affiliate programme from scratch, my marketing director expected affiliates to do everything. Find new customers, educate them, nurture them, then convert them. But we only paid on the last bit. We were also quite happily paying flat fees to content sites, directories and offline media to do the introduction and education part. With affiliates, the assumption was that all of that came free of charge and we'd settle the score at the point of sale. Nobody was being unreasonable, it was just two budgets sitting in different places under different rules.
So the thing I'd want advertisers to take from it is that different publishers do different jobs and they can only deliver against what they're rewarded for. Affiliates have a wide toolkit now, from content and education through to offer-led engagement, email, app notifications, behavioural automation and audience segmentation. Most can do the full job if they're treated as a channel rather than a cost of sale at the bottom of your funnel.
What publishers get wrong is assuming one size fits all, and that goes for their users too. Some people convert in-app, some on web, some off the back of an email, many go through all three. Every tool we have behaves differently depending on who the advertiser is and how their customers buy. The job is to match those up, test, adjust and keep going until it fits.
A lot of campaign planning and spend is calendar-driven, built around predictable moments in the year. How can brands differentiate themselves rather than following the same playbook as everyone else? Specifically, how do you approach Golden Quarter, the peak spend period?
Our traffic climbs the closer we get to the “Golden quarter,” and then again to “Golden week”. That's exactly why we build our own moments through the year and into early Q4, so advertisers have somewhere to land outside the traditional weeks.
When brands back those campaigns it does two things at once. It pulls demand forward and over a few years it's truly retrained our members that every month has something worth waiting for. We now see traffic start building from September, with members planning their spending around our calendar rather than just Black Friday. And it also gives an edge to brands and a place to stand out outside of traditional peak periods – with themed campaigns that help support their own seasonal narrative.
Peak weeks still need investment and I wouldn't pretend otherwise. That's how a brand holds share and stays visible while every competitor is also investing. My advice is to commit early and buy narrow. Premium visibility sells out, and the longer a brand leaves it, the more they pay for whatever's left, which is usually the point where the ROI stops working. A few good placements booked in decent time will beat a November scramble every time. Our research report “The Offer Economy: Golden Quarter 2026” gives some great insight from our members about how and when they spend and which offers work best for which categories.
Marketing mix modelling is the conversation everyone's having right now, as tracking-based attribution gets harder. From your experience on both sides, where does MMM fall short for brands trying to make real decisions?
MMM could be a genuinely good thing for this industry. I used it myself when I was managing affiliates for a brand. However, it’s important to drive home two points:
First, there's no standard for how affiliate goes into a model. Everyone builds their inputs differently, reads the outputs differently, then translates them differently again for their own leadership. It's like solving the same equation using maths from different planets, with everyone walking away certain they've got the right answer. And plenty of brands are still entering affiliate as a single line, when cashback, content and loyalty do completely different jobs. Average them together and the model can't say anything useful about any of them.
Second, no model sees influence. People research, compare and decide without clicking, particularly now trust in click-based advertising is low. Publishers have got much better at the discovery and education work that leads to that purchase decision, and none of that surfaces properly. Not in MMM, not in network tracking, often not even in internal advertiser reporting either.
So my only ask of advertisers is this: Before they act on a model that tells them to cut affiliate, they should go and look at how affiliate was entered into it. If the answer is one line, they haven't measured a channel, they've measured a label.
What does genuine audience insight give a brand that a pure tracking-and-modelling view can't?
A model can tell a brand what happened. It's far less good at telling them why, and the why is what they need if they want to do it again on purpose.
What we can give them is how people actually behave inside our channel. Our members are logged in, they shop around, they compare, and they buy when they're confident they've found the best version of a deal or product. That behaviour has a certain shape, timing, it has categories that over-index, and it has a clear sensitivity to how an offer is built. Knowing that tells a brand when to show up, where to sit and what offer to lead with.
Tracking and modelling will show a brand where something worked. Audience insight is what lets them understand it well enough to repeat it and scale it. It’s exactly that insight that we provide through our Offer Economy reports, where we ask our members how they plan to spend and compare that to what they actually did.
What should a strong working relationship between a brand and a publisher actually look like, and how can brands get more out of that relationship than they currently do?
Brands should look at their publishers individually. A partner list treated as one block is a partner list nobody is getting much out of. It's worth working out what each one is genuinely good at, briefing them on that, and rewarding them for it.
Our job is to behave like a consultant rather than a seller. That means understanding their purchase journey, what they're measured on internally, and where we realistically fit in their calendar. If we don't know their goal, we'll optimise towards ours, which helps nobody.
The thing brands could take more of is honesty. I'd like them to tell us when a campaign didn't land, and we should be telling them when a placement isn't right. I'd rather lose a booking than sell something that won't work, because I'd like them back next year. The partnerships that get the most out of us are the ones where both sides are comfortable saying the awkward thing.
