Interview with Artem Fedin – CEO of aff.studio

In this interview with AffPapa, Artem Fedin, CEO of aff.studio, shared why affiliate performance data is still scattered across a dozen operator dashboards, explained the reporting mistakes that cost affiliates money, and revealed the one metric partners still aren’t tracking.

Yeva: What was the main problem you wanted aff.studio to address when you first started the company?

We didn’t start with a company. We started with two problems inside our own affiliate business.

The first was data: our content team spent its week re-checking the same casino and game details and pasting them into pages by hand. The second was money: our numbers sat in a hundred separate partner cabinets, so one report meant logging into program after program.

Neither was hard. That’s what made it dangerous: death by a thousand cuts, and nobody’s job was to notice. What it cost us was the answer to the only question that matters: which programs actually pay, and where should the next unit of traffic go?

Roman, our founder, went looking for tools. Everyone he asked had the same problems, and nobody had a good answer, so we built our own: a database that keeps casino and game data verified and ready to publish, now CasinoMass, and a tracker that pulls every partner cabinet into one dashboard, now AffStata.

Both stayed in-house for years. We opened them up because of the thing sitting behind both problems: this industry’s software points at players. Almost none of it is built for the people marketing to them.

Yeva: With so much technology available, why do you think affiliates still have to manage reporting across multiple dashboards? What’s the main industry challenge behind that?

Because it isn’t a technology problem. It’s a standards and incentives problem, and no amount of new software fixes that on its own.

Every operator’s reporting is built for the operator. There’s no shared schema across affiliate platforms; “active player,” “NGR,” even “registration” mean different things depending on whose cabinet you’re in. Plenty of platforms still have no usable API, or one you have to request and wait for. Some make you fight 2FA and bot protection just to pull your own numbers out.

Underneath that is leverage. Operators have little commercial pressure to standardize, because the cost of not standardizing lands on someone else. And the affiliate side is too fragmented to demand it; a thousand publishers each solving the same problem privately in a spreadsheet never becomes an industry standard.

So the work gets done, just always by the affiliate. The real question isn’t whether it gets done. It’s whether a person does it by hand or a system does it.

Yeva: From your experience, what are the biggest mistakes affiliates make when reading or comparing performance reports? How can they tell when something doesn’t really add up?

Three mistakes, in order of how much they cost.

Comparing numbers that aren’t comparable. One program reports in UTC, another in GMT+2, a third converts currency on a different date. Put those side by side, and you’ve built a report that looks precise and isn’t.

Reading absolute numbers instead of ratios. Revenue went up because you sent more traffic. That tells you nothing about whether the program got better or worse.

Judging on one month. A single cycle can’t separate a real decline from normal variance.

As for spotting when something’s off: watch ratios over time. Click-to-registration, registration-to-first-deposit, deposit per player, yield per program. When one of those drifts steadily in one direction, and your traffic mix hasn’t changed, something changed on their side.

And this only works if you have your own baseline. If your only record of what happened is the operator’s record of what happened, there’s nothing to reconcile against; you’re not checking the numbers, you’re just reading them.

Yeva: Tracking has become much more complicated over the years. What issues do you see most often, and how can affiliates avoid them before they become bigger problems?

The specific failures are familiar: postbacks that fire late, fire twice, or never fire; cookie-based tracking degrading as browsers keep tightening privacy; mobile-click-to-desktop-deposit journeys that never link; deep links quietly breaking when an operator changes URL structure without telling anyone.

The pattern behind all of them is that tracking fails silently. Nothing errors. You just get a smaller number, and a smaller number looks like a bad week.

The reason it stays hidden is that most affiliates only measure one side of the transaction. Your tracker records the click and waits for the postback. The operator’s cabinet records the registration, the deposit, and the revenue on its own side, independently. While those two agree, you have nothing to check, and the failures live exactly in the gap between them. If you never put the two numbers side by side, the gap is invisible by design.

So the durable fix isn’t a better pixel. It’s a second source of truth to reconcile against. The hygiene still matters: lean on server-to-server postbacks over cookies where you can, keep sub-IDs to one naming convention with no exceptions, and test every new integration with a real end-to-end journey before you scale spend into it, but none of that catches a silent failure on its own. What catches it is reconciling your tracker’s version of the week against the operator’s, weekly rather than at month-end. A drop in attributed volume caught in three days is a bug; caught in five weeks it’s a lost quarter you can no longer prove.

Yeva: AffStata brings data from different sources into one place. From your perspective, what difference does that make for affiliates in their day-to-day work?

The difference it makes is what the week gets spent on. Reconciliation produces no decision; it’s data entry that happens to involve money. When the numbers arrive already aggregated and normalized, one timezone, one currency, consistent definitions, the same person spends those hours deciding which brand earns more traffic, which deal to renegotiate, which program to cut.

There’s a second effect that matters more than the time saved: you notice things. A program sliding for two months is obvious in a single normalized view and nearly invisible across fourteen tabs.

And it makes real comparison possible. You cannot rank programs you can’t put on the same axis, which is exactly the decision an affiliate makes every single day.

Yeva: AI and automation are becoming part of almost every business. How do you see these technologies changing the way affiliates analyze and manage their performance?

The honest answer is that the valuable part is boring. Most of the win right now is in collection and cleaning, pulling data out of platforms that don’t want to give it up, mapping inconsistent inputs to one schema, catching the row that doesn’t reconcile. That’s where automation earns its money today, and it’s not the part anyone puts on a conference slide.

The shift I find more interesting is timing. Affiliate analysis has always been retrospective; you learn in May what April was. Automation moves that toward the present: which program is drifting from its own baseline this week, while you can still do something about it.

One caution. A model sitting on top of unnormalized data will give you confident, well-written nonsense. The sequence isn’t optional: trustworthy data first, then analysis on top.

And none of it replaces the commercial judgment. Knowing why a geo softened, or which affiliate manager actually answers, is not in the dataset.

Yeva: Looking at the industry today, is there a metric or piece of data that you think affiliates still don’t pay enough attention to?

Player quality over time, by program. Retention and cohort value at 60 and 90 days out.

Almost everyone is measured on the front end: clicks, registrations, first deposits, because that’s what gets reported fastest and paid soonest. But an FTD tells you a brand converts your traffic. It says nothing about whether that brand keeps the player. Two programs with identical first-deposit numbers can be worth completely different amounts six months later, and the affiliate sending traffic to the worse one usually doesn’t find out for a year.

Two smaller ones. The gap between what a program accrues and what it actually pays: most affiliates track the first number and budget against it. And the cost of a program in effort: a strong rate that requires three chases a month and a manual invoice is more expensive than its headline percentage suggests.

None of these are hard to measure. They’re just harder than the metrics that arrive by default.

Yeva: As the CEO of aff.studio, what’s been the biggest lesson you’ve learned since launching the company?

That this business runs on trust far more than on features.

When we opened up tools that had been internal, we assumed the value would speak for itself: same industry, same problems, obviously useful. What we underestimated is what we’re asking for. An affiliate connecting a partner cabinet, or making traffic decisions on data we supply, is taking a real risk with their own revenue. That trust isn’t won with a feature list. It’s earned slowly, by being precise about what the product does and doesn’t do, and by not overselling the bits that aren’t finished.

The second lesson: the problem affiliates describe, and the problem costing them money are often different. They ask for time back. What actually hurts is allocating traffic blind. Those need different answers, and you only find that out by talking to far more affiliates than you think necessary, including the ones who tell you your product isn’t for them.

Yeva: Finally, when someone looks at aff.studio today, what do you hope they see beyond it being a reporting platform?

Infrastructure for the business of being an affiliate.

Affiliates get treated as a traffic channel, and the software follows that assumption — built to serve the operator’s view of the relationship. But an affiliate company is a company. It has margins, receivables, supplier risk, concentration risk, and an asset value someone might one day buy. Almost nothing in this industry is built for the person running it that way.

That’s the gap we’re working in. So what I’d hope someone sees is a company building for the operators of affiliate businesses rather than for the operators of casinos — tools and data made for the professional side of this market, which is a lot more serious than the industry’s software gives it credit for.

Reporting is where we started because it’s the plumbing, and nothing else works until the plumbing does. It isn’t where we intend to stop.