TrafficGuard: The World Cup exposed the hidden cost of player acquisition

Matt, TrafficGuard estimates that more than $100m of betting advertising spend delivered no new player value during the World Cup. Was that higher than you expected?

It was probably in line with what we would expect. Based on our market data, we estimate that the online gaming industry spends somewhere between $4bn and $8bn every year on Google PPC. It is important to stress that Google PPC is typically one of the largest, fastest and most successful ways for sportsbook operators to acquire players.

The issue is that within that highly effective spend, you also capture existing customers who are simply returning to log in and would have returned anyway. Sports betting also attracts a significant amount of bot traffic.

So when spend increases around a major event like the World Cup, we would expect the absolute amount being spent on returning users and bots to increase as well. Interestingly, some operators actually reduce budgets during flagship events because they know their existing players are going to return anyway. Others double, treble or quadruple their budgets because they see it as a huge acquisition opportunity.

The challenge is making sure as much of that additional investment as possible reaches genuinely new, real users.

How much advertising spend typically goes towards existing customers or invalid traffic?

It varies enormously by market, operator, format, maturity and market share, but there are some useful high-level benchmarks. When we start supporting a sportsbook, we would typically expect somewhere around 10-20 per cent of spend to be reacquiring customers the operator has already acquired and who would have returned anyway. Then you might have another three-five per cent attributable to bot traffic, which obviously provides no value whatsoever, which represents a material efficiency opportunity. If you can redirect that money into working media that genuinely drives new-user acquisition, first-time deposits and ultimately customer lifetime value, the return can be significant.

Does that mean conventional measures of campaign success can be misleading during events such as the World Cup?

The important distinction is between causality and correlation. If your business is growing, revenue is increasing and media spend is increasing alongside it, your attribution models and platforms may tell you that advertising has driven more users. If you only look at those aggregated, surface-level metrics, you can conclude that all of that advertising investment is incremental to growth. But you need to look underneath that and ask which engagements actually caused a conversion and which conversions would have happened anyway.

A user might click an ad and subsequently deposit, but that does not necessarily mean the click caused the deposit. This is where incrementality becomes important. Would I have had this player, registration or first-time deposit without that advertising engagement? It’s a much more useful question than simply identifying a correlation between clicks and conversions.

So how should operators judge whether their World Cup acquisition campaigns worked?

Immediately after the event, the numbers can look fantastic. You’ve increased budgets, more people are searching, more people are registering and more people are depositing for the first time. As a performance marketer, you can look like the smartest person in the room. The real proof comes later. What is the deposit value and lifetime value of those players after 30 days, 60 days, 90 days or a year, and how does that compare with the customer acquisition cost?

During the World Cup, everybody is bidding and prices rise. You might pay significantly more to acquire somebody who deposits once and never comes back. Immediately after the tournament, that isn’t necessarily visible in the headline numbers. What matters is how many of those players stick around and whether their lifetime value justifies what you paid to acquire them.

Google and other major platforms invest heavily in preventing fraudulent traffic. Why does invalid traffic remain a problem?

I would compare it with programmatic advertising. Nobody would buy through a DSP without third-party verification in place. It is simply considered best practice. We’re not saying the platforms themselves are fraudulent or deliberately generating invalid traffic. But everybody accepts that fraudulent and invalid traffic exists.

If you’re an advertiser buying traffic, ultimately you’re buying revenue. It makes sense to have an independent verification partner whose sole remit is to verify what you’re buying and give you the ability to prevent waste.

Publishers and platforms will rightly say they’re investing in tackling the problem, but this isn’t necessarily the one thing they wake up every morning thinking about. We sit between the advertiser and the channel to provide that transparency and verification.

Are the biggest-spending operators inevitably the most exposed?

The more you spend, the more you’re exposed, simply because there is more opportunity for budget to be wasted. Two per cent of $1,000 might be considered a cost of doing business. Two or three per cent of $10m is extremely material. It’s money that could otherwise have become working media budget and generated a return. There are also differences between markets and verticals. Promotions and bonuses, for example, can attract bot traffic, so operators running those campaigns can be particularly susceptible.

There is further issue around returning customers. As an operator gains market share and a market becomes saturated, you have to ask whether another 10 or 20 per cent increase in spend is reaching 10 or 20 per cent more new users, or whether you’re simply reaching the same users more frequently. At some point you hit diminishing returns.

Is the invalid traffic problem getting worse?

There are probably three things happening simultaneously. First, invalid traffic has always existed. Historically it was often accepted as a cost of doing business, whereas marketers are now paying much more attention to it.

Second, platforms have increasingly moved towards smart bidding. You give the platform more control over optimisation because it has enormous amounts of data, but the quid pro quo is that you have less transparency. Marketers are therefore saying: I’ll use smart bidding, but I still want to understand the incrementality of every click.

Third, the internet itself is becoming increasingly automated and bots are becoming much more sophisticated.

Historically, most bot traffic would identify itself in one way or another. Now bots can come through an ad and interact with a site in ways that look remarkably similar to a legitimate user. Detecting them can involve analysing things such as mouse movements, form filling and device IDs. It is a cat-and-mouse game. Bot sophistication increases and detection technology has to become more sophisticated alongside it. It is an arms race.

Is the World Cup an exceptional case, or should operators be applying these lessons throughout the football season?

It’s an always-on opportunity. The level rises and falls with spend and major events, but inefficient spend exists on a month-to-month and annualised basis. Marketing teams spend days and weeks optimising creative, bid prices, partnerships, landing pages and the rest of the funnel to achieve a one percentage point improvement in conversion.

If you can automate the prevention of three-five per cent of invalid traffic and reduce another five-10 per cent of spend going towards users who would have returned anyway, that’s a relatively straightforward efficiency opportunity. The World Cup magnifies it because budgets are so large, but it doesn’t disappear when the tournament ends.

Latin America attracted enormous advertising investment around the World Cup. Are emerging markets particularly vulnerable?

The best approach is to establish best practice from the beginning and make it part of the standard toolkit. If you’re scaling with returning-user mitigation already in place, analysing lifetime value against acquisition cost at player level and preventing bots, you’re building efficiently from day one.

Brazil is an obvious example of a market where there has been a huge land grab to acquire as many users as possible. That makes commercial sense – invest now and seek the payback later. But you often see a boom-and-bust cycle where businesses spend two or three years focusing heavily on acquisition and then suddenly decide they need to become profitable. They then have to wind everything back.

The opportunity is to compete aggressively for users while baking incrementality and efficiency into that growth from the beginning. More broadly, I think the industry has moved from indexing heavily towards user acquisition to thinking much more seriously about profitable user acquisition.

If the efficiency argument is so clear, what stops operators addressing it?

Part of it is simply priorities. Marketing is demanding. There are countless tools, platforms, analytics systems and problems competing for attention. The other part is KPI alignment. If nobody is asking whether top-line user growth is profitable, why would somebody necessarily focus on it? If the KPI is simply spend and growth, and everything is increasing, nobody may look particularly closely underneath those numbers. Once a business has a clear mandate around efficiency, profitability and incrementality, it becomes a much more obvious opportunity.

Finally, how could AI-driven search change sportsbook customer acquisition?

AI is going to fundamentally change how people interact with the internet. Right now, however, we’re in a period of coexistence. People are reaching online properties through LLMs and AI agents, but they’re also continuing to use traditional search. Search isn’t going to disappear this year or next year, and some behaviours will persist for a very long time. But there is clearly a transition underway.

For operators, the practical implication is that historically they have optimised for SEM and SEO. Increasingly they’re going to have to optimise for agentic traffic as well – making sure they appear within AI-generated results and understanding how those users reach their properties.

So, operators will need to think about SEO, SEM and agentic discovery simultaneously, while keeping pace with how those models develop. The way people use the internet is becoming more complex, and acquisition strategies will have to evolve with it.