In the UK, one paid click on “betting sites” costs an advertiser about $71 on Google. A click on “car insurance”, one of the most fought-over finance terms in the country, costs about $9.
Those figures come from Google Ads keyword data for the United Kingdom, pulled through DataForSEO in October 2026. Other casino terms sit between $50 and $60 per click, while the main consumer finance searches stay under $9. Gambling clicks cost roughly six to eight times more.
Part of that premium is earned, because a depositing player is worth a lot. Part of it comes from how the two largest ad platforms decide who is allowed into the auction, and how little control they leave advertisers once they are in. This article looks at where the extra cost comes from, how programmatic buying through a self-serve demand-side platform (DSP) changes the equation, and where it doesn’t.
Why are gambling clicks so expensive on Google?
The obvious answer has truth in it. Player lifetime value in casino and sports betting is high, so operators and affiliates can afford to bid more than a car insurer can for a single quote request. High prices on their own don’t prove anyone is overpaying.
The less obvious answer is who gets to bid. Google allows online gambling ads only in listed countries and only from certified advertisers, who apply separately for each country. Meta requires prior written permission and proof of a license in each territory, including for affiliate pages that promote gambling.
That leaves few eligible bidders on one surface. In the same UK data, Google rates competition for gambling terms as medium and for car insurance as high, yet the gambling bids are the highest in the sample. My reading of that combination is that a small group of certified advertisers with deep budgets is bidding hard for limited search space. When that happens, price stops tracking value closely.
What does the walled garden add on top of the price?
Cost per click is the visible part. The less visible costs come from how much control an operator hands over once it is inside.
Certification is per country and can be withdrawn if a license lapses, so each new market means paperwork and a wait before the first ad runs. The platform’s own systems shape targeting. Account suspensions are common enough in this vertical that many performance teams plan for them.
Regulators are starting to treat platform opacity as a compliance risk. In March 2026 the Dutch regulator, the Kansspelautoriteit (KSA), published a checklist: license holders must prove with auditable data that at least 95% of the people their ads reach are over 24. According to iGamingExpert, the KSA also named social media as a weak spot and said campaigns must be withdrawn if a platform cannot guarantee that result. An operator that cannot see where its ads ran cannot easily prove compliance.
For an operator, the practical cost is twofold. You pay a premium to enter a small auction, and you get limited evidence about where your ads ran and why the algorithm chose them.
Is programmatic actually open to gambling advertisers?
It is, with conditions, and the conditions matter more than most pitches admit.
Google’s own exchange for programmatic buyers, Authorized Buyers, lists 38 countries where gambling ads can run in open and private auctions, including the UK, Germany, Spain, the Netherlands and the United States. Ads there must target approved countries, never target minors, and lead to a page with responsible gambling information. Outside those countries, gambling ads are rejected unless they run through a direct deal the publisher has opened to gambling. Other supply-side platforms (SSPs) and individual publishers set their own rules, and many publishers block the category by default.
So programmatic is not a loophole. It is a different supply pool with its own gates. The difference is who controls them: in a DSP, the buyer chooses which exchanges, publishers and deals to work with, and can keep a list of approved sites (an allowlist) that accept gambling demand in each market.
No channel helps where the product itself is banned. Brazil is the current example: a provisional measure in September 2026 banned online betting and its advertising, and operators had to take their sites offline within days.
What changes when the operator buys programmatic in-house?
A self-serve DSP gives the operator’s own team the controls a walled garden would otherwise hold. Four of them matter most for gambling.
The first is supply. The team decides which publishers, apps and private marketplace (PMP) deals receive bids. Sports content, betting-adjacent media and adult-skewed sites can be bought directly instead of hoping an algorithm finds them.
The second is bid logic. The team sets the rules: which placements, hours and GEOs to bid on and how much, including the event-driven peaks that dominate sports betting. The DSP then bids automatically within those rules, instead of a platform’s algorithm deciding both.
The third is measurement. Conversions can be passed back from the operator’s own server and feed bid rules the team sets itself, which matters when browser-based tracking loses a large share of events.
The fourth is evidence. Placement-level reports show exactly where an ad ran, which is one part of the record a regulator like the KSA now asks for. The other part, proof of who was reached, still has to come from audience data.
Epom DSP is one example of a self-serve platform built for this model. The operator or affiliate runs campaigns directly, with its own supply choices and its own conversion data feeding optimization. Teams without in-house hands can bring in Epom’s ad ops team for setup and optimization.
| Meta / Google Ads | Self-serve DSP | |
|---|---|---|
| Entry for gambling | Written permission (Meta) or per-country certification (Google) | Exchange and publisher policies per market; Google AdX lists approved countries |
| Who chooses placements | Platform algorithm | Operator, via allowlists and PMP deals |
| Who sets bids | Mostly automated platform bidding | Operator’s rules, applied by automated bidding |
| Conversion data | Platform pixel and server APIs, feeding the platform’s own bidding model | Operator’s own server-to-server postbacks, feeding rules the operator sets |
| Placement evidence for regulators | Varies by surface, often aggregated | Placement-level reporting |
| Where it fails | Uncertified countries, suspensions | Publishers that block gambling; banned markets |
When a DSP pays off, and when it doesn’t
A DSP pays off for teams that can build allowlists, read placement reports and adjust bids every week, or that bring in a partner to do it. Without that work the controls sit unused, and results can fall below what an automated platform delivers, even at its premium.
Search intent is also hard to replace. A person typing “betting sites” is closer to depositing than most display or video audiences, and part of the $71 click price reflects that. A DSP is best used alongside search and social. It can take budget from them where the walled garden’s premium and opacity outweigh its intent advantage.
For an operator in a certified market, the useful question is how much of the media plan has to sit inside an auction it doesn’t control. Even moving 20–30% of display and video budget to placements the team can choose, price and audit gives a benchmark for what the rest is really costing.
The premium is worth testing
Google and Meta will stay central to gambling acquisition, because their intent and reach are hard to match. What they charge for access is the part worth questioning. A DSP does not remove the gates, since exchanges and publishers keep their own, but it lets the operator choose which ones to go through and keep a record of what it bought. In a vertical where regulators now ask operators to show where their ads ran, that record can matter as much as the price.