iGaming PPC Campaign Structure: Budgets, Bidding & Ad Groups That Convert
Most iGaming PPC accounts we audit are not losing money because of bad ad copy. They are losing money because of bad structure. One campaign covering three countries, ad groups with 80 keywords in them, brand and generic terms sharing a budget, and Smart Bidding pointed at registrations instead of deposits. The account looks busy, the spend goes out every day, and the finance team cannot tell you what a first deposit actually costs.
When a click on a casino keyword in the UK or Canada can run well past $30 and a sportsbook term during a major fixture goes even higher, structure is not a housekeeping issue. It is the difference between a profitable acquisition channel and an expensive one. This is the structure we use across our iGaming paid marketing accounts, and the reasoning behind each decision.
Start with the account, not the campaign
Before you build a single campaign, decide how the account is going to be divided. In iGaming, the split is almost never by product first. It is by geography and licence first, because that is where the compliance rules live.
Every regulated market has its own ad copy requirements. UK ads need the 18+ marker and a safer gambling message. Ontario does not allow bonus offers in ads at all. Germany has its own restrictions on what you can say about stakes and speed of play. If a UK campaign and an Ontario campaign share an ad group, one of them is non-compliant by definition, and Google’s policy system will eventually catch it.
So the account hierarchy we build looks like this:
- Separate accounts (or at minimum separate campaigns) per licence and geo.
- Within each geo, campaigns split by intent: brand, competitor, generic and offer terms.
- Within each campaign, ad groups split by a single keyword theme.
- Language separated from geo. A Spanish speaker in the US and a Spanish speaker in Colombia are two different campaigns, not one with two languages ticked.
We covered how the language and cultural side of this plays out in our post on localising campaigns for LatAm and European markets. The short version is that the moment you let geos share campaigns to save time, you lose the ability to control budgets, copy and bids where it matters.
The four campaign types every iGaming account needs
Brand
This is the cheapest traffic you will ever buy and the easiest to lose. Affiliates bid on operator brand terms constantly, and if you are not running brand, someone else’s landing page is sitting at position one for your name. Brand campaigns get their own budget, their own bidding strategy, and an aggressive impression share target. They never share a budget with generic terms, because the moment they do, the algorithm spends the generic money on brand clicks that were going to convert anyway and reports a beautiful CPA that means nothing.
Competitor
Bidding on competitor brand names works in iGaming, but it is a scalpel, not a hammer. Keep it in its own campaign, use exact match only, write copy that talks about your own offer rather than theirs, and expect CPCs that look expensive next to brand. The value is in stealing high-intent users who are already in a deposit mindset. Watch the conversion rate closely and cut the competitors that only drive registrations.
Generic
Terms like “online casino,” “best betting site,” “live roulette” plus the geo modifier. This is where most of the budget goes and where most of the money is made or wasted. It gets the tightest ad group structure, the most negatives, and the longest bidding roadmap (more on that below).
Offer and bonus terms
“Welcome bonus,” “free spins no deposit,” “bet 10 get 30.” High volume, high intent, and the highest share of bonus abusers. Keep these separate so you can measure quality per campaign. In our experience the registration-to-deposit rate on offer terms is often half of what generic terms produce, and if they are mixed together you will never see it.
Ad groups: one theme, one intent, one landing page
The ad groups that convert in iGaming are small. Five to fifteen keywords that all mean roughly the same thing, one clear ad theme, and one landing page built for that theme. “Live dealer blackjack” and “online blackjack” are two ad groups. “Sports betting app” and “sports betting site” are two ad groups, because the landing page and the copy should differ.
For the top ten percent of terms by spend, we go further and run single keyword ad groups. It sounds like overkill until you see the quality score difference on a term that costs $40 a click.
On match types, exact and phrase carry the account. Broad match only enters the picture once a campaign has a mature conversion history feeding Smart Bidding, and even then it goes in its own ad group with a tight negative list, never mixed with exact match terms.
The negative keyword work is where structure is enforced. Three layers:
- A shared account-wide list for terms you never want: “free play,” “no download,” “jobs,” “hack,” “cheat,” “apk,” plus any geo you are not licensed in.
- Cross-campaign negatives so brand terms cannot trigger generic ads and generic terms cannot trigger brand ads.
- Ad group level negatives so “live blackjack” queries do not leak into the “online blackjack” ad group.
Run the search terms report weekly for the first month and fortnightly after that. In iGaming the query landscape shifts every time a new bonus format or a new competitor launches.
Budget allocation: how we split it
There is no universal split, but there is a starting point that has held up across dozens of operator accounts. For a mature account in a regulated market:
- Brand: 10 to 15 percent. Enough to hold impression share above 90 percent. If brand is eating more than this, affiliates are bidding hard on you and it is worth a conversation with them.
- Generic: 50 to 60 percent. The engine of the account.
- Offer terms: 15 to 20 percent, capped and watched for deposit quality.
- Competitor: 10 to 15 percent, only in markets where it is permitted and only after generic is stable.
- Testing reserve: 5 to 10 percent for new geos, new ad groups or new formats.
The number that matters more than the percentages is the daily budget relative to your target cost per first deposit. Google’s own guidance for Target CPA is a daily budget of at least ten times the target, and in our experience that is the floor, not a nice-to-have. A campaign with a $150 CPA target and a $400 daily budget is starved. It will get four or five clicks a day, never accumulate enough conversion data, and Smart Bidding will make decisions on noise. Better to run fewer campaigns properly funded than twelve campaigns on fumes.
Sportsbook accounts get one more layer: budget pacing around the fixture calendar. Weekends, major football days and big events are where the conversions sit. We front-load daily budgets and raise bids in those windows and pull back on Tuesdays and Wednesdays, rather than setting one flat number for the month. Ad scheduling and bid adjustments by day and hour do most of the work here.
Bidding: the three phase roadmap
The single most common mistake we see in iGaming PPC is turning on Target CPA or Target ROAS on day one, pointed at registrations. Smart Bidding will do exactly what it is told. It will find the cheapest registrations on the internet, and the cheapest registrations in iGaming come from bonus hunters and people who never deposit.
Phase one: collect real data
Manual CPC or Maximise Clicks with a bid cap for the first four to six weeks. The goal is not efficiency. The goal is to buy enough traffic to understand the actual cost per first deposit by campaign, and to make sure conversion tracking is clean before an algorithm depends on it.
First time deposit is the primary conversion. Registration is a secondary conversion that you observe but never optimise to. If your affiliate platform or CRM holds the deposit data, set up offline conversion import so Google sees deposits tied back to the click, not just form submissions.
Phase two: Maximise Conversions, then Target CPA
Once a campaign has 30 to 50 first deposits in a rolling 30 days, switch to Maximise Conversions with FTD as the conversion action. Let it run two to three weeks, then move to Target CPA set slightly above the observed cost per FTD, tightening by 10 to 15 percent every couple of weeks as the data supports it. Never tighten a target by half in one move. The learning period resets and spend collapses.
Phase three: Target ROAS, only with real value
Target ROAS is the right end state, but only if you are passing meaningful value into the conversion. That means first deposit amount at minimum, and ideally an early net gaming revenue signal or a predicted LTV score from your CRM. If the value you pass is a flat placeholder, tROAS is just tCPA with extra steps and a more confusing dashboard.
Use portfolio bid strategies with bid caps on generic campaigns. Uncapped Smart Bidding on a term like “online casino UK” can and will push individual CPCs past $60 during competitive periods, and the reporting will not flag it until the month is over.
The structure only works if the landing page does
Structure decides who clicks. The landing page decides who deposits. An account can be perfectly built and still lose money if every ad group sends traffic to a generic homepage. Each ad group theme should map to a page that repeats the promise in the ad, loads in under three seconds on mobile, and puts registration one tap away. We laid out exactly how we build these in our landing page and CRO framework for iGaming offers, and the two pieces are meant to be read together.
Compliance is part of the structure, not a checklist afterward
Every campaign in a regulated geo needs the right certification attached, ad copy that meets that market’s rules, and a landing page that matches. Running gambling ads on Google in regulated markets requires a certified account, and a policy strike on one campaign can put the whole account at risk. This is why we run separate, whitelisted ad accounts per market where the risk justifies it, and why we keep crypto and fiat casino brands on separate structures entirely. The crypto casino compliance rules are different enough that mixing them creates avoidable exposure.
Once the core Google Ads structure is stable, the same logic extends to Microsoft Ads and to the wider network mix. We wrote about how we keep control when scaling across 40+ ad networks, and the principle is the same: separate by geo and intent, fund properly, optimise to deposits.
A worked example: UK online casino, $60k monthly budget
To make this concrete, here is how a mid-sized UK casino brand’s Google Ads account would look on our structure.
Brand campaign, $7,000. Exact and phrase match on brand terms and misspellings. Target Impression Share at 95 percent, top of page. Own negatives to keep out “complaints” and “withdrawal problems” style queries, which get their own small ad group with a support landing page.
Generic campaign, $33,000. Twelve ad groups: online casino, live casino, live roulette, live blackjack, online slots, new online casino, casino app, casino sites UK, pay by phone casino, fast withdrawal casino, and two single keyword ad groups for the highest cost terms. Exact and phrase. Target CPA on FTD after the first six weeks, portfolio strategy with a $45 max CPC cap.
Offer campaign, $10,000. Four ad groups around welcome bonus, free spins, no deposit bonus and deposit match. Separate landing page per ad group. Target CPA set 20 percent higher than generic to reflect the lower deposit quality, reviewed monthly.
Competitor campaign, $6,000. Exact match only on eight competitor brand names, copy focused on the client’s own withdrawal speed and game library. Manual CPC for the first quarter because volumes are too low for Smart Bidding to learn.
Testing, $4,000. New ad group themes, Performance Max trial with brand exclusions, and a Microsoft Ads mirror of the generic campaign.
Every campaign has UK ad copy with the 18+ marker and safer gambling messaging, sitelinks pointing to responsible gambling and terms pages, and offline conversion import from the platform tagging deposits back to the click.
The mistakes that cost the most
Mixing geos in one campaign so ad copy compliance breaks. Optimising Smart Bidding to registrations rather than deposits. Brand and generic sharing a budget so the CPA looks fake. Ad groups with dozens of loosely related keywords sending everyone to the homepage. Daily budgets set so low that no campaign ever exits learning. Switching bidding strategies every week because last week’s number looked bad. Running an entire regulated market on one certified account with no backup.
None of these are exotic. They are the defaults that happen when an account is built quickly and never restructured. The good news is that structure is fixable, and the results usually show within the first billing cycle.
If you want us to look at how your account is currently built and where the budget is leaking, request a free audit and we will walk through it with you.
FAQ
How many keywords should an iGaming ad group have?
Five to fifteen closely related keywords is the range we work in. For the highest cost terms, single keyword ad groups give better quality scores and tighter control over copy and landing pages.
Should I optimise iGaming PPC campaigns to registrations or deposits?
Deposits, every time. Registration is a useful secondary signal, but Smart Bidding optimised to registrations will fill your funnel with bonus hunters. Import first deposit data as the primary conversion and build the bidding roadmap around it.
What daily budget do I need for Target CPA to work?
At least ten times your target cost per first deposit, and enough volume to reach 30 to 50 deposits a month per campaign. Below that, the algorithm is learning from noise.
Can I run brand and generic keywords in the same campaign?
You can, but you should not. Brand converts cheaply and will absorb shared budget, hiding the true cost of generic traffic. Separate campaigns with separate budgets and separate bidding strategies.
Is Target ROAS worth using for casino or sportsbook campaigns?
Only once you are passing real value into the conversion, such as first deposit amount or a predicted LTV score. Without that, Target CPA on deposits is more reliable and easier to manage.
