A new GEO can burn through your budget before you learn anything useful or get killed too early before it has a chance to convert. The difference often comes down to how you structure the test and read the data.scale profitable markets without wasting ad spend.
Testing a new GEO can go wrong in two opposite ways. You may spend a small amount (say, $30), get no conversions, and thus kill a market that simply did not have enough traffic to prove anything. Another scenario is allocating sizable budgets to a campaign that is already performing poorly.
A useful GEO test is the one that shows which step to take next. Rather than increasing ROI right away, it aims at collecting enough information for the subsequent decision, like whether to stop, give the campaign more time, change something and retest, or start raising the budget.
This read shares insights about what to prepare before you start feeding money into your campaign, how to carry out tests, and how to measure the outcomes.
Ready to put your next GEO to the test?
Register as an advertiser at HilltopAds and start testing with quality traffic.
Research the GEO Before You Spend
Before buying traffic, check whether the market gives you enough reasons to test it at all. A full market study is not necessary, and a few basic checks can rule out GEOs with weak demand or obvious barriers. Here’s how to conduct your research:
Gather the available market data
Look at the size of the audience, interest in the vertical, and any data you already have from similar countries. The aim is to see whether there is enough potential to justify a paid test.
See what competitors are running
Use spy tools to find similar offers in the region. Look at the creatives, formats, and landing pages being used. Campaigns that stay active for some time are usually more useful signals than an ad that appeared once.
Look for local demand
Sources like Google Trends and Keyword Planner can show whether people in the GEO are actually searching for the product or topic. Also check whether interest is stable or strongly seasonal.
Study the local market
Look at similar products already sold in the GEO, the main competing brands, local prices, seasonality, and buying habits. For media buyers testing dating offers, it makes sense to check the age of the target audience and what people are after when using such platforms. In more conservative markets, religious or cultural norms may also impact the product positioning.
Check the practical restrictions
A GEO can look promising and still be difficult to run in practice. Verify that the offer is available there, then check language, currency, payment methods, and local advertising rules. Take iGaming as an example. Qatar can be excluded right at the research stage, since gambling is prohibited there. The US market is more fragmented, as regulations vary by state – New Jersey is the one that approved internet gaming, while Utah law bans it.
The point of this research is to remove markets that already show clear problems. Doing so allows you to shortlist the GEOs that are worth spending test budget on.
Helena
Senior Sales Manager HilltopAds
One example of GEOs that we didn’t initially expect to perform well but that ended up being successful was Pakistan (PK) and Brazil (BR). The success came down to a strong combination of factors: the FIFA World Cup, a pre-lander tailored to the event, and a simple subscription conversion flow.
Also read our article on how to scale a campaign without losing ROI:
Run the GEO Test Without Overspending
Once a GEO has passed the basic research stage, the next issue is money. There is no universal $50 or $500 GEO test budget, since a number without campaign economics behind it can be misleading.
Set the Budget Before Traffic Starts
Suppose a proven campaign normally gets conversions at a $5 CPA. A $50 test budget gives room for roughly ten conversions at that benchmark. The actual result may be better or worse, but there is enough space for several conversion events to occur.
Now take an offer with a usual CPA of $50. The same $50 budget buys enough room for only one expected conversion. Zero conversions would hardly prove that the GEO is bad. One conversion would not prove that it works either.
Pros often use the numbers from the campaigns they already know. If a $5 lead comes in regularly, you may see a pattern quite fast. A sale that happens only after a few hundred clicks is a different story. That test needs more traffic and more room in the budget.
The numbers to set before launch:
- The amount you are prepared to spend per day;
- The maximum amount you are prepared to lose on the whole GEO test.
The second number matters when results are poor. Without it, “one more day” can easily turn into several days of spending because nobody decided where the experiment should end.
Define What Counts as a Good Result
The decision rules should also exist before the campaign starts. Take the performance of the working setup as the reference point. You already know roughly what CPA, ROI, and conversion rate make the offer viable.
It’s not necessary for the GEO to conform to those numbers right from the start. Prices and conversion rates may vary country to country, but you will need to set yourself some guidelines. For instance:
- If the agreed test budget is gone and the funnel is still clearly weak, stop.
- Results close to target may deserve more traffic, especially when the conversion sample is still small.
- Scaling comes later, once acceptable CPA and ROI repeat across several conversions.
This also protects the test from emotional decisions. Two quick conversions may tempt a buyer to spend more, while a short run of bad clicks can have the opposite effect. Neither is much of a basis for changing the plan.
Start With What Already Works
Rather than adding several new things to diagnose at once, keep the offer and traffic source familiar. Use the same funnel if you can. For the new GEO, change only what has to be localized.
If you also replace the offer, funnel, source, and ad concept, a poor result will tell you very little about the GEO itself. Imagine launching a new offer in Brazil with a source you have never used before, a new page, and a new creative angle. In such a case, CPA may come back at twice the target.
The reason for the failure could be the GEO or the offer itself. It could also be the weakness of the web page or the need to optimize the traffic source. Since everything changed at once, there is no definite answer. Increasing the budget while changing the target market and creative angle may improve results, but you will not know which change made the difference.
Helena
Senior Sales Manager HilltopAds
We had the opposite experience with GEOs that looked promising but showed poor results in HilltopAds tests with Melbet in the betting vertical. Kyrgyzstan (KG), Tajikistan (TJ), and Kazakhstan (KZ) didn’t deliver the results we expected. We experimented with different rates, tested various pre-landers, and tried different pop traffic models, including classic CPM and SmartCPM, but still didn’t generate deposits. The most likely reasons were limited traffic volumes in these GEOs at the time and strong competition from other active campaigns.
Collect Enough Data to See a Pattern
Days are usually the wrong unit to measure the optimal duration of the test. Three days may produce hundreds of conversions for one test campaign and two conversions for another.
Clicks are not a universal answer either. A funnel converting at 10% tells you much more after 200 clicks than one converting at 0.3%. The amount of data you need depends on:
- traffic volume and price;
- expected CPA;
- conversion rate;
- frequency of the target action;
- funnel length;
- delay between the click and final conversion.
The practical question is whether you are seeing the same result often enough to call it a pattern. This is also where the maximum test budget becomes useful.
If the test reaches that limit and still cannot produce a viable pattern, there should be a good reason to keep spending.
- Register
- with hilltopads
- and get
- 273B+
- of ad impressions!
Analyze the Results and Decide What to Do With the GEO
The final CPA is only one part of the test. To understand what actually happened in the new GEO, you need to look at the test campaign stage by stage.
Check the Funnel From the Start
Begin with the traffic itself. Check CPM or CPC to see how expensive and available the traffic was in the GEO. Then look at CTR, which shows whether users reacted to the creative in the first place.
After the click, check what happened on the landing page. Did users stay, scroll, register, fill in a form, or move to the next step? From there, look at the intermediate conversions, final conversions, and, finally, CPA and ROI.
Reading the results in this order makes campaign optimization easier and helps you see where performance starts to break down. A low CTR, for example, may mean that the creative does not fit the local audience.
If CTR is healthy but users do little after the click, the problem is more likely to be further down the funnel. The landing page, the offer itself, or traffic quality may need another look.
Compare the GEO With a Working Benchmark
Use a successful GEO with the same or a similar campaign as your reference:
- Compare traffic cost first.
- Then CTR.
- Then conversion rate.
- Then CPA and ROI.
The useful part is seeing where the new GEO first starts to fall behind. Two GEOs can both end up with a $30 CPA for very different reasons. In one, clicks may cost more with the landing page converting well. In the other, traffic can be cheap, and most users may leave after the click.
Those cases need different fixes. The first may improve with cheaper placements, while the second calls for a closer look at the landing page, offer, or traffic quality.
Read our previous article about the difference between contextual and behavioral targeting:
Stop, Continue or Scale
After reviewing the funnel, put the GEO into one of three groups.
| Decision | What the Data Looks Like |
| Stop | Results repeatedly miss the acceptable range and the campaign has consumed enough budget to make the problem clear |
| Continue | The test still lacks enough conversion data, or performance is close enough to target to justify another controlled sample |
| Scale | CPA/ROI remain acceptable across a meaningful volume of traffic and conversions |
These are campaign decisions rather than permanent labels for countries. A GEO that fails today can become interesting later. Traffic prices change, and so do offers. New payment methods appear, as well as competitors enter and leave markets. Seasonality can also create a completely different environment.
It is also worth keeping a short record of why the test was stopped. “CTR was normal, but checkout conversion stayed 60% below the reference GEO after X spend” gives a much clearer picture than marking the GEO as unsuccessful.
Helena
Senior Sales Manager HilltopAds
When it comes to testing Tier 1, Tier 2, and Tier 3 GEOs in practice, there are some differences advertisers often underestimate. For Tier 1, I wouldn’t recommend testing complex flows such as iGaming, cams, or AI with a budget below $500. Ideally, you should allocate at least $1,000 per GEO. Lighter verticals such as apps, sweepstakes, eСommerce, and subscriptions can be effectively tested in Tier 2 and Tier 3 markets.
The key principle is simple: the more GEOs you test and the more complex the conversion flow, the larger the budget you need to get meaningful results. With pop traffic, it’s also important to use a pre-lander, ideally with several variations for testing. For other ad formats, strong, attention-grabbing creatives are essential.
Read the successful cases of our advertisers on launching campaigns for different GEO:
Scale a Winning GEO Gradually
Good results during a test can change once the budget goes up. Higher spend may bring more expensive traffic or users who convert worse than the first group.
It’s crucial to increase spend in steps instead. After each increase, give the campaign enough time to settle and check the same numbers again:
- Conversion volume;
- Conversion rate;
- CPA;
- ROI;
- Traffic cost and quality.
If performance stays inside the acceptable range, make the next move. If CPA starts climbing faster than the extra conversion volume justifies, stop increasing the budget and investigate. The best result may be a profitable GEO at $300 per day rather than an unprofitable one forced to $1,000.