How do I find the best traffic sources for AI offers?
- 📅 Updated October 23, 2026
- ⏱️ 8 min read
- 📚 Answer 2 of 2
By Masha Gebber, Help center editor
Short answer
Give every source and placement its own Source ID and sub-IDs, then compare them on earnings per click over a complete period with enough conversions. Split by country and device before deciding, scale the clear winners and fix or drop the rest.
- You can only find your best sources if each one has its own labels.
- Compare on earnings per click over complete periods, not on clicks.
- Country and device splits often reveal winners hidden inside average sources.
- Scale winners gradually and keep testing one new idea at a time.
Once sales start coming in, the most valuable question you can ask is: where exactly do they come from? Almost every affiliate has one or two sources that earn far more than the rest, and many have a source that quietly loses money. Finding them is mostly a matter of labeling well and reading your reports the right way.
You can't compare what you can't separate. Before anything else, make sure each placement has its own link:
| Level | Field | Example |
|---|---|---|
| Channel | Source ID | reddit, blog, discord, exoclick |
| Property or campaign | sub1 | r-aiart, review-site, t1-mobile |
| Placement | sub2 | pinned, verdict, below-player |
| Creative or angle | sub3 | banner-03, memory-hook |
If several placements share one link today, create separate links now and replace them. From today on, your reports will tell you which one earns. See where to find your tracking link.
Clicks show attention. Earnings per click shows value. A source with modest clicks and high EPC is often your best source.
- Open the flexible breakdown report.
- Choose the last complete 30 days.
- Group by Source ID, then sub1, then sub2.
- Show clicks, conversions, earnings and EPC.
- Sort by EPC.
See how to read EPC and conversion rate for what the numbers mean.
Rank only rows with enough conversions to trust. As a rule of thumb, a placement needs at least several conversions in the period before you compare it with others. Rows with a handful of clicks or one lucky sale belong in a "too early" list, not in your ranking.
Tip: if most of your placements are "too early", extend the period to 60 or 90 days, or wait. Patience beats guessing.
Averages hide winners. A source that looks average can contain an excellent segment:
| Split | What it reveals |
|---|---|
| Country | Tier 1 visitors from a source may earn several times more than the rest |
| Device | Mobile and desktop can convert very differently, especially on paid traffic |
| App | Your audience may strongly prefer Sugarlab AI or SpiceStars AI |
| Landing page | One page may suit a source much better than another |
Example: a tube campaign earns $0.20 per click overall. Split by country, Tier 1 desktop visitors earn $0.55 and everyone else $0.12. The affiliate moves budget to Tier 1 desktop and the campaign's overall EPC doubles.
Renewals change how you should judge sources over time.
If you run RevShare, your best sources keep earning after the first month. A placement that looks average in its first 30 days can become your best once renewals arrive.
- Compare RevShare placements over at least three months.
- Look at earnings from older clicks, not just new sales.
- Communities and SEO often look better over time than in their first month.
EPC is the main number, but a source's real value includes a few more things:
| Signal | Why it matters |
|---|---|
| Approval rate | High rejections mean unpaid sales and account risk |
| Stability | A source that earns steadily beats one that spikes and vanishes |
| Effort | Hours per week needed to keep it running |
| Control | Sources you own, like a site or community, are safer than rented ones |
| Growth potential | Can it realistically double, or is it already at its limit? |
Tip: score each source from one to five on these signals once a quarter. The best source on paper isn't always the best business.
Put every placement with enough data into one of three groups:
| Group | Signs | Action |
|---|---|---|
| Scale | Top EPC, stable over two periods, low rejections | Give it more traffic, effort or budget |
| Fix | Decent traffic, below-average EPC | Test a better landing page, creative or app |
| Drop | Low EPC after a fair test, or high rejections | Pause it and move the effort elsewhere |
Write the decision down with the date, so you can check the effect next month. Decisions you never measure are just guesses with extra steps.
- Organic sources: create more of the same content, such as more reviews in the cluster that earns or more events in the community that converts.
- Placements: copy the winning format to similar pages or channels.
- Paid sources: raise budget gradually, add similar placements and keep winning zones in their own campaign.
See how to scale without losing quality.
A blogger finds that comparison articles earn three times more per click than single reviews, and that the verdict box beats sidebar banners everywhere. Next month's plan: four new comparisons and verdict boxes on every review.
A community owner finds that weekly events convert far better than the pinned resources post, and that Tier 1 members prefer SpiceStars AI while others prefer Sugarlab AI. Next month's plan: two events a week, with the app matched to each event's theme.
A media buyer finds that two zones on one network produce most of the profit, while mobile traffic in Tier 3 countries loses money. Next month's plan: a whitelist campaign for the winning zones and Tier 3 mobile excluded.
Keep one row per source and update it monthly:
| Source | Clicks | Conversions | EPC | Approval rate | Hours per week | Decision |
|---|---|---|---|---|---|---|
| Review site | 4,200 | 61 | $0.74 | High | 6 | Scale |
| Subreddit | 1,100 | 22 | $0.88 | High | 4 | Scale |
| Pop campaign | 9,500 | 30 | $0.14 | Low | 3 | Fix or drop |
Your best sources point to your next ones:
- Look for neighbors. If a subreddit converts, related subreddits or a Discord server for the same niche may too.
- Follow the questions. What your best audience asks in comments can become new content or new placements.
- Match the format. If honest reviews earn most, write comparisons and alternatives in the same niche.
- Read the guides. The best traffic sources guide compares eleven channels.
Test one new source at a time, with its own labels, and give it a fair period before judging.
At the start of each month, spend thirty minutes on this routine:
- Update your scorecard with last month's clicks, conversions, EPC and approval rate.
- Compare with the previous month. Which sources grew, which shrank and why?
- Pick one source to scale, using the method that fits its type.
- Pick one source to fix or drop, and write down the test you will run.
- Pick one new source to test, only if your current sources are stable.
- Set a date to check results, usually four weeks later.
Three decisions a month, each measured, is enough to keep improving without drowning in data.
Traffic and conversion patterns shift through the year. Holidays, Valentine's Day, summer and the end of the year can all change how audiences behave. Compare each month with the same month last year when you have the history, and don't drop a strong source because of one seasonal dip.
- Using one link for several placements.
- Ranking sources by clicks or traffic volume.
- Deciding after a week or a handful of sales.
- Ignoring country and device splits.
- Dropping RevShare sources too early.
- Scaling three things at once, so you never know what worked.
How often should I review my sources? Weekly for a quick look, monthly for decisions about budget and effort.
What if all my sources have similar EPC? Then look at volume and effort. A source that needs one hour a week is better than one that needs ten, at the same EPC.
Should I drop a source with low EPC but no cost? Not automatically. If it takes little effort and brings some sales, it may be worth keeping while you focus energy elsewhere.
How many sources should I run at once? As many as you can manage well. For most affiliates that is two or three before hiring help.
Can I see which app each source prefers? Yes. Give each app its own links on the same placement and compare EPC after a fair period.
Contact support if a source's results look impossible, such as conversions without clicks or clicks from countries you never targeted, or if you want help reading your reports. Include the source, period and a screenshot.
Early on, a single source is your whole business, and you learn it by feel. Once you run several sources, intuition stops working: the source you enjoy most isn't always the one that pays most. A labeled, regularly reviewed set of sources turns your time into a portfolio you can manage, and makes the path to VIP and PRO faster, because more of your effort goes into the placements that bring approved sales.