How do I know if a channel is profitable?

Efri · 4 min read

Save

Here is how to know if a channel is actually profitable.

Step 1: Fix your tracking first

You cannot know what works if you do not know where customers came from. Put UTM tags on every link you share. Use the same naming every time.

Capture the click ID from each platform: gclid for Google, fbclid for Meta, li_fat_id for LinkedIn, msclkid for Microsoft, ttclid for TikTok. These IDs let you match a visit on your site to a specific ad click.

Connect your CRM to your ad data. When a deal is created, copy the original source onto the deal record. This ties a channel to real revenue, not just a click.

Then check your numbers against your bank account. Add up what all your platforms claim. Compare it to what you actually collected. The gap tells you how much is being double-counted.

Step 2: Define profit correctly

Profit is not revenue. It is revenue minus cost. For each channel, calculate return on ad spend and cost to acquire a customer.

Platform-reported ROAS is often inflated by 30 to 50 percent. That happens through double-counting and loose attribution. Telling your finance team that a platform says 4.8x is not proof. The platform is grading its own homework.

The only number that matters is incremental ROAS. That is the revenue that happened because of the ads, divided by what you spent. It leaves out sales that would have happened anyway.

This is why incremental numbers are always lower than platform numbers. It is also why incremental numbers can be compared across channels and platform numbers cannot.

Step 3: Know your baseline

Before you test one channel, you need to know how much revenue comes from marketing at all. Run a full holdout of all marketing in a small part of your market. Ten percent is a safe start.

Compare revenue in that group to a matched group that still sees ads. The difference is what marketing is actually driving. The rest is demand that would have happened anyway.

Baseline is the revenue you would get with no ads. It comes from your brand, repeat customers, word of mouth, and season. Incremental is the extra revenue the ads cause. Every test is about separating the two.

Step 4: Run holdout tests

The only way to know if a channel creates new revenue is to run a controlled test. A holdout test compares a group that sees ads to a group that does not. If the group with ads converts more, the ads caused that difference. That difference is the incremental lift.

Geo tests are the gold standard. They do not depend on the platform's own data. They use your real sales numbers.

Here is how. Split your markets into two groups. Run ads normally in one group. Turn them off in the other. Compare revenue between the two. The difference is the real lift from that channel.

Start simple. Turn a channel off in one region. Keep it on everywhere else. Watch what happens to revenue. Do not test in your biggest market. Pick a region that is about 15 to 20 percent of your spend for that channel.

That is enough to see a signal without risking your whole business.

Match the test length to your sales cycle. If it takes 60 days to close a deal, a 30 day test tells you nothing. You are just counting deals that were already in motion.

Lock in the rules before you start. Agree with your team on what counts as a win, a loss, or a tie. Agree on what you will do in each case. Do this before you go live.

Step 5: Measure incremental ROAS

Take the revenue you lost in the holdout region. Divide it by the spend you cut. That is your true return.

Benchmarks vary. For video, CTV has the highest median incremental ROAS at $1.38. YouTube is $1.26. Linear TV is $0.75. Use a benchmark to pick which channel to test first. Then measure your own number.

Step 6: Move your budget

Cut spend from channels below your target. Add spend to channels above your target. Stay budget neutral where you can. Move money, do not just spend more.

Some channels claim more credit than they create. Branded search, retargeting, email, and affiliate often fall in this group.

If you pause branded search tomorrow, many of those customers still find you through organic search or by typing your name. Prospecting social, non-brand search, video, and audio often create demand.

They tend to measure at or above what the platform claims.

Step 7: Test again every quarter. Results change. Saturation, competition, and audience fatigue all shift the numbers. Run the cycle every quarter. Aim for at least five tests a year. Complex brands run twenty or more.

What you need

You need a way to split your audience or your regions before ads run. You need clean tracking to measure each group. You need CRM integration to tie revenue back to the test.

Tools that help: Measured and Haus for geo holdouts. AppsFlyer or Kochava for mobile. Meridian GeoX as an open source option.

The bottom line. Platforms will never tell you which channel is profitable. Each one claims the sale. The only way to know is to turn it off and watch what happens. Build the tracking. Define profit correctly. Know your baseline. Run holdout tests. Measure incremental ROAS. Move your budget. Test again next quarter. That is how you stop guessing and start knowing.