Estimating a monthly attribution bill from your own numbers

Gather three numbers first

Every pricing model uses one of three counts, so collect all three once and reuse them for each vendor you look at.

• Installs by source. Take last month's installs from the app stores and from your ad platforms. Split them into installs from paid ads, installs from your own links and everything else.
• Events per month. List each in-app event you plan to send to the tool and how many times it fired last month. Your product analytics or your backend has this.
• Monthly active users. The number of distinct people who opened the app in the month. See MAU for how the count is defined.

If the app has not launched, use your plan for the first three months and mark the estimate as a guess.

Match your numbers to the vendor's unit

Read the pricing page for the billing unit, then pick the row that applies.

If the page shows no rate, you cannot finish the formula. Stop there and use the questions to ask when a vendor shows no price.

A worked example with published figures

The example below uses Airbridge Core because its figures are public: $40+/mo, 500,000 data points included, then $0.0001 per data point, where one data point is one event Airbridge collects for your app.^[1][2] The app is hypothetical, and the result is an estimate from published figures.

Say a fitness app records these events in a month:

• 20,000 installs
• 6,000 trial starts
• 2,500 purchases
• 700,000 other standard events it has chosen to track

1. Add the events: 20,000 + 6,000 + 2,500 + 700,000 = 728,500 data points.
2. Subtract the included amount: 728,500 − 500,000 = 228,500 extra.
3. Multiply by the rate: 228,500 × $0.0001 = $22.85.
4. Add the base: $40 + $22.85 = $62.85 for the month.

Now remove half of the "other" events because no report uses them. The total becomes 378,500, which is under 500,000, and the estimate returns to the $40 base. The event list moved the bill, and the number of users did not change.

Turn the estimate into a range

A single figure hides the risk. Run the formula three times.

1. Today. Last month's real numbers.
2. Planned. The month you expect six months from now, using your growth plan.
3. Peak. Your busiest month, such as a launch or a seasonal campaign.

If the peak figure is far above the planned one, read how the vendor handles volume above the plan in plan limits and overage before you choose.

What to do next

Add the items the formula leaves out, which are listed in costs that sit outside the plan price. Then start a trial and compare the usage the tool reports after two weeks with half of your monthly estimate. A large gap means your event count or the vendor's counting rule differs from what you assumed.