Why did my cost per purchase increase?
Enter two periods of online sales numbers and see which part of the chain moved your cost per purchase.
When cost per purchase rises, the cause is one of three links: reaching people costs more, fewer people click, or fewer clicks become purchases. A tired ad shows up as a falling click rate, a busy auction as a rising CPM, and a weaker page or broken tracking as a falling conversion rate. Enter two periods of the same length, use purchases as the result, and the tool splits the change and drafts a client message. Compare it with your break-even cost per result to see whether it still pays.
Made with Cofoundr ยท free Meta Ads tools
Runs in your browser. Nothing you enter is sent to us. Your last numbers are remembered on this device only.
Cofoundr shows the change in cost per result against the previous period on every client page, and alerts you when it climbs past your limit.
How it works
- Cost per result is spend divided by results. It is the product of three things: what it costs to reach 1,000 people (CPM), how many of them click, and how many clicks become results.
- Each of the three is compared between the two periods and shown as a percentage change.
- The share of the change is worked out so the three shares add up to 100 percent. A driver pushing the other way shows a negative share.
How to use the result
The biggest driver tells you where to look, not what is wrong. CPM points to the auction, audience and placements. Click rate points to the creative and offer. Conversion rate points to the landing page, tracking and traffic quality. Compare two periods of the same length, and avoid a period with a big sale or outage in only one of them.
Common questions
Should I use purchases or add to carts as the result?
Use the result that matters to the client, usually purchases. Use the same one in both periods.
Why compare two periods?
A cost on its own says little. The change between two comparable periods shows which part of the chain moved, so you can fix that part.
Can results mean leads, purchases or messages?
Yes. Use whatever the account counts as a result. Use the same result type in both periods.
What if results are higher than clicks?
That can happen with on-Facebook lead forms and messaging. The split still works, but read conversion rate as results per click.
Related
Find the budget your clients are wasting
Connect Meta. Cofoundr scores every client account and flags wasted spend. 50 free credits for 7 days, MCP included, no card.
Free Meta ads auditSee features
Where the time goes for an agency running 4 Meta ad accounts
Scenario: 15 minutes of checking and 30 minutes of diagnosing per account per day over 22 working days, plus 2 reports per account at 3.5 hours each. Your numbers may differ.