Collect a full quarter of spend by channel
Use at least three months, so one odd week does not decide anything. For each channel, add up everything you paid to run it: ad spend, tools, freelancers and the time of the people who work on it.
Time is the cost people leave out. If you spend ten hours a week on LinkedIn posts, put a value on those hours. A channel that looks free usually is not.
Count new customers by source
Open your CRM and count customers who started in the same period, grouped by source. Use first touch, or the source your sales team recorded on the first call. Pick one rule and keep it.
If your source field is empty on half the deals, fix that first. No calculation survives that. HubSpot and Pipedrive both let you make the source a required field when a deal is created.
Calculate CAC for each channel
The formula is simple. Total cost of the channel divided by the number of new customers it produced.
Take an example with invented numbers. Search ads cost 6,000 and produced four customers, so CAC is 1,500. Content cost 3,000 in people time and tools and produced three customers, so CAC is 1,000. Outbound cost 4,000 and produced one customer, so CAC is 4,000.
Put the results in a Google Sheets table with one row per channel. Looker Studio can chart it later if you want.
Compare each CAC with what a customer is worth
CAC means little alone. Compare it with first-year gross profit from a customer won in that channel. A CAC of 4,000 is fine when the average customer brings 20,000 in gross profit, and a disaster when it brings 3,000.
Add one more column: months to earn the CAC back. If a channel takes longer to pay back than your cash allows, it is too expensive for you, even if it looks good on paper. The calculation of how much you can afford per lead sits in Calculate how many leads you need and what you can afford.
Adjust for lead quality and timing
Two corrections stop you making silly decisions.
- Quality: if one channel produces customers who churn in three months, its real CAC is higher than the table says.
- Timing: content and SEO take months to pay back, so judging them on one quarter is unfair. Compare them over six months, and paid channels over the same six months.
Where you have fewer than about ten customers in a channel, treat the number as a hint and not as a verdict.
Check marginal CAC before adding money
The average CAC of a channel tells you what it did. The marginal CAC tells you what the next pound will do. Compare the last increase in spend with the extra customers it brought. If spend rose by a third and customers rose by a tenth, the channel is saturating.
This is why moving everything to your best channel backfires. See Dominate one channel before adding the next for how to scale in stages.
Move 10 to 20 per cent, then wait
Take 10 to 20 per cent of the budget from your worst channel and give it to your best. Leave it for a full cycle, which for most B2B teams is one to two quarters. Then repeat the calculation.
Keep a minimum in every channel that works at all, so you can restart it. Put the move, date and reason in your decision log, so the next review knows what you did and why. To try something new with a small slice of the freed money, see Test a new channel each quarter.
Common mistakes
- Using ad spend only and ignoring people's time.
- Trusting last-click data, which gives all the credit to the final touch and underrates content and referrals.
- Cutting a slow channel after one quarter.
- Moving half the budget at once.
- Recalculating only when someone complains about spend.
How you know it works
Blended CAC falls over two quarters, or stays flat while volume grows. Your channel table is updated every quarter in under an hour. Every budget move is logged with a number behind it, and you can explain each one in a sentence.
