How do I compare advertising costs across channels?

The problem with the platform metrics is that they measure different things. Cost per click is priced by competition for attention. Cost per thousand impressions is priced by reach. A channel with cheap clicks and poor intent can cost more per customer than an expensive one where the clicks come from people ready to buy. Comparing the headline figures directly leads you to the wrong channel confidently.

Cost per enquiry is the first honest comparison. Total spend on a channel divided by the enquiries it produced. That single figure puts a search campaign, a social campaign, a directory listing, and a sponsored newsletter on the same scale, and it usually reorders them.

Cost per customer is the one that matters. Cost per enquiry divided by the share of enquiries that convert, which differs by channel more than people expect. Referral enquiries convert at a far higher rate than cold traffic, which is why a channel producing fewer, better enquiries frequently wins on this measure while losing on every metric the platform reports.

Getting the enquiry count requires attribution, and attribution for a small business is mostly a question rather than a technology. Ask every enquiry how they found you and write the answer down. Six months of that record beats any analytics configuration, because it captures the conversations, recommendations, and offline moments that no tracking reaches.

Include everything the channel actually costs. Advertising spend, management fees, the time you spend on it valued at something, and any tooling. A channel that appears cheap on spend alone can be the most expensive once the hours are counted, and time is the scarcest input in a first year business.

Compare against what a customer is worth rather than against each other alone. A cost per customer of two hundred is excellent if the average customer spends two thousand and poor if they spend two hundred and fifty. Without that comparison the ranking tells you which channel is cheapest and not whether any of them are viable.

Allow enough time before judging. Most channels need a quarter to produce a readable number, and decisions made on two weeks of data are noise. The exception is a channel producing nothing at all, which is usually apparent quickly.

Watch for channels that assist rather than convert. Somebody may find you through social, search your name later, and enquire directly. The last click gets the credit and the first one earned it. This is where a simple asked question outperforms tracking, because people will tell you they had seen you around.

Then act on the comparison rather than collecting it. The point of ranking channels is to move money from the worst to the best, and a business that measures carefully and changes nothing has bought a report rather than an improvement.

Record the comparison somewhere it persists, since the useful pattern only appears across several quarters. A single quarter tells you which channel performed, and four consecutive quarters tell you which one is reliable.

Be careful about killing a channel on one bad period, particularly one with a long consideration cycle. The rule worth applying is a full quarter of data and a specific reason, rather than a bad month and a hunch.

Compare paid channels against the unpaid ones honestly, because referral and repeat business usually win on cost per customer by a wide margin and rarely appear in the comparison at all.