When prices move, almost every dashboard lies. Three figures hold up, and they are simpler than the ones people usually watch.
Follow ratios, not amounts
Three ratios are enough, and they fit in one table.
The number of units sold in relation to the number of people reached.
The acquisition cost expressed as a product equivalent: how many items have to be sold to cover the spend.
And the share of sales attributable to a given piece of content, compared with a period without it.
Those three numbers survive price variation, platform changes and tool outages, which no automatic dashboard does.
Why attribution lies
It does not lie out of malice, it lies by construction.
Somebody sees a video on a phone, mentions it to somebody else, looks up the brand two days later on another device, then buys in store or by message. No tool links those four moments.
The result is well known: paid advertising takes credit for what the content triggered, and the content looks useless because it ticks no box.
Two practical answers. Compare periods rather than clicks: the same conditions, with and without a campaign. And ask the question directly, at the moment of purchase or in the conversation: how did you hear about us.
The indicator almost nobody uses
The number of incoming messages.
In this market, a large share of sales starts with a conversation. The volume of messages received, their nature and their quality say immediately whether content is working.
Three things to note each week: how many messages, how many are about a specific product, how many end in a sale.
It is free, immediate, independent of any tracking tool, and it reacts within hours where dashboards take three weeks.
The cost nobody counts
Internal time.
A campaign mobilises invisible hours: brief, exchanges, approval, sending the products, archiving, answering the messages generated.
Those hours appear nowhere, and they often add up to more than the price of the videos. A brand that counts them sometimes discovers that a less creative but better organised campaign earns more.
Count them once, honestly, on a real campaign. The figure obtained changes the way you buy.
What does not count
Three indicators to ignore.
The number of views, which measures distribution and not interest.
Likes, which often come from an audience that will never buy.
And followers gained, unless your model rests on an audience, which is not the case when you are buying files for advertising.
Those three figures share one flaw: they rise when a campaign is pleasant, not when it is effective.
What you look at, and in what order
| Indicator | What it tells you | When to read it |
|---|---|---|
| Three-second view | Whether the hook holds | From 1,000 views |
| Full view | Whether the edit holds | At 5,000 views |
| Click to the product page | Whether the promise is clear | After two days |
| Purchase | Whether the product follows | After a week |
What distorts a reading of the numbers
Comparing two videos run on different budgets, or judging a creative over a weekend. A test is read over seven full days, not over one evening.
What you do with each result
A weak hook is replaced, a weak middle is shortened, a weak ending is rewritten. You do not throw out the whole video for one segment.
The minimum table
One line per campaign, six columns.
The start and end date. The number of creatives run. The spend, also expressed as a product equivalent. The units sold over the period. The number of incoming messages. And a free column: what we learnt.
That last column is the valuable one after a year, because it survives changes of team, of tool and of currency.
