Facebook ads: what they actually produce, and the hour that follows

Facebook advertising is the first marketing spend for most Algerian businesses, and often the only one. It is also the worst measured, not for want of tools but because everybody looks at the wrong figure.

What a campaign produces here, in the overwhelming majority of cases, is neither an order nor a visit to a site: it is a message. And a message is a person waiting, with a question, for a length of time you decide without thinking about it.

Everything else — the targeting, the creative, the budget — comes after that sentence. An excellent campaign pouring forty messages into an inbox nobody opens before the next morning is still an excellent campaign, and an entirely wasted spend.

This article will give you no cost per result: no dated, checkable series exists for this market. It gives you the real mechanism, the two figures that matter, and what we refuse to launch even when the budget is ready.

The button under the post is the default path, and it is the most expensive per result obtained. It was designed to be easy rather than to be effective, and that difference is written nowhere.

It optimises for what looks like success: reactions, comments, shares. Those three can be had cheaply from people who will never buy, and they produce a satisfying screenshot rather than a sale.

A campaign built in the ads manager lets you state what you actually want: a message, a call, an order. The platform then goes looking for the people who do that thing, and they are a different population from the ones who react.

The gap shows within a week. A boosted post gets three hundred reactions and four messages; the same amount with a message objective often gets forty, with far fewer reactions and a less flattering report.

The button is not useless for all that: it serves to show information to people who already follow you — an opening, a change of hours, an item out of stock. The mistake is using it to acquire customers.

The ads manager gives you a cost per message or per registered order. It cannot tell you what happens at the customer’s door, and that is precisely where the leak is.

If some of the parcels come back refused, your real cost is the budget divided by the accepted orders, not by the registered ones. The gap between those two numbers is large enough to overturn the conclusion of a campaign you believed was profitable.

The direction of the market, meanwhile, is dated: over one year the number of online merchants grew 26%, card transactions on the internet 38%, and the amount paid online 179%, while the card base grew only 9%.

The useful reading is in the gap between the last figure and the other three: the cards already existed, what is changing is usage. Paying before dispatch is no longer an exotic request, and an order paid in advance is an order that cannot be refused at a door.

What we advise and what we refuse: offer prepayment as an option with a small advantage rather than as an obligation, and measure the cost per accepted order for a month before touching the targeting. We refuse to optimise a campaign whose acceptance rate nobody has counted.

Your ad is seen with no sound, in a feed, by a moving thumb, on a phone held at arm’s length. The first second decides everything after it, and vertical format is not a fashion but the shape of the screen.

The text in the image has to fit in a thumbnail and in the language your buyer thinks in. The shortest words win, and one line in French followed by one line in Arabic works better than a sentence attempting both at once.

Then the price. A hidden price produces the largest number of messages and the lowest closing rate: it is a way of buying conversations you are going to lose. Showing a price filters, and filtering is exactly what you want to pay for.

There is finally an omission specific to this market: an ad showing a product without saying where it ships from or how long it takes to arrive. Those two lines clear more objections than any slogan, and they cost nothing to add.

The test is free: look at your own ad on your phone, sound off, at normal distance. If you cannot tell in two seconds what it is, what it costs and how to get it, no budget will repair that.

What we will refuse: launching a message campaign for a business that cannot answer within the hour during the hours the ads run. We say so before the quotation, while it still costs us a project, because after that it only costs the client.

We will refuse to sell the boost button as an acquisition method, and to build narrow interest targeting on a small budget in a market this size. Those are the two easiest things to sell on this page.

We will refuse to optimise a campaign whose accepted-order rate nobody has counted, and to promise a cost per result before a month of measurement. A figure announced before the measurement is an estimate dressed as a commitment.

What we do: the objective chosen from what you sell; the price and the delivery time visible in the creative; comments handled in the same hours as messages; the page and the account in your name; and the cost per accepted order as the figure reported.

And what you should do without us this week: for seven days, count three things — messages received, how many waited more than an hour, and how many orders were accepted. A notebook is enough, and those three numbers will settle the campaign better than any targeting.