Algeria AgencyDigital marketing
The video that answers a question still earns in two years
The platform is also a search engine. It is the one surface in this pillar where work done once keeps working.
The article accompanying this one splits the platform into two products under one name: an engine and a television. It then spends eleven of its twelve sections on the television — the skippable advertisement whose first five seconds are the product.
This page deals with the other half, the engine. People type a question, the platform offers them videos, and one of them can be yours — three months after you published it, and then two years after.
That is the difference that matters across this whole pillar. On the other platforms what you publish is distributed for a few days and then disappears, and the two articles either side of this one are precisely about surfaces where nothing persists. Here a well-placed video goes on being found without anybody pushing it.
So we will look at which questions deserve a video, where to find them, how to title them, what belongs in the first thirty seconds, at what rhythm to publish, and — the most honest section in the article — which trades this does not work for at all.
Two products, and this one is not paid for
The first obstacle is vocabulary. "Advertising on this platform" and "being found on this platform" name two activities that share only the name of the site.
Advertising is bought, adjusted, stopped, and ceases to exist the day you stop paying. That is the companion article’s subject, and it is a good subject.
The engine is not bought. You enter it by publishing something that answers a question people ask, and the video stays findable afterwards, indefinitely, with no budget.
What that changes for a small business is considerable. Advertising spend stops when cash gets tight, and everything stops with it; a video published a year ago goes on bringing people in during the month you spent nothing at all.
It is also what makes it slow. Nothing this article describes produces an effect this week, and a business that needs customers this month should read the companion article rather than this one.
The video that answers a question
The form is narrow and describes itself in a sentence: one precise question, put in the title, and the answer given straight away.
It is not a company presentation, not a tour of the premises, and not an atmosphere video. Those three have their place elsewhere; none matches what somebody types into a search bar.
The test is easy to apply before filming: would somebody type this phrase? "How to tell if a car battery is dead" gets typed. "Our workshop has welcomed you since 2015" does not.
The second half of the test is more demanding: the answer has to be complete. A video that sets out the question and closes with "contact us to find out more" is disguised advertising, and it is treated as such — watched for twelve seconds, then left.
That requires accepting something uncomfortable: you are going to give away what you know, including to people who will use it to manage on their own. The section on what it returns explains why it pays anyway.
The questions are already in your messages
Finding subjects is the part everybody overcomplicates, and the most reliable source is already yours.
Read back thirty messages you have received and note every question in them. Not the price enquiries — the questions: does it hold on tiles, how long does it take to dry, can I do it myself, what size should I take.
Do the same with what you are asked at the counter or on the phone. A question you have explained a hundred times out loud is exactly a question hundreds of people type, and your answer is already worn smooth.
The sign of a good subject is irritating and reliable: it is the question that wearies you. If you sigh inwardly on hearing it, that is because it comes round often, and "comes round often" is the only qualification that counts here.
A keyword tool is not needed to start and is misleading at the beginning. It gives national or regional volumes, often for phrasings nobody in your trade uses, while your inbox holds real questions worded the way people ask them.
The title is the question, written the way it is asked
The title is not a space for expression. It is the field the search reads, and the rule is to fill it with the question itself.
Write it as somebody would type it, not as it would be drafted in a brochure. People type an incomplete, direct phrase; an elegant, allusive title, or one built around the company name, matches nothing that is being searched for.
Avoid wordplay, the rhetorical question and the title that promises without saying. Those work on platforms where distribution is pushed at people; here the person comes looking, and they either recognise their own question or move on.
A point about language that matters here more than elsewhere: people type in French, in Arabic, and often in both within one phrasing. There is no single solution — publishing two videos, one per language, works better than one that mixes, because a title can only be in one language at a time.
Finally, do not change a title that works in order to improve it. A video that has been found for a year is found because its wording matches what people type, and that match is not visible from your screen.
The first thirty seconds give the answer
This is where this article contradicts its neighbour most sharply, and the contradiction is real rather than apparent.
In an advertisement, the first five seconds exist to hold somebody who asked for nothing and is about to skip. That is a problem of hooking, and the companion article treats it in full.
Here the person asked the question, clicked, and wants the answer. Holding them is not the problem; making them wait is. An introduction, an animated logo, a company presentation before the answer send away exactly the people you managed to bring in.
The structure that works is therefore inverted from habit: the short answer first, in one or two sentences, then the explanation, then the edge cases, and the business at the end if it belongs there at all.
That structure has a side effect worth knowing: people sometimes leave after thirty seconds, satisfied. That is a good result, even though the watch-time indicator says otherwise — they got what they came for, from you.
What dates, and what does not
A video that stays findable for years ends up being found in a world that has changed, and an answer that has become false costs more than it earned.
What does not date: a method, a gesture, a principle, a diagnosis. How to recognise a fault, what the steps are, what to check before buying. Those answers stay true as long as the trade exists.
What dates fast: prices, lead times, product references, administrative procedures, anything depending on a supplier or a regulation. Keep those out of the video itself and put them in the text underneath, where they are corrected in a minute.
This section treats something other than the corresponding section of the article on filming, and the difference is worth stating. There it is raw footage staying reusable in a future edit. Here it is an already-published video whose answer stops being true — a different asset and a different failure.
The review is done once a year, in an hour, on the most-watched videos only. An answer that has become false is corrected by republishing or by taking down, and taking down is better than leaving an explanation online with your name attached to it.
One question per video, never a compilation
The temptation is to group: "ten tips for…", "everything you need to know about…". It is the form that looks most generous and it is the least effective here.
The reason is mechanical. A compilation matches no typed question exactly; it matches ten approximately, and the approximation loses against a video that answers one of them precisely.
The second reason is on the viewer’s side. Somebody looking for one precise thing does not want to work through eight minutes to find the minute that concerns them, and a timestamp in the description only half compensates.
The third is practical: ten short videos are produced separately, over ten sessions, whereas a compilation demands a whole afternoon and is therefore never made.
There is one exception and it comes later: once a dozen videos exist, an overview video pointing to them makes sense. It is then a consequence of the work rather than a way of avoiding it.
The text underneath, and what nobody puts in it
The description is treated as an administrative form and is in fact part of the mechanism, for three distinct reasons.
The first is that it is text, and text is read by machines that do not watch the video. One or two sentences restating the question and its short answer are worth more than a slogan.
The second is that it is where the perishable items the previous section kept out of the video belong: an indicative price, a reference, an address. They are corrected without touching the picture.
The third is the most neglected: it is the only place where somebody who has just been given a useful answer for free can find how to reach you. A clickable number and an address, at the top of the description, not at the bottom.
Add timestamps as soon as the video runs past two minutes. They cost five minutes to write and they turn a moderately watched video into one where people reach the part that concerns them.
What it returns, and when
The calendar has to be stated clearly, because this is where most attempts stop — not through failure, but through discouragement at the wrong moment.
The first weeks produce almost nothing, and that is not a signal. A video takes months to find its audience, because the platform offers it sparingly at first and then more if the people who see it watch it through.
What comes afterwards is cumulative, and it is the property distinguishing this approach from everything else in the pillar: January’s video goes on working while you publish June’s, and the twelfth arrives into a catalogue already receiving visits.
We quote no figure to support that, and it is better to say why than to stay vague. The question is what a video returns in two years; answering it needs a long series on Algerian channels that nobody publishes, and a foreign figure would describe another market in another period.
What you can measure yourself, on the other hand, answers exactly that question: each of your videos’ views in its second year against its first. That reading costs ten minutes a year and it settles, for your trade, what no source will settle.
The channel, and why a single video disappoints
A video published alone on an empty channel performs worse than an identical video on a channel holding ten, and that has to be known before concluding the method does not work.
Part of the effect is human. Somebody who finds a useful answer goes to see what else is there; an empty channel interrupts that movement, and the companion article already notes that an empty channel gets noticed.
Another part is how the platform works: it readily offers other videos from the same source to somebody who has just watched one. With no other videos, that offer has nothing to make.
The practical consequence is that the approach has to be judged on a series and not on a piece. Publishing three videos and concluding is as uninformative as judging a shop on its first three days.
It is also what makes the rhythm matter more than the quality of each piece. Twelve adequate videos spread over a year are worth more than three very good ones, and that sentence is roughly the reverse of what one would say about an advertisement.
When this approach does not work
It does not suit everybody, and the list of cases where it fails is short and recognisable.
The first is a product nobody tries to understand. An impulse purchase, an everyday food item, a fashion article: people do not type questions about them, so there is no search to capture.
The second is a trade whose questions are entirely local. "Is this restaurant open on Friday" is a real question and it is not asked here — it is asked of the business listing, which the article on local search treats.
The third is a business that cannot hold a rhythm of one video a month for a year. That is not a judgement: it is a constraint on time, and a method presupposing twelve months of regularity does not suit somebody without it.
In those three cases the money and the time sit better elsewhere — on the business listing, on reviews, or on the paid advertising the companion article treats. Recognising it early costs a conversation; recognising it after a year costs a year.
What we do, and what we refuse
What we do is bounded. We read back thirty of your messages to extract the questions, we keep twelve, we write the titles and the short answers, and we come back once a year to revise what has stopped being true. The filming is yours, by the method in the article on filming.
We refuse to promise a date on which this starts to return. We do not have one: the question needs a long series on Algerian channels that nobody publishes, and putting a delay on it would mean converting a foreign figure into a market that is not its own.
We also refuse to produce a video about a question nobody asks. It is the commonest request — a company presentation, a tour of the premises — and those are legitimate objects that simply do not belong to this approach: nobody searches for them, so the engine will offer them to nobody.
And there is one thing to do tonight without us, which decides all the rest: read back thirty messages and note the questions in them. If you find twelve that recur, the method has an object; if you find only three, your trade is among those the previous section sets aside, and that is a free answer in half an hour.
Frequently asked questions
Do we have to pay for the video to be found?
No, and that is the difference between the two halves of this platform. Advertising is bought and ceases to exist when you stop paying; a video answering a question stays findable afterwards with no budget. In exchange it is slow: nothing here produces an effect this week, and a business needing customers this month should read the companion article.
How many videos, and at what rhythm?
One a month for a year, rather than three very good ones. A single video on an empty channel performs worse than an identical one on a channel holding ten — people go to see what else is there, and the platform readily offers other videos from the same source. Judge the method on a series, never on a piece.
In French or in Arabic?
People type in both, often mixing them within one phrasing, and a title can only be in one language at a time. Two videos, one per language, work better than one that mixes. Start with the language most of your messages arrive in — that is information you already have.
Is this not giving away our expertise for free?
It is, and that is the condition. A video that sets out the question and closes with "contact us" is disguised advertising, watched for twelve seconds and then left. Some people will manage on their own; those who cannot or will not now know you, and a clickable number at the top of the description is what turns that into a call.
What do we do with a video whose answer is no longer true?
Correct or take down, once a year, on the most-watched videos only. Taking down is better than leaving a false explanation online with your name attached. Better still is avoiding it at the shoot: prices, lead times and references go in the text underneath, where they are corrected in a minute, and never in the picture.
Which trades does this not work for?
Three cases. Products nobody tries to understand — impulse purchases, everyday food, fashion. Trades whose questions are purely local, which belong to the business listing. And businesses that cannot hold twelve months of regularity. Recognising the case early costs a conversation; recognising it after a year costs a year.
Where we come in
Thirty messages reread give you the real questions, not the price requests. The next one is harder: writing the title the way somebody would type it.
- We pull the questions out with you and keep the ones that can be filmed.
- We write the title in the customer’s words, not in the trade’s vocabulary.
- We revisit the video at a year if the question has changed shape.
If nobody asks that question, do not film it: a video about a subject you find important finds no audience.
Read next
YouTube ads: the first five seconds are the product
This is the only platform where you are heard, and the only one where you buy time. The skip is not a failure, it is a deadline.What advertising budget to set — and how to know
There is no published rate card for Algeria. Here is what the available data actually lets you conclude.A live campaign: the moves of the first week
The budget is set and the money is going out. What to look at, what to change, and above all what not to touch yet.
Let us talk about your project
A free audit, no commitment: we look at your online presence and tell you what is holding it back.