Algeria AgencyDigital marketing
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.
On this platform you buy neither an impression nor a click: you buy attention by the second, and you only pay if somebody stays. It is the most honest model in the pillar and the worst exploited.
From that follows one constraint that decides everything else: the person can leave after five seconds, and most will. Those five seconds are not a hook meant to prevent the departure; they are the five seconds within which your message has to have been delivered in full.
There is also a particularity this series has not had until now: you are heard. In a social feed the sound is off by default and everybody knows it. Here the person was watching a video with sound, and they still have it.
This article contains no chart, and that is deliberate: we know of no dated, checkable series for Algeria on viewing, watch time or cost per view. Section 9 explains that refusal in the place where a figure is precisely what is expected.
You are buying time, not impressions
The dominant format is the one that can be skipped after five seconds, and it is billed in a way nobody takes seriously: you only pay if the person stays past that threshold, or if they click.
That means a video skipped by a thousand people costs you nothing. It is mechanically true, and it inverts the whole logic of production: your risk is not being disliked, your risk is being watched by people who will never buy.
The consequence is counter-intuitive and worth writing plainly: a long video that keeps a third of its audience can cost less per result than a short one everybody skips, because you only pay for the people who stayed.
Another consequence, more useful still: put the price, the place and the exact nature of the offer in the first seconds. You thereby filter out the people you do not want to pay for, and the skip becomes your ally rather than your opponent.
Finally it changes how you judge. A high skip rate is not bad in itself: it is free filtering. What is bad is a high watch rate with no measurable effect, because there you paid for attention and got nothing.
The skip is not a failure, it is a deadline
The commonest fault is treating the five seconds as a hook meant to retain. In go a rhetorical question, an animated logo, a mood line — and the person leaves knowing nothing about you.
The right way to use them is the opposite: assume the video stops at the sixth second and write it that way. Who is speaking, what is offered, for whom, at roughly what price. It is not elegant, it is effective, and it costs no more to film.
What comes after the five seconds no longer addresses the same audience: it addresses the people who stayed, so people who are interested. That is where demonstration, detail, proof and the address belong.
An editing rule follows and it is easy to check: cut your video at the fifth second and watch that fragment on its own. If it is not enough to understand what you sell, the video needs redoing whatever the quality of the rest.
And one detail systematically forgotten: the logo at the end serves nobody. The vast majority of your audience will never see the end, and the place for your name is the beginning.
This is the only platform where you are heard
All the advice in this series about muted sound, compulsory subtitles and burned-in text applies to social feeds. Here it does not apply in the same way, and that should be said clearly.
The person was watching a video they chose, usually with sound, sometimes with earphones. They are in a listening disposition you will not find on any other platform in this pillar, and that disposition is a resource.
It means the voice becomes a working tool again. A sentence said clearly in three seconds lands better than text to be read, and it lands in your customer’s language — dialect included, which is an advantage silent video cannot offer.
That does not excuse you from subtitles, for two practical reasons: some of the audience watches at reduced quality or in a public place, and a burned-in subtitle survives every condition. Burn them into the image rather than relying on a setting.
The combination that works is therefore double: the voice carries the message, the subtitle preserves it. It is more work than a silent video, and this is the only place in the pillar where that extra work is paid back.
Two products under one name: a search engine and a television
This platform is used in two ways with almost nothing in common, and the same budget buys two different kinds of attention there without the interface telling you.
The first is a search engine. Somebody types "how to change a gasket", "review of this model", "tutorial", and watches a video to learn. That is intent, in the sense of the article on search advertising, and it is the most profitable context.
The second is a television. Somebody is listening to music, following a match, letting videos run on. The attention is real but it is not available: your advertisement interrupts a song, and intent is not on your side.
Both have their use. The search context suits a service somebody wants to understand before buying; the entertainment context suits a reputation being built, or a product discovered by being seen.
The practical decision is not to mix them in one campaign, because the average of the two describes neither. Two separate campaigns, with two different videos, cost the same money and produce information you can act on.
What video does well, and what it will not do
This format excels at three things, and it matters to name them because they are rarely what it gets asked for.
The first is demonstration: a product in use, a movement performed, a machine running. Anything understood by being seen, that would take a paragraph to write, is cheaper and more convincing here.
The second is the place: showing a workshop, a room, a kitchen, a warehouse. In a market where the customer’s first question is often "do these people actually exist", thirty seconds of a real place is worth more than a page of text.
The third is the person: somebody speaking, with their face and their accent. It is the fastest way to make a small business credible, and it needs nothing but a phone held straight.
What this platform will not do: produce an immediate order. Nobody interrupts a video to fill in a form, and a supplier promising you direct sales at volume from video is selling a mechanism that does not exist here. The effect arrives later, elsewhere, and section 9 explains how to see it.
The channel is your profile, and an empty channel is noticed
As on the other platforms, an interested person goes to check who you are, and where they check is your channel. It is the same mechanism as the Instagram grid, with one difference: expectations here are lower and emptiness is more visible.
A channel containing nothing but the advertising video is a clear signal: this business paid an agency for a campaign and does not otherwise exist. It is not fatal, but it cancels part of the trust the video had just built.
The remedy is modest and needs no production: four or five short videos, filmed on a phone, showing what you actually do. A job in progress, a delivery, a preparation, a before and after. Polish matters less than existing.
Arrange them into playlists named with your customers’ words rather than your own. "Prices and timings", "how we install", "our work" are understood; an internal product-range name is not.
And fill in the channel description like a bio: what you do, where, and how to reach you. It is the most consulted and least completed place on the whole platform.
Production: a phone is enough, a corporate film is not
The budget mistake specific to this format is the institutional film: three minutes, music, drone shots, a slogan, and no information. It costs a lot, it goes down well in a meeting, and nobody watches it past the fifth second.
What works is filmed with what you have: a phone held straight or on its side, daylight, a microphone brought close — even an earphone microphone beats the built-in one from two metres away.
Produce two versions of the same content, and that is the only technical constraint that matters: a vertical one for short placements, a horizontal one for ordinary viewing. The same shoot gives you both if you frame wide.
Useful length is not a rule but a consequence: as long as you have things to show, and not one second more. A ninety-second demonstration that keeps its audience is better than a twenty-second edit that said nothing.
And one habit that saves a great deal: film several short sequences in a row, once a month. Almost all the cost of a session is in setting it up, and leaving with six sequences instead of one divides that cost by six.
Subtitles, language, and what survives a bad connection
The language of your video is a commercial decision rather than a reflex. Spoken dialect lands better aloud than written French, and it carries a familiarity nothing else replaces in local advertising.
But speech alone is fragile. Some of your audience watches at reduced quality to save data, some in a place where they cannot turn the sound up, and some do not understand the variety you are using.
The countermeasure is single and it settles all three cases: subtitles burned into the image, large, at the top or in the middle rather than at the bottom, because the bottom of the screen is taken by the interface and the buttons.
Write those subtitles in whichever language your customer reads fastest, which is not necessarily the language spoken in the video. A voice in dialect with a subtitle in French is a very common combination here, and it works.
Finally, test your video in real conditions: on a phone, at low quality, without earphones, in a street. It is free, it takes five minutes, and it reveals unreadable text nobody had seen on a computer screen.
What can be measured, and why this article quotes no figure
We give no cost per view, no average watch time and no completion rate for Algeria here. That is not an omission: we know of no dated, public, checkable series on those three quantities for this market.
The figures in circulation come from three families and none is usable. Global averages, which aggregate unrelated markets. Platform data repeated without a date. And agency studies, which describe their own campaigns and get published because those went well.
What you can measure yourself is more modest and exact. The number of billed views, meaning people who stayed past the threshold. The share of those who watched to the halfway point. And, the only genuinely useful measure, the change in how many people search for your name.
That last indicator is the right one for this format, and it is free: if your video is working, more people type your company’s name in the following weeks. You see it in your site’s statistics and in the searches that lead to you.
And the measurement that cuts across everything: ask at the counter and on the telephone where the person saw you. On this platform automatic attribution is the weakest in the pillar, because the effect arrives days later and by another route.
Targeting: what works here and what wastes
Three ways of targeting exist on this format, and they do not have the same value in a market this size.
The first is the subject of the videos being watched, and it is the most useful: you choose the context rather than the person. A do-it-yourself video, motoring content, a make-up tutorial — the context states the interest better than a demographic box.
The second is re-engagement: showing the video to people who visited your site or saw an earlier video. It is the best spend on this format when it is available, because those people already know who you are and the five-second threshold is crossed more often.
The third is a list of specific channels, and that is where the waste happens. Choosing ten channels cuts the audience to a size where frequency explodes within days, exactly like the narrow targeting described in the article on Facebook.
The rule that follows: broad on context, narrow on re-engagement, and never narrow on both at once. And check where your advertisement actually ran — the placement report exists, it is rarely opened, and it contains surprises nobody chose.
What to check before signing
The first question is about the five seconds: show me the beginning of the video, on its own. A supplier who starts with an animated logo has not understood the format they are billing you for.
The second is about separating the contexts: will the campaign distinguish search from entertainment? If not, you will get an average that describes neither and on which no decision is possible.
The third is about production: how many sequences for the price asked, and are they delivered in two formats. One horizontal video for everything is an incomplete deliverable on this platform.
The fourth is about measurement: which figure will act as judge. If the answer is "views", ask how a paid view will be distinguished from a commercial effect, and listen for whether the phrase "brand searches" appears.
The fifth is the usual test: what they would refuse to do. The right answer here is "I would refuse to produce a three-minute film", and it costs whoever gives it the most profitable line on their quotation.
What we do, and what we will refuse to do
What we will refuse: producing a multi-minute institutional film intended to be run as advertising. It is the commonest and most profitable commission in this format, and almost nobody watches it past the fifth second.
We will refuse to deliver a video whose first five seconds contain neither the trade, nor the offer, nor an order of price. We will also refuse to mix the search context and the entertainment context in one campaign, because the average describes nothing.
We will refuse to promise direct sales from video, and to quote a cost per view "typical in Algeria". No dated series establishes one, and a figure invented in a quotation is a promise the report will later have to contradict.
What we do: the whole message in the first five seconds; the voice in the customer’s language and subtitles burned in at the top; two formats from one shoot; search and entertainment separated; and the change in searches for your name as the judge.
And what you should do without us this week: film three twenty-second sequences on your phone — a product in use, your premises, and you explaining a price. Then watch them on somebody else’s phone, without earphones, at low quality. You will know immediately whether there is a campaign to run.
Frequently asked questions
Should the video be short or long?
As long as you have things to show, not one second more. You only pay for people who stayed past five seconds, so a long video keeping a third of its audience can cost less per result than a short one everybody skips.
What goes in the first five seconds?
Who you are, what you offer, for whom, and an order of price. Write the video as though it stops at the sixth second: most people will leave there, and they should leave having understood.
Are subtitles needed if the sound is on?
Yes. Some of the audience watches at reduced quality, in a public place, or does not understand the spoken variety. Burn them into the image, at the top or in the middle: the bottom of the screen is taken by the interface.
Can we film on a phone?
Yes, and it is often better than a corporate film. Daylight, a microphone brought close, and wide framing so one shoot yields a vertical and a horizontal version. Polish matters less than the information given.
What does a view cost in Algeria?
We quote no figure: no dated, checkable series exists for this market. What circulates comes from global averages, from platform data with no date, or from agency studies describing their own good campaigns.
How do we know whether the campaign achieved anything?
Look at whether more people search for your company’s name in the following weeks — that is free and visible in your statistics. Then ask at the counter where people saw you: automatic attribution is the weakest in the pillar here.
Where we come in
Three twenty-second clips tell you whether a message holds in five. That is this platform’s only question, and a phone answers it.
- We watch your attempts and say which one lands inside five seconds.
- We speak your buyer’s language and subtitle in the other one.
- We choose between targeting content and targeting search, with the reason written down.
We promise no direct sale from a video: this channel makes you known, and attributing an order to a view would be invention.
Read next
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.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.