Algeria AgencyDigital marketing
Instagram ads: your grid is the landing page
Here people look at your profile before deciding. Nine posts settle it, and no budget repairs an empty grid.
Instagram and Facebook are run from the same tool, billed the same way and sold in the same sentence. They are nonetheless two different places, and the commonest mistake is running the same thing on both because the box was ticked by default.
What changes is not the technology, it is what the person does. Under a Facebook ad, they open the comments to make up their mind. On Instagram they tap the account name, and the first nine posts of your grid decide instead of your ad.
That moves the work to before the campaign. A grid with no prices, no visible product and three reposted quotations turns an advertising budget into a curiosity expense: people look, understand they cannot buy anything, and leave.
We are not repeating here what the article on Facebook advertising already established — the boost button, account ownership, the hour after a message, cost per accepted order. All of it applies unchanged. This article deals with what is specific to this platform.
A third the size, and not the same audience
The first decision comes before any budget, and it is whether your buyers are here at all. It is a question answered first by a national figure, which is not enough, and then by your last ten customers, which is decisive.
According to figures published by DataReportal for Algeria, Instagram has 12 million users, against 21.1 million for TikTok and 25.6 million for Facebook.
So this is not the same scale, and the reasoning "smaller, therefore cheaper" is wrong: a narrower pool empties faster, frequency climbs sooner, and an audience cut down further by fine targeting becomes unusable within days.
On the other hand this audience is younger, more urban and far more concentrated on things that are looked at: clothing, beauty, prepared food, furniture, interiors, gyms — anything a photograph alone can make somebody want. For those trades it is the best platform in the country.
And the reverse has to be said just as plainly: if you sell industrial equipment, services to businesses, or anything bought by people over fifty, you probably should not be here. That is the sentence nobody will bill you for.
DataReportal, Digital 2026 Algeria — October 2025 figures
The same ads manager, and that is exactly the trap
Both platforms are run from one tool, and that tool proposes by default to run everywhere. It is presented as an optimisation, and it is one from the point of view of cost per impression; it is not one from the point of view of what people see.
A square image built for a feed appears cropped in a vertical placement, and a vertical video built to be seen full-screen appears tiny in a feed. The same file cannot serve both placements without one of them being served badly.
The commonest fault is also the easiest to correct: text placed at the bottom of the image. In full-screen placements that area is occupied by the interface, which means by the account name and the button — so your price disappears behind a button.
The production rule fits in one sentence: a vertical version with the text in the upper third, a square version for the feed, and nothing important in the first hundred or the last two hundred pixels of the height.
If you can only produce one format, produce the vertical one and run it only where it is native. One placement fewer costs less than a creative truncated on all of them.
On Instagram your profile is the landing page
This is the central fact of this page and it has no equivalent elsewhere. Faced with an ad that interests them, the person does not click the link first: they tap the account name to see who you are.
They then see nine posts, without sound, as thumbnails, in three seconds. Those nine thumbnails answer three questions they will never ask out loud: do these people actually exist, is what they sell what I think it is, and is anybody still looking after this.
The practical consequence is blunt and free: before spending anything, look at your own first nine thumbnails on a phone. If one of them is a reposted quotation, a greetings card, or a recruitment notice, it is occupying the place of a product.
The correction needs no photographer: nine posts showing nine things you genuinely sell, with the price in the caption. It is less flattering than a grid harmonised by a filter, and it converts incomparably better.
And the order matters, because the top-left thumbnail is the most looked at: put there what you want to sell this month, not what you are proudest of.
The bio is the only clickable thing, and it is one line
Your whole profile contains exactly one clickable place, and it holds whatever you put in it. It is the platform’s most badly used constraint.
Three pieces of information have to fit in the text: what you sell, where you are, and how to order. "Online shop" says nothing; "women’s shoes, Bab Ezzouar, 48h delivery" says everything and reads in a second.
The single link has to lead to what the ad promised rather than to a home page. If your campaign is about a product, the link goes to that product; otherwise you paid for a visit in order to offer a navigation exercise.
The phrase "price by direct message" is a choice with a cost, and it is worth knowing: it multiplies messages and collapses the closing rate, because it attracts the people comparing and repels the people buying. The Facebook article explains why a displayed price filters.
Finally, check the number and the address from a phone that is not yours. A wrongly entered call button is an advertising spend ending in a misdial, and that lasts months before anybody notices.
Saves are worth more than likes
The platform shows you likes because they are numerous and reassuring. They describe almost nothing: approving costs one gesture and commits to nothing afterwards, and a post much liked by people who will never buy looks exactly like a post that works.
Two other figures are available free in the statistics of every post, and they are worth far more. Saves: somebody puts your product aside to come back to it, often to show it to somebody else before buying.
Shares by direct message: somebody sends your post to one particular person. It is the platform’s strongest signal, because it stakes the sharer’s reputation with somebody they know.
Use those two figures as a production compass rather than as a report. The most-saved post of the month tells you what to photograph next month, and it says it better than any briefing meeting.
One counter-intuitive consequence: a post with few likes and many saves is a success, and a post with the reverse is entertainment you funded.
Photography is the cheapest lever on this page
It is not an advertising question, and it nonetheless decides half the results. On this platform the quality of the image is the entry price, and it requires no equipment — which makes it awkward to sell, since there is nothing to bill beyond an hour of attention.
Three things are enough and they are free: daylight coming from a window, a plain background — a wall, a sheet, a table — and the product clean, framed on its own, without the clutter of the back room.
The second rule is quantitative: three or four photographs of the same item from different angles beat one perfect photograph. People want to see the back, the sole, the stitching, the inside; every missing angle becomes a message to handle.
The third is a scale: put something familiar beside the object, or show the object worn. A great many returns and refused parcels come from a misunderstood size, and a worn photograph costs less than a return.
We will quote no conversion figures tied to photography: we know of none dated and checkable for this market. But the trial is free — retake six photographs by a window and compare the saves over two weeks.
Video builds the audience, the grid builds the trust
The two formats on this platform do not do the same job, and judging them on the same figure is the mistake that gets the working one abandoned.
Short video is what the platform shows to people who do not follow you. That is where a new audience is, and it is judged on reach and on new followers — not on today’s sales.
The grid fetches nobody. It is consulted by people who have just discovered you, and it is judged on what happens next: messages, saves, profile visits that turned into something.
A video with a hundred thousand views and not one sale is not a failure: it is an audience brought in front of a grid that had no answer. The problem is two screens further on than where it is usually looked for.
Production that holds up over time is therefore lopsided: a few videos a week for reach, and a stable grid showing what is for sale. Inverting the effort — a heavily worked grid and no video — produces a handsome account nobody discovers.
Stories are the closest thing to a price list
Ephemeral posts are seen only by people who already follow you, which makes them a poor acquisition tool and the best selling tool on the platform. They address an audience made up entirely of people who have already decided to look at you, and that audience is the least solicited of all.
The format that works is dull: a photograph, a price, an available size, and one sentence saying how to order. That is not creative work, it is a daily display, and it is exactly what your customers expect from a shop.
The part almost nobody uses is free: the permanent highlights at the top of the profile. They can hold a catalogue by category, a size guide, the delivery zones and times, and photographs of orders delivered.
Those highlights then work for you at every new visit, including the ones your advertising has just paid for. One hour of tidying, once, and one question fewer in every conversation, for months.
And one precise use is worth flagging: put there the answers to the three questions you receive most often. You will know which by looking at your messages not as customer service but as a list of things your profile failed to say.
Partnerships: what to ask before paying somebody
Paying somebody with a following to show your product is a widespread practice here, sometimes effective, and almost always bought on the wrong criterion.
The follower count says nothing. It can be bought, it ages, and it includes accounts that will never see the post. Offering it as a pricing argument is the sign that the person is selling an audience they do not measure.
Ask three concrete things about their last three collaborations: the saves and shares on the post, the number of messages the client received, and whether the client came back a second time. The third answer is worth the first two.
Ask also what will remain: is the post kept, can you run it as advertising from your own account, and for how long. A collaboration erased after twenty-four hours buys the value of a passing display and is often billed at the price of a campaign.
What we refuse: buying followers for your account, and paying for a collaboration on the strength of a claimed audience alone. Both produce a presentable figure and no customers, and the first damages your account’s reach for a long time.
What can be measured here, and why an agreed word beats a dashboard
The four figures that matter are the ones in the Facebook article — messages received, the share handled within the hour, orders accepted, origin asked of the customer — and they apply here without modification.
What is specific to this platform is that attribution is weaker still. The journey happens entirely inside an application, usually on one device, and it sometimes ends in a private conversation that nothing connects to your campaign.
The countermeasure is homemade and it works: an agreed word. A campaign says "mention window when you write to us", or carries a discount code specific to that channel. You then get an exact measurement of what that channel produced, depending on no dashboard.
Two different codes for two campaigns are worth more than a twenty-page report, and one person noting the codes in a notebook will give you a figure the platform cannot.
And the platform’s most useful measurement stays free: profile visits set against what happens after them. Many visits and few messages point at the grid, never at the campaign.
What to check before signing
The first question is about placements: will both platforms be run with the same files? If the answer is yes, ask why, and listen for whether the word "cropping" appears.
The second is about the grid: what do they propose to correct before the first campaign runs? A supplier talking about budget without having looked at your first nine thumbnails has not looked at your account.
The third is about production: who photographs, where, and how often. It is the most decisive line of a quotation on this platform, and it is often the line that is missing from it.
The fourth is about the figures reported: ask explicitly for saves and shares rather than likes. The answer tells you whether your supplier reads the platform or their own report.
The fifth is the usual test: ask what they would refuse to do. On this page the answer worth having is "I would refuse to run ads before your grid shows products", and it costs whoever gives it a month of billing.
What we do, and what we will refuse to do
What we will refuse: running ads on this platform for a business whose work cannot be photographed, or whose buyers are not here. We will say so before the quotation, and we would rather lose the campaign than make it succeed in a report and fail in your bank account.
We will refuse to launch before the first nine thumbnails show nine things you sell, with prices. It is a correction that bills almost nothing and changes results more than targeting does, which explains why it is rarely proposed.
We will refuse to buy followers, to pay for a collaboration on the strength of a claimed audience, and to run one single file across every placement while calling that an optimisation.
What we do: a vertical version and a square one, with the text in the upper third; a bio that states the trade, the place and the delivery time; highlights tidied into a catalogue; saves and shares as the steering figures; and one agreed word per campaign so measurement depends on no dashboard.
And what you should do without us this week: open your profile on a phone that is not yours, look at the first nine thumbnails for three seconds, and count how many show a product with a price. Then retake six photographs beside a window. It is free, and it decides more than the budget.
Frequently asked questions
Should we run ads on Instagram and Facebook at once?
Yes, but not with the same files. A square image is cropped in vertical placements and text at the bottom goes behind the interface. A vertical version and a square one are enough; failing that, keep the vertical and reduce the placements.
Is our business right for this platform?
If it photographs well and is bought by relatively young, urban people, yes: clothing, beauty, food, furniture, sport. If you sell to businesses or to an older clientele, your budget will work better elsewhere.
Do we need a professional photographer?
Not to begin. Window light, a plain background, the product on its own, and three or four angles per item are enough to get ahead of most accounts. Add a worn photograph or a scale reference: many returns come from a misunderstood size.
How many followers do we need before advertising?
The count does not matter; the grid does. Nine posts showing nine products with their prices are worth more than ten thousand followers in front of an account where it is unclear what is for sale.
Which figures should we look at?
Saves and shares by direct message, post by post, rather than likes. Then profile visits compared with messages received: many visits and few messages point at your grid, not at your campaign.
How do we know what the campaign actually brought in?
Through an agreed word or a discount code specific to that channel, written down by hand when an order arrives. Attribution is especially weak here because everything happens inside an application, and a notebook will give you an exact figure the dashboard cannot produce.
Where we come in
Nine thumbnails seen from somebody else’s phone tell you what a buyer sees. They do not tell you whether your business belongs on this platform at all.
- We arrange the grid so nine things you sell are visible with their prices.
- We rewrite the bio: the trade, the area, and how to order.
- We say plainly when photography cannot serve your product.
If what you sell does not photograph, no budget repairs that: we will tell you where the same money works better.
Read next
Which post to fund: the signal nobody looks at
The best advertising image already exists on your profile. How to find it, what funding does to it, and what not to produce.Running a Facebook campaign: the structure, the wait, one change
Campaigns break mostly from being managed too much. How to structure one, how long to leave it alone, and what to change when you change something.Facebook ads: what they actually produce, and the hour that follows
Here a Facebook campaign produces messages, not orders. And the cost per message means nothing if nobody answers.
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.