Algeria AgencyIndustries
The first ninety days: what order to do it in
Everybody knows what to do. Almost nobody knows in what order — and the order is what decides the result.
The article beside this one gathers what holds true for all twenty trades in the series. It is a distillation, and a distillation has no order: it gives principles, all true at once.
What almost everybody starting out is missing is not the list. It is the sequence. Doing the right thing at the wrong moment costs the same time as doing it at the right one, and produces nothing.
This article proposes twelve weeks, in an order where each step rests on the one below it. There is nothing sophisticated in it: the merit of a sequence is that it prevents you starting with whatever is enjoyable.
One thing depends on your trade, and it is the first section: what you sell decides what has to be verifiable first. The rest of the order is the same for everybody.
Why the order matters more than the list
Every start-up checklist looks alike, and they are broadly right. A page, a listing, a number, photographs, written answers, a measurement: nobody seriously disputes that those things are needed.
What separates a business that did this work from one that failed at it is almost never the list. It is the order it went in, and the natural order is the wrong one because it follows enjoyment.
The enjoyable order starts with the logo and the page, continues with the photographs, and arrives after six months at the question of who answers the phone. The useful order does the opposite, and it looks disappointing for three weeks.
The construction rule is a staircase: a step goes down only if the one below it exists. A handsome page pointing at a number nobody answers is not half a success, it is expenditure that increases the number of disappointed people.
It is also what makes the sequence testable. At each stage the question is the same — does what I have just done stand without what I have not done yet? — and when the answer is no, the step is premature however interesting it is.
Where your trade sits, and why it changes the first step
One thing in this sequence depends on what you do, and it is what a customer can verify before buying. A shop is verified on its stock and location, a service on its references and availability, a producer on capacity and compliance.
The composition of the economy says who this article is addressed to. The commercial register held 2,419,913 operators at 10 December 2025, of which 42.11% in retail, 37.6% in services and 14.45% in production.
The marked figure here is services, and it is marked for a precise reason: it is the broadest category and the least visible. A shop has a window, a producer has a product; a service business has nothing the customer can look at before paying, which shifts the whole sequence.
The practical consequence is short. If you sell a service, the most urgent step is not the page, it is proof: something verifiable, today, by somebody who does not know you. If you sell goods, it is availability and the address.
This figure says the composition of the register and nothing else. It does not say how many businesses succeed, nor in which category it is easier to start, and it must not be read that way — it serves only to locate the reader who found no page under their own trade.
CNRC, position at 10 December 2025
Week 1: the number that rings, and who answers it
The first step is invisible, which is why it gets skipped. It consists of deciding which number a person dials to reach you, and who answers when they do.
A number is chosen once and does not change. It will appear on a sign, on invoices, on packaging, in the memory of people who will call you back in two years — and each of those places costs something to correct.
The choice that counts is this: does the number belong to the business or to a person? A personal number works perfectly for three years, then somebody leaves or changes their life, and the customers follow the number rather than the business.
Then you decide who answers, in what hours, and what happens outside them. "Somebody always ends up picking up" is not a decision, it is an average — and the calls it loses are the ones arriving at the worst moment, which is while you are working.
This week’s test is brutal and takes thirty seconds: ask somebody who does not know you to call your business at three in the afternoon and at seven in the evening, and listen to what happens.
Week 2: one verifiable sentence about what you sell
The second step is a sentence. It says what you do, for whom, and where — and it has to be verifiable by somebody with no reason to believe you.
"We accompany our clients towards excellence" is not that sentence. "We repair industrial sewing machines, in Algiers and its outskirts, on site within forty-eight hours" is one, and it has the property of being false if it is false.
That constraint is useful because it forces you to choose. A verifiable sentence excludes what you do not do, and that is precisely what most businesses starting out refuse to do, for fear of losing an enquiry.
It will be reused everywhere: on the listing, on the page, in the first message of a conversation, in the mouth of whoever answers the phone. A business that describes itself in three different ways depending on where you look is not yet a business to somebody discovering it.
Write it in the two languages your customers use, and have somebody outside read it. The question to ask them is not whether it is handsome; it is what they think you sell after reading it once.
Weeks 3 and 4: one place, and which one
The third step is choosing one place where you are found, and one only. That is not thrift, it is a condition of truthfulness: two places maintained by a busy person become, within six weeks, one up-to-date place and one wrong place.
The choice is made by observation rather than preference. If people write to you on a messaging app, that is it; if they call after seeing a business listing, the listing is it. The right place is where the questions come from, not the one you find most professional.
For the great majority of businesses starting in Algeria, the business listing does more work than anything else. It carries the address, the hours, the number and the reviews, and it appears without being looked for.
What that place needs to contain is five things: the sentence from week 2, the address, the real opening hours, the number from week 1, and a way to write to you. Nothing else is necessary at this stage and everything else can wait three months.
The real hours deserve particular insistence because they are the most frequently wrong and the most expensive information there is: somebody who travels to a closed door does not come back, and will never tell you they came.
Weeks 5 and 6: the five questions, written once
By this stage people are writing to you, and you discover they all ask the same questions. There are five, they are specific to your trade, and you will know them after two weeks of conversations.
The step consists of writing them once, properly, and reusing them. Roughly what does it cost, do you deliver, where exactly are you, how long does it take, and what payment do you accept: it is almost always those, in a varying order.
An answer written once is better than one improvised fifteen times, for a reason that has nothing to do with time saved: it is consistent. Two customers comparing what you told them must find the same thing.
The answer about price is the hardest and it is the one not to dodge. If you cannot give a price, say what makes it vary and give an order of magnitude — "from" is information; "it depends" is not, and it reads as a refusal.
Put those five answers where whoever answers can find them, not in your phone. It is the first time in this sequence that something becomes transferable, and that is what makes it important.
Weeks 7 and 8: the proof you are allowed to show
Proof is what every new business lacks, and it is also the area where lying is easiest. The rule is simple and holds the whole section: show only what happened.
What happened is more plentiful than people think. A finished job photographed, premises, a workshop, a vehicle, a machine, a person doing the work: none of that is an invented testimonial and all of it is verifiable.
What is forbidden is not forbidden by honesty alone; it is also a question of risk: an invented figure, a review you wrote yourself, a client reference quoted without agreement. The third is the most frequent and the most dangerous, because it commits somebody else.
Agreement is asked for, obtained in writing, and covers something specific: the name, the photograph, the nature of the work. A verbal agreement given in a good moment does not survive a later disagreement, and that is exactly when it matters.
If you have nothing, say so and replace proof with a commitment. "We have no references in this sector yet, here is what we commit to and what happens if we fail" is a defensible position; a borrowed reference never is.
Weeks 9 and 10: the account belongs to the business
This step produces nothing visible and is the most profitable in the sequence at three years. It consists of checking that everything you have created belongs to the business rather than to a person.
The list is short: the number, the business listing, the domain name if you have one, the platform accounts, the email address that receives enquiries. Each has an owner, and that owner is almost always whoever created the account one evening.
The mechanism is always the same and does not depend on anybody’s good faith. Somebody leaves, falls out, falls ill or changes number, and the business finds itself in front of its own listing unable to edit it.
The correction costs an hour at the start and becomes very hard later. An email address in the business’s name, access shared with at least two people, and a written note of who holds what: that is all this step asks.
Write that note and keep it somewhere other than a phone. It is the document you will look for on the day you need it, and that day will be precisely the one on which the person who knew is unreachable.
Weeks 11 and 12: the first measurement
The last step is to count something, and to count it by hand. A three-month-old business does not need a dashboard; it needs to know where its customers come from.
The question is put to every person who buys, once, without pressing: how did you hear of us. The answer goes down in three words in a notebook, and after a month you have information nobody can sell you.
That count beats any measurement tool at this stage, for a simple reason: it covers the channels that leave no trace. Word of mouth, the sign, the neighbour, somebody who saw you working — none of those appears in a dashboard and they are often the main ones.
The second number to keep is enquiries left unanswered. It is unpleasant, and it is the most actionable of them all, because it measures neither the market nor the competition: it measures something entirely within your control.
Three months of those two numbers is enough to decide where the first advertising dinar goes. Without them, that dinar goes wherever whoever calls you is selling, which is the commonest way to start and the least effective.
What not to do in these ninety days
Do not start with advertising. Paying to send people to a number that does not answer and a sentence that says nothing is the only expense in this list that actively makes things worse, because it produces disappointed people instead of producing nothing.
Do not create five accounts on five platforms. One living account beats five of which four show a last post from seven months ago — and that silence is read as a signal about the business, not about the account.
Do not promise a lead time you have never met. The first customer who believes you and is let down costs twice: the sale, and the account they will give of it, and at three months old that account travels fast because there is no other.
Do not buy software. At three months a notebook and a spreadsheet cover everything this sequence asks, and the subscription money is better spent on the step you have not taken yet.
And do not skip week 9. It is the only step whose absence is never visible until it is too late, and the next section explains why.
What collapses when a step is skipped
Each skipped step produces a characteristic defect, and it is worth knowing how to recognise them because they diagnose from the outside, on any business.
Skipping week 1 gives a business people find and cannot reach. The symptom is a respectable view count and zero enquiries, and it is systematically blamed on the advertising or the market when it is a telephone.
Skipping week 2 gives a business nobody knows exactly what it sells. The symptom is a majority of off-topic enquiries, and it is corrected with a sentence, not with more visibility.
Skipping weeks 5 and 6 gives a business that answers well while the owner is answering and badly the moment they are busy. The symptom appears exactly when trade starts to work, which makes it particularly cruel.
Skipping week 9 produces no symptom for years, then produces one, final. The person holding the accounts is no longer there. It is the only step in this list whose omission cannot be made up, and the only one that produces nothing visible on the day you take it.
What we do, and what we refuse to do
What we do in these ninety days is modest and comes to three things: we write the week 2 sentence with you in both languages, we set the business listing up properly once, and we do the week 9 ownership note — the step nobody asks for.
We do not sell advertising to a three-week-old business. It is the expense most often refused in this trade and the easiest to justify: sending people to a chain that is not ready produces disappointed people, and a disappointed person costs more than a stranger.
We do not build a website at this stage either, unless your trade requires one — a catalogue, a booking, an online sale. For most businesses starting here the listing does more work for nothing, and saying so costs us a piece of work.
We promise no result at ninety days. We know neither your market nor your competition, nobody knows them in advance, and a figure advanced now would become the measure of work that has not started.
Finally, eight of the twelve weeks described here need nobody. Deciding a number, writing a sentence, writing five answers, asking every customer how they heard of you: that is free, it is the substance of the sequence, and it is what is missing in most of the businesses that call us after a year.
Frequently asked questions
Where do we start if we genuinely have nothing?
With the number and who answers it, before anything visible. The test is available today: ask somebody who does not know you to call your business at three in the afternoon and again at seven in the evening, and listen. Anything built on top of a number nobody answers increases the number of disappointed people.
Do we need a website from the start?
For most businesses starting here, no. The business listing carries the address, the hours, the number and the reviews, and it appears without being looked for. A site becomes necessary when your trade requires one: a catalogue, a booking, an online sale. Otherwise it can wait three months without harm.
When can we start advertising?
After weeks 5 and 6, never before. Paying to send people to a number that does not answer and a sentence that says nothing is the only expense in this list that actively makes things worse: it produces disappointed people instead of producing nothing, and a disappointed person costs more than a stranger.
What do we do with no references to show?
Say so, and replace proof with a verifiable commitment: what you undertake to do, and what happens if you fail. A finished job, premises, a workshop, a machine, somebody at work — all of that happened and can be shown. A borrowed reference or a review you wrote yourself is never defensible.
Why is week 9 so important if it produces nothing?
Because it is the only step whose omission gives no symptom for years and then one final one: the person who held the accounts is no longer there and the business can no longer edit its own listing. An hour at the start, very hard afterwards.
How do we know where customers come from without a measurement tool?
By asking, once, every person who buys, and writing three words in a notebook. That count beats any tool at this stage because it covers the channels that leave no trace — word of mouth, the sign, the neighbour — and those are often the main ones.
Where we come in
Have somebody who does not know your company ring it today. What they get in thirty seconds is what the market gets too.
- We write the verifiable sentence in both your working languages.
- We build the business listing once, properly, on your own account.
- We leave you the eight weeks out of twelve that need nobody.
We will not sell you advertising three weeks in: it is the spend regretted fastest when nothing is yet ready to receive it.
Read next
Your sector is not on the list: what holds for everybody
Nineteen articles about nineteen different trades. Here are the eight things that turned out to be true in all of them.Why your website is slow — and what it costs you
Site speed is not measured on your desktop, and that is where the whole misunderstanding starts.A website does not break down, it goes out of date
Almost every website failure is silent: nothing stops, nothing alerts, and nobody notices before a customer does.
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.