This is the digital marketing code of conduct I work by. And before I begin it, I offer you a salute.

Not a courtesy. In military custom a salute is not deference paid by the junior to the senior — it is a mutual acknowledgement that each of us knows his place and keeps to his limits. That is why an officer returns it to a private, and why it is not waived by difference in rank.

I offer it because you are about to read what I say about myself — and you are entitled to know where these rules come from before you read them.

This is not a decorative phrase. It is the foundation every army is built on. A weapon in the hands of a man without doctrine is not strength — it is a danger to the one carrying it before the one facing it.

Which is why the order of training is settled: discipline first, then the weapon. The weapon is not issued to anyone who has not passed the first. And the rule in training is that what a man endures in camp decides what he is worth in the field — a man who takes discipline lightly in peacetime takes duty lightly when it counts.

Doctrine is not a restraint on capability, as those who have never served tend to assume. It is the condition of its survival. A disciplined unit still has its capability after the first engagement. A unit that let its discipline go is beaten from the inside before the enemy reaches it.

I found that same rule, word for word, in a field where no one carries a weapon.

Since I am borrowing the vocabulary of the military, honesty requires that I mark its limit before I borrow it.

That is true where it belongs. And it is precisely what does not belong here.

Deception is legitimate in war because the other party is an enemy who means you harm. The man in front of me in my work is no enemy — he is someone who trusted me with his money and his reputation because he has no way of judging what I do.

The enemy may be deceived. The client may not.

Anyone who borrows the discipline of that school and borrows its stratagems along with it has turned his clients into adversaries while believing he is doing well by them.

What I take from it is the discipline, the rules of engagement, the honesty of the report, and the responsibility that rank does not lift. The stratagem stays where it belongs — on a battlefield that does not exist in my work.

Professions that touch people’s money have systems of accountability. A doctor answers to his board, an accountant to a regulator, a serving officer to a disciplinary code that does not do favours.

Digital marketing has none of that. No licence to withdraw, no register to be struck from, no body to answer to. Anyone who wants to present himself as an expert today needs nothing more than to say so.

That leaves the client facing a question he has no means of answering: how do I know this man is not using a method that will take my account down in six months?

Where no authority imposes a doctrine, what remains is for the practitioner to write his own and be held to it.

This document does not vouch for me — a man’s testimony about himself is not accepted. It is a document that can be used against me. If I breach a clause in it, there is a published text with a date on it.

Military practice does not confuse two things, however similar their results:

Reconnaissance reads what is exposed, gathers what has been put on show, and builds its assessment on what can be seen. It is a duty, not a merit — a commander who engages without it has failed in his job.

Intrusion enters where it has no permission.

What separates them is not the difficulty of the act, nor the chance of being caught. It is that the first reads what was offered to it, and the second takes what was not.

This is the line I stop well short of

In “competitor analysis” the two get confused often, sometimes in good faith. So I set the boundary in plain terms:

Legitimate reconnaissance	Intrusion I will not go near
Reading a competitor’s published site	Accessing an ad account
Analysing their search rankings	Entering an analytics dashboard
Studying ads they run in public	Viewing a client list
Examining their published content and prices	Any internal system
Analysis tools that work on open data	Buying data obtained that way

Loose talk about “hacking” attaches itself to people who work in this field, and I think the answer to it is plainness rather than hints. So I will say it once, without qualification:

And I will add this: were I offered a project at twice my rate on condition that I cross that line, I would turn it down. Not because I fear exposure, but because a man who sells his line once has no line left to sell the second time.

Technical capability is not permission. A man who has the capability and nothing to stop him is short of doctrine, not skill.

Talking about integrity costs the speaker nothing. What costs something is publishing what harms you.

So I am not asking you to believe any of the above. I am asking you to check these five facts:

These lines are not crossed. Not at a client’s request, not under competitive pressure, and not for a project of any size.

People say this field has a “grey area” that is hard to judge. That is true of some of it, and overstated for most.

Most of what gets called grey is not unclear to the person doing it. It is unclear to the person it is being done for. Vagueness is not a property of the act — it is a space inhabited by people who would rather not be asked.

So I do not test a method by asking “is this explicitly against the rules?” That is the question of a man looking for a way out. I test it with two questions:

First: can I explain it in full to the client, to the platform, and to any professional body — without leaving out a single detail?

Second: if it came to light tomorrow, would it harm the client I did it for?

If the answer to the first is no, or to the second is yes, then the area is not grey. It is perfectly clear, and I was the one trying not to see it.

And when the client himself asks for the questionable route

It happens. And the request usually comes without bad intent — a man has heard there is a faster road, and does not know its price.

My duty then is not refusal alone. It is to explain what he cannot see: that the result will come, that it will go, and that what remains is the suspended account, the burnt domain, or the reputation that is hard to repair.

Refusal without explanation is disdain. Explanation with refusal is advice.

Automation is not the problem. It is among the most useful things in this field: thousands of rows processed in seconds, a repetitive task run without error, and time freed for work that needs a mind.

The dividing line is a single question: what does the automation do to reality?

Speeds up something that genuinely happens	Development
Manufactures activity that never happened	Manipulation

Same tool, same code, same technical competence. The whole difference is in the intent and the effect.

AI tools have made one person capable of what used to take a team. I use them to the fullest: data analysis, audience research, market research, developing ideas, automating processes, analysing sites, improving content, and building the tools themselves by hand.

Greater capability means greater responsibility, not less.

A model is wrong with confidence. It gives you a figure with no source in the voice of certainty, invents a reference in correct form, and offers a conclusion that is sound in shape and rotten at the base — and it does not tell you when it has done so.

Hence: the output is a draft, not a result. It is reviewed, every figure and every source in it is verified, and the responsibility stays with the man who signed it, not the machine that produced it.

Anyone who hands a client an output he has not read has not used AI — he has used it as cover for not doing the work.

Disciplined work does not begin with execution. It begins with an appreciation of the situation: what the position is, what is available, what the realistic timeframe is, and what cannot be guaranteed at all.

What gets hidden most in this profession is not what we do, but what we cannot do. So I state it:

• Google AdsVisits and calls within 24 to 48 hours, but judging the campaign takes two to four weeks.
• Meta AdsEngagement from the first days, real conversions from week three or four, and the true result from month two.
• SEOThree to six months for real effect, longer in crowded sectors. The shortest sensible engagement is six months, and anyone promising you less is selling you time, not results.
• Google Maps & local presenceFour to eight weeks.
• Content marketingEngagement from month one, search results from month two to four.
• Website designThree to six weeks, and a single landing page in one to two weeks.
• Building AI toolsTwo to six weeks depending on scope.

These timeframes are written on my service pages, so any promise faster than them creates a contradiction you can see with your own eyes on this same site.

I do not claim I get nothing wrong. Anyone who claims that has already got the first sentence wrong.

But there is a distinction anyone who has served knows:

An error of judgement is reading the situation with the information you have and reading it wrong. It happens, it is corrected, it is learned from, and it goes into the report as it is, without cosmetics.

An error of intent is knowing and doing it anyway.

The first gets reviewed. The second never does, because the man responsible never files it.

Do not believe any of the above. Check it:

• Run Digital Maraya on my own brand. The tool is on this site, and my result in it is published.
• Open any of my tools and read what it says it does not do. The limits are written inside each one.
• Ask the AI assistant whether Ashraf appears in ChatGPT. You will get an answer against my interest, because I wrote that into its instructions.
• Verify the intellectual property certificate by its number with the Ministry of Economy and Tourism.
• And ask me to walk you through a method I have used. If I cannot explain it in front of you, or in front of the platform, or in front of any professional body — it is a method I should never have used in the first place.

I may disagree with you. I may tell you your idea will not give you what you expect. I may refuse to run a method I consider misleading, however much you insist. And I may tell you the result needs time rather than hand you a quick promise that suits you today and costs you in two months.

I may point you towards what earns me less: if you need customers quickly, SEO is not your answer right now and advertising suits your stage better — and I will say so even if you came asking for SEO.

This is not making things difficult. It is the difference between a man who sells you what you asked for and a man who gives you what you need.

I work in a field that changes every year: new tools, new platforms, algorithms rewritten, and AI moving faster than anyone can comfortably follow.

One thing I do not want to change: the doctrine I work by. Tools change and methods change, and a doctrine that changes was never a doctrine — it was a circumstance.

And if you find anything in my work that contradicts what is written in this code of conduct, tell me. This document is worth nothing if it cannot be used against me.

And I close as I began — with a salute.