Engagement proposal · 24 August 2026
Slow
Assets
A rebuild, a publishing engine, and a partnership — priced so the middle one runs long enough to work.
| Website rebuild | Next.js, bilingual English and Arabic, hosted and maintained by us | $400once · half to start |
|---|---|---|
| Monthly retainer | An article every three days, the podcast, the posts, and everything the site runs on | $199per month · 6-month minimum |
| Referral commission | On anyone you introduce, for as long as they stay a client — paid to you | 20%of what I bill them |
| Paid acquisition | Meta. Strategy, creative, two shoot days a month. Not now — section 07 | 20%of ad spend · $2,000 floor |
| Video editing | Per finished video, if you decide to shoot them. Section 08 | $20each · as they happen |
All figures in US dollars. The first two are the engagement; the last two are optional and separate, and nothing above depends on them.
01 · The recommendation
Build the things that keep working
after you stop paying attention to them.
What I am not selling you
- Daily posting, daily ideation, and a camera pointed at you most mornings
- Distribution you rent from a platform that changes the terms every quarter
- Reach that decays to nothing within days of publishing
- A machine that stops the week you get busy — and you will get busy, because your product is you, in a room, for a full day at a time
- The month where managing the content becomes more work than managing the business
What I am
- A site you own outright, on infrastructure that is somebody else's problem
- An entity that search engines can resolve: who you are, what you sell, where, for how much
- A library that gets more valuable every month rather than less
- Work that keeps producing enquiries while you are in a training room in Amman with your phone off
- The thing a language model quotes when someone asks it who trains sales teams in the Levant — which is a question being asked, today, by people who used to open Google
The catch, said once and up front
Slow is slow. Nothing in this document produces a result in three weeks. The first ten articles will be read by almost nobody, and that is the expected outcome, not a failure of the method. Both search engines and answer engines reward a track record, and a track record is the one input that cannot be bought or accelerated — only accumulated.
Month four is where it starts to look like anything. That is not a hedge to protect myself later. It is the reason the retainer asks for six months instead of one, and the reason it is priced the way it is.
02 · What changed
The survey said keep WordPress.
This document says Next.js.
That is a reversal, and it is the first thing your current developer will notice, so it goes at the front rather than buried in a scope list. The survey of 16 August closed section 04 with one decision it could not make on your behalf:
“Who edits the site after launch? It does not change the platform, only where the budget goes.”
The survey, section 04
That question offered two answers and you gave a third. Not you edit it, and not quite we maintain it, but we take all of it — the site, the hosting, the content, and the publishing schedule. Once nobody at Skills 4 Excellence opens an editor, the case for WordPress is gone, because that case was never technical. It was about who could operate the thing.
The question was about budget. The answer turned out to be about the platform.
The argument for WordPress was never about WordPress
It earns its keep when a non-technical owner needs to change a paragraph at 9pm
without calling anyone. That was the whole of section 04, and it remains a good
argument — for a site its owner edits. It stops applying the moment publishing
is my job and my invoice. Nothing else on the survey's list of assets was ever
WordPress-specific: schema, semantic HTML, llms.txt and a content
library are all just server-rendered pages, and they are easier to guarantee when
the build is under my control.
A second language moved from Phase 3 to core scope
Because the market did. Syria, Jordan, Iraq and the Gulf are not an eventual
expansion of this business, they are where it already works, and half of that
audience will read an English page and quietly leave. Proper bilingual routing
— separate URLs per language, correct dir="rtl", per-language
metadata and hreflang — is a structural feature in Next.js and
a plugin negotiation in WordPress. Polylang on a theme works. It does not work
well, and it is the first thing to break in a redesign.
Content now arrives as a pipeline, not as a person typing
An article every three days is roughly 120 pages a year landing on a schedule. Handled as files rather than database rows, each one gets a preview link you can comment on before it exists publicly, a version history, and a rollback that takes seconds. That workflow is section 04 of this document, and the site has to be built for it rather than have it bolted on.
I will not carry a number I cannot control
The survey's most expensive single finding was that your server takes 1.5 to 3.1 seconds to return the first byte, and it said plainly — at cost to itself — that this was hosting and not framework. That is still true, and it is still the largest available win. It simply stops being an argument for staying put once the hosting is mine, because I am not taking responsibility for response times on infrastructure I did not choose.
What has not changed
- Every finding in the survey still stands. None of them were platform artefacts and none of them go away by moving
- The Phase 1 fixes still happen first, on your existing site, in week one — before the rebuild starts
- Your gold stays. Same hue, restored to the chroma it should have had, so it passes contrast instead of failing it. It is recognisably yours
- Your testimonials, your photographs, your domain and your email all carry over untouched
What this costs you to switch
- The ability to edit pages yourself in a browser. You are trading it deliberately, and if that ever stops being acceptable, say so and I will build you an editor rather than pretend the trade did not happen
- Your existing plugin licences, which were doing the damage anyway — Elementor, ElementsKit, Premium Addons, TablePress
03 · The rebuild
A website pointed at the market you
actually sell to
Modern, fast, and bilingual by construction rather than by translation plugin. Everything underneath it — hosting, delivery, security — moves to us and stops being something you have an opinion about.
Two languages, five countries, one site
Lebanon, Syria, Jordan and Iraq, and then the Gulf. Those are different buyers with different procurement habits, but they share a language, and right now the site does not speak it. The Arabic edition is not a translated copy sitting at a subfolder nobody links to — it is a full parallel site with its own URLs, its own metadata, its own structured data and its own entry in the sitemap, so it can rank on its own terms in its own market.
The same page, both editions
Leadership and sales training, built for how business is actually done in the Levant and the Gulf.
Twenty years across pharmaceuticals and insurance, in Lebanon, Jordan and Iraq. Delivered on site or online, in Arabic or in English.
تدريب القيادة والمبيعات، مصمّم لطريقة العمل الفعلية في بلاد الشام والخليج.
عشرون عاماً من الخبرة في قطاعي الأدوية والتأمين، في لبنان والأردن والعراق. تُقدَّم البرامج حضورياً أو عبر الإنترنت، بالعربية أو بالإنجليزية.
Set in Tajawal
and marked lang="ar" dir="rtl", so a screen reader switches voice and the
browser mirrors the layout without being asked. Translation is done by a person and
reviewed by you — a machine translation of a leadership programme reads, to an Arabic
speaker, exactly as badly as you would expect.
Every form goes to WhatsApp, not to an inbox
This is the one decision in the rebuild that is regional rather than technical. In this market a corporate enquiry that arrives by email waits; the same enquiry on WhatsApp gets answered within the hour. You already run the business from WhatsApp. Building a very fast website that deposits leads somewhere you do not look would be reproducing finding 02 of the survey with better typography.
So every form — programme enquiry, coaching request, callback — opens WhatsApp with the name, the organisation and what they asked for already written into the message. Nothing to log into, nothing to check, no notification you have to remember to enable.
The email address still gets built properly, for two reasons: a corporate L&D buyer will eventually need to send a purchase order to an address, and the survey found the founder's published address bouncing. It simply stops being the channel the business depends on. Karen holds the mailbox — she administers it and forwards you the two or three a month that actually need you.
Things that stop being yours
- Hosting and deployment
- CDN and edge caching, so a visitor in Riyadh is served from near Riyadh
- TLS certificates, and their renewal
- Firewall, bot filtering, and rate limiting on the forms
- Daily backups, and a rollback that takes about a minute
- Uptime monitoring that pages me, not you
- A performance budget enforced at build time, so the site cannot quietly get slow again the way the last one did
You are not asked to hold an account with any of it, or to have a view on it. If it breaks at 2am, it is mine.
Carried over from the survey
- The Phase 1 fixes, done first on the existing site: the dead
carousels, the bouncing address, the failing contrast, the
[2023]placeholder - Hand-written schema — Organization, Person, Service, Offer, Review, FAQ — rather than the Yoast stub
- Fifteen programme pages replacing the fee table, each with a price and a currency symbol
- One reconciled vision and mission, stated once
- WCAG 2.2 AA, verified rather than claimed
llms.txt, a clean sitemap, and redirects from every existing URL so nothing you have already earned is lost
| Design and build, bilingual, to launch | $1,600 |
|---|---|
| Arabic edition — routing, RTL, translation review | $450 |
| Schema and entity layer | $250 |
| Migration, redirects, launch | $200 |
| What this would be quoted at | $2,500 |
The gap is real and I am not going to dress it up as a launch promotion. Section 06 is where I explain what I want instead of the difference.
04 · The engine
One article every three days,
and everything else falls out of it
This is the part that produces the result, and the part most arrangements abandon in month two. So it is built as a loop with exactly one person's judgement in the middle of it — yours — and everything around that automated or mine.
The topic
Pulled from a ranked list of questions your buyers actually type, filtered by what this site can realistically rank for in the next year. Not what is interesting — what is winnable.
MineThe draft
Researched, written, and put on a preview link before it exists publicly. Written with a language model, and I am not going to pretend otherwise — I cannot write in your voice, and neither can it.
MineYour review
You tell me what is wrong: the claim that is not how it works in Amman, the example that should have come from pharma, the sentence you would never say out loud. About twenty minutes. Nothing publishes without this step.
Yours — the gatePublish
Corrected, published, submitted, and linked into the rest of the library. Three days later, the next one. Ten a month, about a hundred and twenty a year, indefinitely.
MineYour voice
You read the piece aloud and talk around it — the story that did not fit in the text. Fifteen minutes into your phone, whenever suits. Optional per article.
YoursEverything downstream
Your recording becomes the podcast episode: edited, artwork, published, distributed. The article becomes three LinkedIn posts, one a day until the next one lands, written and handed to you in a document for you or Karen to schedule.
MineWhy three days, specifically
Three days is not a rhythm chosen for comfort. It is the fastest cadence that still reads as a publishing operation rather than a dump. Forty articles appearing in one week is a pattern search engines are explicitly built to detect, and the consequence is not a warning — it is silence, applied to the whole domain. Ten a month arriving on a schedule, month after month, is the profile of a firm that publishes.
It also divides exactly into the thing downstream: three days per article, one post per day, no gaps and nothing invented to fill them. You are never asked what to post today, because today's post already exists.
About the machine, stated plainly
The drafts are written with a large language model. That is how ten articles a month is possible at this price, and concealing it would be both dishonest and trivially easy to catch.
It is also, on its own, not good enough. High-volume publishing of unreviewed model output is the clearest pattern search engines now demote, and it deserves to be demoted, because it is indistinguishable from every other firm's unreviewed model output. What makes this different is steps 3 and 5 — your corrections, your examples, your recorded voice on the piece. That is not a formality performed to satisfy an algorithm. It is what makes the article worth reading, and the algorithm is only trying to detect the same thing I am.
Which is why every article is attributed to you as a named,
resolvable author, tied to the same Person entity the survey recommended
building — linked to your LinkedIn, your AMI certification, and the twenty years the
current site buries. The claim that this is Fadi Tuffaha's expertise has to be true, and it
has to be machine-checkable. Both halves are in scope.
The byline as it will appear on every article, and the structured-data properties that will carry it. Those four fields are the difference between content that is merely published and content that is attributable — which is what both a reader and a language model are checking for.
What this actually costs you in time
The real objection to any content arrangement is not the price, it is the suspicion that it will quietly become a second job. So here is the whole of it, measured rather than promised.
Between four and six hours a month, depending on how many pieces you decide to voice. Call it an hour and a quarter a week, in whatever gaps you already have.
05 · The retainer
What arrives every month
One number covers the engine, the podcast, the posts, and everything the website runs on. Itemised, because you should be able to see what each part is worth on its own.
| Ten articles — researched, drafted, revised, published | $550 |
|---|---|
| Podcast — editing, artwork, publishing, distribution | $200 |
| Thirty LinkedIn posts, written and handed over | $150 |
| Site maintenance, updates, fixes | $100 |
| Hosting, CDN, TLS, firewall, backups, monitoring | $60 |
| Quarterly measurement against the survey's targets | $40 |
| What this would be quoted at | $1,100 / month |
Six months, then monthly
The minimum is not a lock-in device and I would drop it if I thought the arrangement could be judged sooner. It cannot. Six months is roughly sixty articles, twenty-odd episodes, and the first point at which there is enough search data to see whether the curve is bending. Judged at eight weeks this will look like a failure, because at eight weeks it is one.
After the six months it runs month to month, and either of us can end it with thirty days' notice. Everything published stays yours: the articles, the episodes, the site, the domain, the analytics history. There is no version of this where leaving costs you the library.
What is not in it
A price this low only survives if the edges hold, so here they are. I would rather be boring about this now than awkward about it in month three.
- Paid advertising of any kind — that is section 07, with its own terms
- Video editing — section 08, twenty dollars a piece
- New pages beyond the site as built. A new programme page or a campaign landing page is a number, quoted when you want one
- Translating your existing course material, workbooks or slide decks
- Design or build work for anything that is not this website
- Being in the room. On-site days, delivery support and client meetings are yours — except where section 06 applies
Everything on that list is a conversation and a price, not an argument. What the list protects against is the version where it all quietly becomes included, the arrangement stops being worth doing, and I get slower and stop replying. That is how these end, and naming it is the cheapest way to prevent it.
06 · Why the price is what it is
This is partner pricing. Here is what I want instead of the difference.
You have now read two numbers that do not add up: a twenty-five-hundred dollar build for four hundred, and an eleven-hundred dollar month for a hundred and ninety-nine. I would rather explain that than let you wonder about it, because an unexplained price that low is a good reason to distrust the work behind it.
It is not an introductory rate and it does not expire into a real one. It is what this costs when the relationship is worth more to me than the margin — and I am going to tell you exactly which relationship, so you can decide whether you want to be in it.
You send me the ones who need it
Your consulting work, your list and your training rooms put you in front of firms whose web presence is in roughly the condition yours was in on 16 August. When it comes up — and it comes up — I am the name you give them.
Of everything I bill a client you introduced, for as long as they remain a client. Not a finder's fee on the first invoice — a share of the account, paid monthly against invoices you are welcome to see.
I show up as your technical side
When it helps you win the room, I am on the call: the person who answers the website, search and AI-visibility questions while you stay on the training. You get to walk into larger accounts with a build capability behind you, without having hired one.
An exclusivity, or a debt
You are not obliged to refer anyone. There is no target, no minimum, and nothing in the retainer changes if you never send a single client. If it turns out there is nobody to refer, we both find that out and the price stays what it is.
$199 is not what this work costs. It is what it costs when I would rather have the next ten clients than the difference.
07 · Later
Paid acquisition
Not now, and not part of anything above. It is written down because we discussed it, and because you should know the terms in the month before you want it rather than during.
Meta, and why it is Meta
For a Levantine and Gulf audience buying corporate training, Meta is where the attention is and where the cost of a qualified conversation is lowest, by a distance. Google captures people already searching for a named solution, which for leadership training is a thin and expensive layer. LinkedIn has the right job titles and prices itself accordingly. Meta has effectively everyone, in Arabic, on their phones, at a fraction of the cost — with the best targeting and creative-testing machinery of the three. That is the answer for this market and this offer. It would be a different answer for a software product in Germany.
The fee: 20% of spend
| Monthly ad spend | My fee | |
|---|---|---|
| $2,000 | $400 | The floor — everything included below |
| $3,000 | $600 | |
| $5,000 | $1,000 | |
| Below $2,000 | — | Declined. See below |
Below $2,000 a month in spend, I will not run this.
I want to be precise about why, because it sounds like holding out for a bigger commission and it is not that.
You sell high-ticket corporate programmes. Nobody buys one from a single advertisement. The path is several steps — something that earns attention, something that proves you know the subject, something that asks for a conversation — and each of those is a separate audience with its own creative, which has to gather enough data for the platform to learn who responds to it. Starve them and every layer stays stuck in the learning phase, and the algorithm never receives a signal strong enough to optimise against.
The result of a thousand-dollar month is not a small result. It is no result, and a thousand dollars gone.
Two thousand is the point where each layer can be funded properly and the thing has a real chance. Below it I would be taking a fee to run something I already know will fail, and then having a difficult conversation in month three. I would rather have the easy one now.
What $400 a month buys at the floor
- The strategy: the offer, the audiences, the funnel structure, and what each layer is for
- Creatives made — written, designed and cut, not briefed out
- Two shoot days a month, on location with you, for new footage
- Campaign landing pages and forms, wired to WhatsApp like the rest of the site
- Conversion tracking that actually attributes, so we can tell what worked rather than guess
- Weekly iteration on what is live, and a monthly report in cost per qualified conversation
On webinars, since you did not seem to want one
The true thing first: for education, coaching and consulting, a live webinar is the strongest selling instrument there is. In ninety minutes it does what no landing page can do at all — it lets the buyer experience you teaching, which is the actual product.
You did not seem to want to run them, and I am not going to spend a package arguing with you about it. So, the alternatives, honestly ranked:
- A recorded video sales letter — most of the mechanism, none of the calendar
- A free short course by email — slower, but it builds a list you own outright
- A straight lead form to a call — the simplest, and it converts the worst
All three work. The first works best of the three. It is genuinely your call, because you are the one who has to show up.
Entirely separate from everything above. The retainer does not require this, and this does not require the retainer. Either can start, stop, or never happen without touching the other.
08 · Also available
Video editing, $20 a piece
If you want to take the core idea of an article and say it to camera in your own words — which, for your business, genuinely does work, on Instagram and on LinkedIn both — I will edit it. Cut, captions in Arabic or English, titles, sound, colour, delivered ready to post. Twenty dollars a video, billed as they happen, no commitment and no minimum.
The honest limit: I cannot do the part that matters. The idea, the framing, and the first fifteen seconds that stop someone scrolling are yours, and no editor supplies them. If you shoot it, I will make it look right. If you would rather not be on camera at all, skip this entirely — nothing else in this document depends on it, which is rather the point of the whole approach.
09 · What happens next
Four steps, and the first one is a message
You say yes, to all of it or to part of it. The build and the retainer are separable. The build on its own is a perfectly reasonable answer and I will not treat it as a negotiating position.
The Phase 1 fixes go onto your existing site. The dead carousels, the bouncing
address, the failing contrast, the [2023] placeholder. Those happen whatever
else you decide, and they are already inside the $400.
The rebuild. You see it on a preview link throughout, in both languages, on your own phone. It does not launch until you have said yes to it.
The engine starts. First article on day one, the next on day four, and the first measurement report ninety days later — against the same numbers the survey was measured with, not against impressions.
Say yes to the parts you want. This is not a bundle.
Everything here can be taken separately. If today's answer is the $400 rebuild and nothing else, that is a real answer and I will build it properly. If it is the rebuild and the engine, that is the one I would pick. The advertising can wait until there is something worth advertising to.
Reply on WhatsAppWhich is, appropriately, the channel the whole document argues for.