Digital marketing · GCC

Marketing Joined Up, Not Scattered

A digital marketing agency for GCC businesses, in Arabic and English. We do not sell channels. We fix the handoffs between the advert, the page, the form, the follow-up and the report, because that is where the money leaks.

5 seams, not channels

The handoffs between advert, page, form, follow-up and report. Each loses enquiries unnoticed

Fixed fee

never a share of spend

A percentage of media spend pays us to spend more, not better

3 months

initial retainer period

Enough to fix the seams and see one clean reporting cycle

Arabic

written, not translated

Advert copy, pages and follow-up written in Arabic from the start, English second

What this page argues

Key takeaways

Marketing rarely fails inside a channel. It fails at the joins between one team’s work and another’s.
A seam has two owners, which means none. Naming one owner per seam beats a bigger budget.
Every seam can be tested in an afternoon. The tests are on this page.
Channel choice follows the buyer. We say in writing when one you like is wrong for you.
We are a studio in Istanbul. No Gulf office, address or landline, and we will not pretend otherwise.
Everything we start gets a purpose, a review date and a condition that ends it.

What is on this page

Most marketing fails at the seams

Look at marketing that is not working and you usually find competent people producing decent work. The adverts read properly, the page looks right, the salespeople are not lazy. No single task is broken.

The joins are broken. Marketing is a chain of handoffs: one person writes the advert, another built the page it points at, a developer built the form, an inbox receives what it sends, a salesperson answers, an account manager reports a month later. At each handoff the work leaves somebody who understands it for somebody who understands only their own part.

A chain we have walked more than once. An Arabic advert, correctly written, pointing at an English page that is slow on mobile data, which is how the audience arrives. The form rejects the country code the visitor typed, with an error in English. What survives reaches an address created by an employee who left. In the report this looks like a healthy lead count at an acceptable cost.

Nobody did a bad job, and the chain still lost almost everyone. So we do not open by asking which channels you want: our Google Ads work and our SEO work begin by walking the chain on a real phone, in both languages. The website is a seam, not a separate project.

We will not tell you what share of your budget the seams lose, because we do not know, and nor does anyone who quotes a figure. We can show you where loss is possible, and how to check each point.

The seams, and how to test them

Each row is a handoff: the five between first touch and enquiry, plus the reporting loop that sets next month’s spend. The last column is a test you can run this week.

The handoff	What breaks there	Who owns each side	How to test it
Advert → landing page	An Arabic advert opens an English page, or one too slow on mobile data.	The media buyer and the web team, separately.	Click your advert on a phone, on mobile data, in each language.
Landing page → form	Eleven fields where four would do; a phone field refusing a Gulf number.	The developer, and sales, who needs the fields.	Submit a real enquiry by phone in each language, then mistype the number.
Form → inbox or CRM	Alerts reach a departed employee, a spam filter, or a silent CRM rejection.	Hosting owns delivery, sales the destination.	Send a test enquiry weekly; confirm it reaches where a salesperson looks.
Enquiry → follow-up	The reply comes days late, or in English to an Arabic enquiry.	Marketing creates it, sales answers it.	Log every enquiry and its first human reply for a month.
Outcome → ad platform	Nothing goes back, so the platform keeps optimising towards form fills.	Sales owns the outcome, the buyer the platform.	Check whether the platform knows last month’s won deals.
Report → next decision	Read, admired, filed. The losing activity runs again.	The agency reports, the client holds the budget.	Name one thing each of the last three reports changed.

Run all six and you have a list of faults a named person can fix on a stated date. Tell us what you are running and we will walk it.

First touch to enquiry

The boxes are the parts people are proud of. The circles are the parts nobody was asked to own.

Six steps, five seams. A failure at any circle looks identical to a lack of demand.

Which is why falling enquiries get misdiagnosed as a market or a creative problem. Both are interesting. A form quietly refusing a country code is not, and it costs more.

Why nobody owns a seam

Ownership inside a marketing function is usually clean. The agency owns the adverts, a developer owns the site, sales owns the phone, somebody owns the CRM. Each can say what they are responsible for.

A gap between two responsibilities has two owners, and a thing with two owners has none. Ask who is responsible for the advert and the page saying the same thing: the media buyer says the page is not theirs, the developer says the advert is not theirs, and both are correct. The seam is unowned by agreement, not neglect, and it worsens as suppliers multiply.

Our fix is unglamorous: for each seam, one name, one test, one frequency, written down. Sometimes the name is ours. Often it should not be, because the person who can change what happens between the enquiry and the call works for you.

Channel fit in the Gulf

What each channel is good at, bad at, and who it reaches. A channel that is wrong for you is a more useful finding than one that is merely plausible.

Channel	Genuinely good for	A poor fit for	The buyer it reaches
Google Search	Somebody who has decided they need this and is choosing a supplier.	Demand nobody searches for yet; tiny niches.	An active buyer, comparing suppliers today.
Instagram	Showing a product, a place or finished work. Recognition before need.	Specifications, and brands with nothing new.	Consumers and younger decision-makers, browsing.
TikTok	Reach at low cost, when the product is visual.	Procurement, regulated services, brands short of video.	Broad consumer audiences, heavily young.
LinkedIn	A named job title inside a named company. Partnerships, recruitment.	Local consumer trade: the click cost rarely repays a small sale.	Corporate, institutional, government-adjacent.
WhatsApp	Closing the enquiry: questions, catalogues, quotes, photographs.	Cold outreach, and lists built without consent.	Anybody who already contacted you. The Gulf default.
Email	Keeping the customers you have: prices, availability, real updates.	Cold prospecting here, ignored for a call.	Customers and contacts who asked to be listed.

Two notes. A channel you cannot answer is worse than one you never opened, so first we ask who replies, in which language, on which days. And weekend patterns differ between the Gulf states, so scheduling is set per market. If you sell online, the store build belongs in the same conversation.

Choosing channels, and saying no

The order of the questions matters more than the answers. Where is the buyer already, before where we would like them to be. Are you capturing demand or creating it, because those need different channels and very different patience. What does it cost to feed a channel every week. And who answers when it works.

That last question kills more channel ideas than any cost calculation. Short video needs new material continuously and looks abandoned the week you stop. A messaging channel needs somebody replying the same day, in the language it arrived in.

Sometimes a client wants a channel we think is wrong, usually one a competitor is visibly using. We neither agree nor refuse. We write down why, in a page: who it reaches, what feeding it costs, what we would give up to fund it. Then, if you still want it, we run a bounded test with a budget cap and a stop condition agreed in advance.

We say no to ourselves too. If your enquiry flow is broken or your tracking is guessing, we advise fixing those before adding spend. Where the money sits also differs by business type: a supplier selling to procurement in Dubai and a consumer brand split the same budget in almost opposite ways.

Where the budget goes

One monthly budget, drawn twice. An illustration of the shape of the decision, not data: nothing here is a benchmark or a recommended split.

Illustration only. Your split comes from your buyers, your sales capacity, and what you can produce.

The supplier puts most of its money where intent is already stated, and on writing that answers a technical question better than a competitor does. The consumer brand spends most on making things worth looking at, because there the creative is the campaign. Copying the wrong shape is the expensive mistake, usually made by copying a competitor who sells to somebody else.

Arabic across the whole chain

Bilingual marketing in the Gulf usually breaks at a seam, not in a translation. The advert is Arabic and the page is English. Or the page is Arabic but the form labels are not. Or everything is Arabic until the confirmation email, which arrives in English from a system nobody configured. Or the chain is Arabic and the person who calls back is not.

Each is a handoff, not a language failure, so we treat language as a property of the chain: an enquiry that begins in Arabic stays in Arabic through the page, the form, the confirmation and the first reply. If your team cannot staff that, we say so early.

Two things we insist on. Arabic copy is written, not translated: machine translation hurts relevance in search and reads as translated, which costs trust at the moment you ask for a phone number. And right-to-left is designed in, so fields, validation and error messages mirror properly. That is interface work as much as copywriting.

What we handle

Six areas. We can run all of them, or one beside a team that runs the rest.

Strategy and channel plan

Who the buyer is, where they already are, what is ruled out and why. Written down, so you can disagree with a sentence, not a mood.

Content in Arabic and English

Articles, service pages, product copy and social material, written in each language rather than translated, at a rate you can sustain.

Paid media

Google Search and Performance Max, plus paid social where the buyer is. Your accounts, your billing, a fixed fee.

Organic search

Technical fixes, bilingual keyword work and content that earns rankings over months. The slowest channel, usually the cheapest.

Pages and tracking

The seam nobody wants to pay for. Pages that load on mobile data, forms that accept Gulf numbers, a site you can change yourself.

Reporting and review

A monthly view of enquiries and cost per enquiry, open dashboard access, and a meeting whose output is a decision rather than a document.

Measurement, honestly

Reporting is the last seam and the one most often faked. The usual failure is numbers individually true and collectively meaningless: impressions, reach, engagement rate, a cost per lead counting submissions nobody called.

We instrument the chain instead: enquiries by source, what happened to each, how long until somebody replied, and what share a salesperson would call genuine. That last figure changes decisions, and only your team can supply it.

Attribution is a model, not a measurement. A buyer who saw a video in March, searched your name in May and called from a WhatsApp link in June will be credited to one of those, confidently, by any platform. We state the model and what it over-credits, and where we can we feed outcomes back, so the platform optimises towards customers rather than form fills.

Two constraints shape what we can collect. Consent is one, and Google documents the mechanics in its consent settings guidance. Data protection law is the other, and the UAE publishes an overview of its regime. Speed belongs here too, because a slow page loses the enquiry before tracking fires: Google’s Largest Contentful Paint guidance sets the thresholds we build against.

Retainer, project or in-house

Three ways to get this done, including the one that does not involve us.

Arrangement	Cost shape	Speed	When it ends	Where it fits
Monthly retainer	A flat fee, busy month or quiet. Easy to keep paying after it stops earning.	Slow first month, spent on repairs, then steady.	The work stops, the assets stay, in your name.	Content, paid media and reporting: weekly decisions compound.
Fixed project	Fixed price, fixed scope, paid in stages.	Fastest route to one defined outcome.	Nothing to unwind. Run it yourself or hand it on.	A strategy, a campaign build, an audit, landing pages.
In-house hire	Salary, tools, recruitment, management. Hardest to reverse.	Slowest to start, because hiring takes months. Then fastest daily.	The knowledge leaves with the person.	Volume enough to keep a specialist busy weekly.

The honest summary: a project is the right first step more often than agencies admit, because it leaves you something you keep and no year-long commitment. A retainer earns its place when there is something to do every week. Website prices are published on the pricing page; marketing scope is quoted per business, because no honest package price exists for work whose amount depends on what the first fortnight finds.

How an engagement runs

01 Walk the chain first

Before any spend changes we go through the seams on your real setup: click your adverts on a phone, submit your forms in each language, find where enquiries land, read the last three reports. About a week.

02 Fix the cheapest leaks

The repairs that cost least and return most: the form field, the delivery failure, the language break, the missing conversion signal. Rarely impressive, and they routinely move the numbers more than a campaign would.

03 Then spend, where the buyer is

With the chain holding, we agree the channel plan and budget shape in writing, with what we are deliberately not doing listed beside it. Each channel gets a purpose, a review date and a stop condition.

04 Review on a date that exists

Monthly, against enquiries and cost per enquiry, including your sales team’s judgement of quality. The output is a decision: continue, change, or stop.

How we decide what to stop

Most marketing budgets are a record of things once started. Nothing gets stopped, because stopping requires somebody to say a decision was wrong, while continuing requires nobody to say anything.

We remove that asymmetry by writing the stop condition at the start, when nobody is defending anything. Every activity gets three lines: what it is for, when we look at it, what would end it. A new channel test might read: to find whether procurement buyers respond here, reviewed after eight weeks, stopped if no enquiry a salesperson rates as genuine has appeared.

At each review we ask four questions in order. Is it producing enquiries somebody would take a meeting from? If not, is there a mechanical reason, or were we wrong about the audience? Is it consuming attention out of proportion to its result? And would we start it today? The last is the most useful, because it strips out money already spent, which is not a reason to continue.

Three things we stop quickly. Activity whose only product is reporting. Channels producing volume of the wrong kind, worse than silence because they consume the attention real enquiries needed. And anything only one person understands. Stopping frees budget, and attention, which matters more: a plan with four things gets done, a plan with eleven gets reported on. We write these up on the blog.

“Budget is rarely lost inside a channel. It is lost in the gaps between them, where nobody is looking.”
The argument this page is built on

When we are the wrong choice

Say no to us if any of these are true

You want the fee charged as a share of media spend. It pays us to grow your spending, not your results.
Nobody on your side can answer an enquiry within a working day, in the language it arrived in.
You want a guaranteed number of leads, or a guaranteed position. Anyone promising either is guessing.
Procurement requires a Gulf office, address or landline. We are in Istanbul.
Your site cannot be changed without a long approval chain. If we cannot touch the pages, we cannot fix the seam.
You want the reporting to look good more than to be right.

Questions we are asked

Do you work on a retainer or per project?

Both. Strategy, campaign builds, audits and landing-page sets are fixed-price projects. Content, social, paid media and reporting run as a retainer, because those are weekly decisions rather than a deliverable.

Can you work alongside our in-house marketing team?

Yes, and often the best arrangement. We can lead the plan or fill one gap, such as paid media or Arabic content. We first agree in writing who owns each seam, because that is where two capable teams lose enquiries.

Which channels should we be on?

Only the ones your buyers already use, and fewer than you expect. Business-to-business in the Gulf is often search, LinkedIn and writing that answers a technical question. Consumer brands, usually Instagram, TikTok and paid search.

How do you report, and how often?

Monthly, against enquiries and cost per enquiry rather than impressions and reach, with your sales team’s view of quality included. Dashboard access is open throughout.

What is the minimum commitment?

Projects are fixed scope. Retainers run an initial three months, roughly what it takes to fix the seams and see a clean month of reporting.

Will you ever tell us to spend less?

Yes, and we have. We have advised pausing paid media until landing pages and tracking were fixed, because spending into a broken chain buys more lost enquiries.

What do you do when we want a channel you think is wrong for us?

We tell you in writing before agreeing: one page on who that channel reaches, what it costs to keep fed, and what we would give up to fund it. If you still want it, we run a bounded test with a budget cap, a review date and a stop condition agreed in advance. Occasionally it proves us wrong, which is worth the money. What we will not do is run it for a year while neither of us decides.

How do you decide what to stop doing?

By writing the stop condition before we start, when nobody is defending anything. Each activity gets a purpose, a review date and a result that would end it. At review we ask whether it produces enquiries your salespeople would take a meeting from, whether there is a mechanical reason or we were wrong about the audience, whether it consumes attention out of proportion to its result, and whether we would start it today. Money already spent is not a reason to continue.

Do you have an office in the UAE or Saudi Arabia?

No. Filweb is a studio in Istanbul working Gulf hours, and we will not rent an address to look local. If procurement requires a locally registered supplier, we are the wrong choice.

Who owns the ad accounts, the analytics and the content?

You do, from the first day. We work inside your accounts rather than behind our own, and content, history and tracking stay in your name.

Do you write Arabic, or translate it?

Write it. Copy and pages are written in Arabic as their own version, with English in parallel rather than as the original. Machine translation hurts search relevance and raises click costs.

We already have an agency running ads. Can you just fix the seams?

Yes, and a sensible thing to buy alone. We walk the chain, list the faults, fix the ones that are ours, and specify the rest.

How quickly will we see results?

Seam repairs fastest, sometimes within a fortnight, because they recover enquiries you already pay for. Paid media takes six to eight weeks to optimise on real data. Organic search takes months; anybody promising otherwise is guessing.

Do you guarantee a number of leads?

No. We commit to the work, the method and the reporting, because those are ours to control. Enquiry numbers depend on your market, prices, offer and how fast you answer the phone.

Let us look at the whole picture

Tell us what you are running and who you want to reach. We will walk the chain, name what is leaking, and say whether you need us.