Illustrative scenario: not a client case study
Separate Fact From Interpretation From Promise
A sentence like "we are the leading digital marketing agency in Dubai, trusted by dozens of businesses to deliver results that matter" mixes three different kinds of claim: an unverifiable superlative, an interpretation, and an implied promise. Nobody can cite that sentence safely, because there is no single checkable fact inside it.
Rewrite it as separate statements: a fact ("gang&lani media has worked with businesses across Dubai and Abu Dhabi since our founding"), a described capability ("our services include SEO, paid media and social strategy for UAE service businesses"), and, separately and clearly labelled as such, a promise or opinion about outcomes. A journalist or an AI system can quote the fact confidently; they will usually skip or heavily qualify the promise, which is the correct outcome.
Show Real Authorship, Not an Anonymous "We"
A page with no named author, no visible expertise, and no way to verify who wrote it is a weak citation source, both for a human journalist checking credibility and for an AI system that increasingly weighs signals of genuine authorship. Where you can, name the person responsible for the content, with a short, honest description of their role, not an inflated title.
This doesn't mean every blog post needs a full biography. It means the site as a whole should make clear who stands behind the claims, through a consistent about page, named contacts for specific services, and content that reads like it was written by someone who actually does the work described.
Use Primary Evidence Instead of Restating Received Wisdom
A huge amount of marketing content on the web restates the same secondhand claims about "UAE consumers" or "Gulf shoppers" without any original source. This kind of content is nearly impossible to cite responsibly because it has no traceable origin, and citing it just propagates an unverified claim further.
Where you can, cite something checkable: a specific, named piece of public guidance, such as Google's published guidance on AI features and websites or Bing Webmaster Guidelines, a government portal's stated licensing requirements, your own direct observation from work you've actually done, or a clearly labelled opinion. Do not invent a statistic or a study to sound more authoritative; an invented number is worse than no number, because it eventually gets checked and damages trust when it doesn't hold up.
Keep a Claim Register So You Know What You've Said
As a site grows, it becomes easy to lose track of every specific claim made across dozens of pages, an old service description here, a stated turnaround time there. A simple claim register, a spreadsheet listing each specific factual claim, which page it's on, and when it was last verified, prevents a small business from accidentally contradicting itself across its own website.
Review the register whenever a core fact changes, pricing, service scope, team size, credentials, and update every page that states it, rather than fixing the page someone happens to notice. This is the same discipline as entity clarity, applied to claims rather than identity facts.
Write Numbers and Dates the Way a Citation Needs Them
A number without a date attached is not a durable fact; a stated percentage says nothing about when that was true or where it came from. If you state a number at all, state where it came from and, ideally, when it was published, so anyone quoting you can judge whether it's still current.
The same applies to your own operational facts. "We respond to enquiries within one business day" is a checkable, citable claim; "we respond quickly" is not. Specificity is what makes a sentence worth quoting.
Make Corrections Easy to Find and Trust
Every business occasionally publishes something that turns out to be wrong or out of date. A page that shows when it was last reviewed, and that corrects errors visibly rather than quietly editing them out, is a more trustworthy source for both journalists and AI systems, because it demonstrates the content is actively maintained rather than published once and abandoned.
Build a light process: when a factual error is found, fix it, note the date of correction where appropriate, and check whether the same error was copied onto any other page. Treat a caught mistake as useful information about where your content process needs tightening, not as an embarrassment to bury quietly.
What to take away
- Separate verifiable fact, interpretation and promise into distinct sentences.
- Name the person behind the content; anonymous "we" pages are weak sources.
- Cite real, checkable evidence, never an invented statistic or study.
- Keep a claim register so facts stay consistent across every page.
- Attach a source and, where relevant, a date to every number you publish.
- Show when a page was last reviewed and correct errors visibly.
Frequently asked questions
Does this apply to every page, or just blog articles?
It matters most on pages likely to be quoted or extracted: service pages, FAQ pages, about pages, and anything answering a specific customer question. A purely visual landing page carries less of this burden.
Isn't separating fact from promise just weaker marketing copy?
It's different marketing copy, not weaker. A clearly stated fact next to a clearly labelled promise is often more persuasive than a vague superlative, because it gives the reader something concrete to evaluate rather than something to be suspicious of.
What counts as a "primary source" for a UAE business?
Your own direct experience and data, official government or regulator guidance, and named, dated, publicly verifiable publications from recognised platforms, such as search engine guidance documents. Secondhand marketing claims repeated across other agencies' blogs do not count.
How often should the claim register be reviewed?
Whenever a core operational fact changes, and as a fallback, at least twice a year even if nothing has obviously changed, since small drifts, an old price, a discontinued service, tend to go unnoticed longer than expected.