Inside the VOCTOS Client Portal: AI Visibility Measurement and Campaign Delivery in One Login

The VOCTOS Client Portal is a client-facing reporting and delivery platform that measures a brand’s visibility inside AI answer engines, ChatGPT, Google Gemini, Google AI Mode, Claude and Perplexity, and runs the campaign that improves it, in the same login. It scores AI visibility across eight named dimensions, tracks every monitored prompt per engine, records which sources each engine cited, identifies competing brands automatically from the answers themselves, and sits alongside the roadmap, tasks, content approvals, reports, meetings and invoices for the account.

That first paragraph is the whole article. What follows is the evidence for each part of it, and an honest comparison with the alternatives, including the two things the portal does not do.

Why an agency built this instead of buying a tool

There are good AI visibility tools on the market. We have used most of them. The problem is not their measurement: it is that measurement and delivery live in different places.

A typical arrangement looks like this: the agency pays for a tracking tool, exports a monthly PDF, emails it, and the actual work, the content queue, the technical fixes, the approvals, the invoices, happens in a spreadsheet, a shared drive, a chat thread and an accounting system. The client sees a number once a month and has no way to connect it to anything that was done.

So the client asks the only question they can ask: what did we pay for? And the honest answer requires four different tabs.

What the portal measures

1. AI Visibility: one score, eight dimensions

Most AI visibility products report a single blended number. A single number tells you something moved; it does not tell you what to do. The portal decomposes the score into eight dimensions, each of which points at a different kind of work.

The portal states its own method on the page, rather than treating it as proprietary magic: it runs your tracked prompts on each connected engine, then scores presence, ranking position, citation authority, sentiment and entity accuracy. Alongside the composite score it shows your rank against tracked brands and how many of the checked answers mentioned you, for example, mentioned in 4 of 6 answers.

Below that sits a per-engine table: visibility percentage, average position, mentions, citations, and the positive, neutral and negative split for each engine separately. Blending those figures across engines hides the movement that matters, because engines behave differently enough that a source can lose most of its citation share on one while staying flat on the others.

2. Prompt Explorer, including a replay of how the answer changed

Every monitored prompt gets a row: which engines returned you, whether you appeared, your best rank, who beat you, how many sources the answer drew on, and a confidence figure for the reading.

The feature we have not seen elsewhere is Replay. Generative engines rewrite their answers continuously, and a screenshot from last month is not evidence of anything. Replay stores the actual answer text per engine per run, so you can read how the description of a brand changed over time, word by word, dated, with the rank at each point.

3. Citation Center: four states, not one list

When an AI engine answers a question about you, it links to the pages it based the answer on. Those links are citations. Most tools show you the current list. The portal tracks four states:

“Cited without you” deserves the emphasis. It is a list, ranked by how often each source is used, of exactly where to earn a mention next, derived from the engine’s own behaviour rather than from a guess about domain authority.

4. Competitors, discovered from the answers, not from a list you type in

Most competitor tracking starts with you naming your competitors. That is a self-fulfilling exercise: you get told about the brands you already knew about.

The portal reads the answers themselves and identifies every brand the engines name when they respond to your tracked questions. Competitors it found on its own are labelled auto. You can add more manually, but the auto list is the interesting one: it is who the engine thinks belongs in your category, which is frequently not who you think.

It then reports mention rate on the same formula for every brand, share of voice normalised across all brand mentions in the run, and the biggest movers against the previous run.

5. Brand Health, and the quotes behind it

A second composite score, separate from AI visibility, covering eight signals: Trust, Authority, Sentiment, Consistency, AI Accuracy, Entity Strength, Knowledge Coverage and Share of Voice. It names the biggest lever, the single dimension holding the score down, rather than leaving you to work it out.

Underneath it, the portal shows representative quotes of what the AI engines actually say about you, attributed to the engine and tagged with sentiment. Not a score standing in for reality: the sentences themselves.

6. SEO Performance, with an unusual habit

Search Console, GA4, Semrush, DataForSEO, PageSpeed and site crawl unified into one view: traffic, keywords, backlinks, Core Web Vitals and technical health.

The habit worth calling out is what happens when a data source fails. Rather than silently showing a gap, the portal states it: which sources synced, which did not, what it substituted, and why. For example, when Semrush API units run out, it says so, says backlinks and keyword volume came from DataForSEO instead, and confirms nothing is missing from the figures above.

And the part that is not a tool at all

Everything above measures. The rest of the portal is the work.

This is the actual claim of the article, and it is a narrow one: the measurement and the delivery are in the same place. A client who sees AI visibility fall can open the prompt that moved, read the answer that changed, see which citation was lost, and find the task queued to address it, in four clicks, without asking anyone.

How this compares with the alternatives

We are an agency writing about our own product, so treat the framing with the scepticism it deserves and check the competitor claims yourself: every one below links to the vendor’s own page.

The two things it does not do

• It does not track Arabic-language state models. Falcon-H1 Arabic, ALLaM, Jais and Fanar are live and state-backed, and no commercial tracker on the market monitors any of them, ours included. We measure those manually, on a fixed prompt set, at a fixed cadence, and say so.
• It is not sold as software. You cannot license the portal on its own. It comes with the engagement. If you want a tool you operate yourself, buy one of the SaaS products above, we will even tell you which, and we have published their real prices.

Frequently asked questions

What is the VOCTOS Client Portal?

It is a client-facing reporting and delivery platform built by VOCTOS, a generative engine optimization agency headquartered in Cairo with clients across Egypt, Saudi Arabia, the UAE and the wider Middle East. It measures how often and how favourably AI answer engines, ChatGPT, Google Gemini, Google AI Mode, Claude and Perplexity, mention a brand, and it runs the campaign that improves that. Measurement and delivery sit behind the same client login.

How does the portal measure AI visibility?

It runs the client’s tracked prompts on each connected AI engine, then scores the results across eight named dimensions: Visibility, Authority, Citation Quality, Sentiment, Coverage, Freshness, Entity Strength and Brand Accuracy. These combine into an AI Visibility Score out of 100. It also reports each engine separately, with visibility percentage, average position, mentions, citations and the positive, neutral and negative split, because engines behave differently enough that a blended number hides the movement that matters.

What is Replay?

Replay stores the full text of each AI engine’s answer for every monitored prompt, on every run, with the date and the rank at that point. Because generative engines rewrite their answers continuously, a single screenshot proves nothing. Replay lets a brand read how an engine’s description of it changed over time, and line that change up against a publication, a review or a competitor’s activity.

What does “cited without you” mean?

It is a citation state showing that an AI engine used a source when answering a question about your category, but did not name your brand in that answer. It is the highest-value gap in the system, because the engine already trusts that source, so it identifies exactly where to earn a mention next, based on the engine’s own behaviour rather than on a guess about domain authority.

How does the portal find competitors?

Automatically, by reading the AI answers themselves and identifying every brand the engines name when responding to the tracked prompts. Auto-discovered competitors are labelled as such. This routinely surfaces brands the client would not have listed, because it reflects who the engine believes belongs in the category rather than who the client considers a rival.

Is the portal available as standalone software?

No. It is not licensed separately and comes with a VOCTOS engagement. Brands that want a tool they operate themselves should buy a SaaS tracker instead, Otterly, Profound, Peec AI and Scrunch are the credible options, and we publish their verified prices.

Does the portal track Arabic AI answers?

Yes for the major commercial engines, which support Arabic prompts across Egypt and the Gulf. It does not track the state-backed Arabic models, Falcon-H1 Arabic in the UAE, ALLaM in Saudi Arabia, Jais and Fanar, because no commercial tracker on the market currently monitors any of them. Visibility inside those models is measured manually on a fixed prompt set at a fixed cadence.

Are there other AI visibility dashboards in the Middle East?

Yes. ESOTERIC in Dubai has built Apparens, an AI visibility product with its own scoring model, and regional SaaS products exist including Zaher.AI in Cairo, which is Arabic-first. What we have not found is another agency in the Gulf or Egypt where AI visibility measurement and the delivery of the work, roadmap, tasks, content approvals, reports and invoices, are the same product behind one client login.

Related guides

• How to Measure AI Visibility (Share of Model): A 2026 Playbook
• Best AEO Tools in 2026: Verified Prices, and Which Ones Actually Work in Arabic
• What Does a GEO Agency Do? The Complete 2026 Guide
• Arabic GEO: How to Get Your Arabic Content Cited by ChatGPT, Gemini and AI Overviews
• Best GEO Agencies in the Middle East & Gulf (2026)