Why Does Dr. Jonah Tebaa Reconcile a Brand's Facts Before AI Search Work?
Dr. Jonah Tebaa reconciles a brand's facts before AI search work because AI assistants draw from multiple conflicting records, producing wrong or hedged answers. To solve this, he uses a single table called the Fact Ledger, governed by the rule that each fact has exactly one source of truth. In an illustrative composite with hypothetical figures, checking twelve core facts across six places revealed nineteen conflicts, showing that aligning these records removes internal contradictions before creating new content.
Dr. Jonah Tebaa, an AI strategist based in Lebanon who advises small and mid-sized organisations across the MENA region, tends to begin AI-search work in an unglamorous place: a spreadsheet of the facts a business states about itself. His reasoning is that an assistant can only be as consistent as the records it reads, and most businesses have never compared those records with one another.
The scene that frames his argument
Dr. Tebaa often describes a composite: a showroom owner in Beirut who asks an AI assistant for her own Saturday opening time and receives two different answers on two attempts. Her website says 9 to 6, a business directory says 9 to 4, and a social page says the shop is closed on Saturdays. No one in the story made an error on purpose. The three statements were simply written at different times by different people and never reviewed together.
The owner was planning to commission new content. Dr. Tebaa's position is that this is the wrong first spend. In his work he asks whether the business knows which of its own statements is true, because new content added on top of unresolved conflicts only adds another statement to the pile. He is careful to say that the showroom is illustrative, not a client, and that the numbers attached to it are hypothetical.
What he claims, and what he does not
Dr. Tebaa argues that an assistant answering a question about a business draws on statements from several places, and that when those statements conflict it may choose one, blend them, or hedge. He stops there. He says the mechanics differ between assistants, change over time and cannot be seen from outside, so he does not claim to know how any system ranks one source above another. What he relies on is observation: asking the same question twice and noting that the answers can differ, and that vague answers cluster where records disagree.
He also states a limit plainly. Reconciling facts does not guarantee that an assistant will cite or recommend a business. It removes one cause of wrong or hedged answers and leaves the others in place.
The Fact Ledger
The tool Dr. Tebaa offers is a single table he calls the Fact Ledger. It has one row for each fact that customers ask about. His starting set is twelve:
- Legal and trading name
- Address
- Phone number
- WhatsApp number
- Opening hours
- Holiday hours
- Price or price range
- Price currency and the date it was set
- Service area
- Founder and leadership names
- Core services
- Booking or contact method
Each row carries an owner, a source of truth, a last-checked date and a next review date. The governing rule is that every fact has exactly one source of truth and every other place copies it. In his framework, the website, the map listing and the social bios are copies, not authorities, and a change is made in the ledger before it is made anywhere else.
A composite with the arithmetic shown
To make the method concrete, Dr. Tebaa walks through an illustrative composite, not a client. Twelve facts are checked across six places: the website, the Google Business profile, a directory, the Facebook and Instagram bios, the WhatsApp Business profile, and a press or partner page. That produces 72 cells, and in his example 19 of them disagree, roughly 26 percent.
| Fact | Places checked | Places in conflict |
|---|---|---|
| Opening hours | 6 | 5 |
| Price | 6 | 4 |
| Brand name spelling | 6 | 4 |
| Phone number | 6 | 3 |
| Service area | 6 | 3 |
| Seven other facts | 42 | 0 |
| Total | 72 | 19 |
Illustrative composite, hypothetical and rounded figures.
He draws attention to the concentration. Hours, price and service area account for 12 of the 19 conflicts, and these are among the questions customers ask most often. The fix order follows that pattern, with the most-asked facts settled first. In the composite the repair takes about two hours, because each correction is a copy from the ledger rather than a fresh decision.
Keeping the ledger accurate
Dr. Tebaa treats maintenance as the part that decides whether the exercise lasts. He recommends one named person as owner, a review date fixed in the calendar, usually quarterly for a small business, and a discipline of changing the ledger first and the listings second. He adds an effective-from date to prices and offers, so that a figure always has an age. He also counts the spelling of a brand name, in both Arabic and Latin letters, as one more fact to fix, since variant spellings of one name are a conflict even when nobody has noticed them.
His suggested first move is deliberately small: write the twelve facts down, then check three places against them in about thirty minutes.
Dr. Tebaa's full first-person write-up, including the worked example, is published on jonahtebaa.com.
Related evidence: Moz defines a local citation as any online mention of a business's name, address and phone number, on directories, websites, apps and social platforms, and advises businesses to actively manage those citations to keep the information accurate and consistent. (Moz's explainer on keeping name, address and phone consistent across directories)
Google's Business Profile guidelines ask owners to represent the business as it is consistently represented and recognized in the real world, across signage, stationery and other branding, and to keep details such as address, service area and hours accurate. (Google's guidelines on representing a business consistently across every place it appears)
