71/100
This shows the overall SEO and technical health of the 24 reviewed pages.
01 · Starting point
A clear view
of the sample.
The scores summarize what was measured on 24 selected pages. They help prioritize the work; they are not grades for every URL or forecasts of traffic and sales.
64/100
This shows how ready the site's information is to be understood and cited by AI answer engines—not how often the brand currently appears.
24 / 393
This was a strategic sample, not a review of the entire site. A repeated template problem may affect more pages, but a full crawl is required before treating it as sitewide.
✓What is working
- All 24 reviewed pages responded.
- Most sampled content was available in the initial HTML.
- Every reviewed page had a unique title and canonical tag.
- The XML sitemap was valid and the tested crawler profiles were not blocked.
!What needs attention
- Page clarity: headings and search summaries are missing on many reviewed pages.
- Business data: some structured data is unreadable and the restaurant entity is incomplete.
- Accessibility: image uses need individual review and the correct text treatment.
- AI clarity: important facts need shorter, self-contained answers.
02 · Priority opportunities
What we found.
Why it matters.
Four patterns explain most of the opportunity in the reviewed sample. Each one can be translated into a practical task.
Pages need clearer labels
16/ 24 without one clear H1
Eighteen of the 24 pages also lacked a brand-written search summary.
Simple example: the menu page should clearly introduce itself as Tahini's menu and briefly explain what customers can find there. Large decorative text is not always the same as a real page heading.
Search and AI business data needs repair
6/ 31 structured-data blocks unreadable
The initial HTML also lacked a complete Restaurant and Location description.
Simple example: this information works like a digital business card. The brand, each location, address, hours and telephone should be connected clearly instead of making a search system assemble the facts itself.
Images need an accessibility review
534/ 568 image uses lacked an alt attribute
This is an inventory signal, not 534 confirmed unique images that all need written descriptions.
Simple example: a food photo may need “Chicken shawarma wrap with garlic sauce and pickles.” A decorative background should be marked as decorative. Each use needs review; the same text should not be copied everywhere.
Important facts need answer-ready wording
Cleareranswers needed
Reviewed copy did not consistently state the main fact first or work as a short, self-contained answer.
Simple example: “Tahini's is a Mediterranean fast-casual restaurant serving shawarma, bowls and catering” explains the business more clearly than a slogan alone. This helps people and AI tools understand the page; it does not guarantee a citation.
03 · Action plan
A practical
three-phase path.
Start with the clearest repairs, add approved business information, then validate the finished work and measure again.
1
Fix the clearest technical issues
Add one clear heading and a useful search summary to priority pages, then repair the structured-data blocks that could not be read.
Example: give the menu page a clear “Tahini's Menu” heading and summary, repair its hidden data and test the final page again.
2
Strengthen restaurant, location and content signals
Connect approved facts to the correct locations, improve key answers, publish the prepared AI guidance files and review important images for accessibility.
Example: connect the London location to its approved address, telephone and hours, then publish the prepared llms.txt files as a simple guide to Tahini's official pages.
3
Validate and measure
Recheck each agreed change, document what passed and repeat the same AI visibility test after 60 days.
Example: compare the same HubSpot test before and after implementation. Movement shows a change in that tool; it does not by itself prove sales impact.
04 · AI visibility baseline
How Tahini's appears
in AI search today.
HubSpot's AEO Grader compared how OpenAI, Perplexity and Gemini represented Tahini's. The result is a useful snapshot: the brand is understood, but visibility and market presence vary considerably by provider.
57.7%
Simple combined view
The three-provider average is about 58 out of 100.
Calculation: (49 + 68 + 56) ÷ 3 = 57.7. This equal-weight average makes the three results easier to summarize in conversation. It is our presentation summary, not an additional official score issued by HubSpot.
AI technical accessOne door is open.
One door is open.
Two guide files are missing.
This check is separate from the HubSpot score. It shows whether AI citation crawlers can enter the site and whether Tahini's gives them a simple map of its official information.
robots.txt allows the tested citation crawlers
The site has a robots file, and the audit found 0 of the 6 tested citation crawlers blocked.
Example: the front door is open. OpenAI, Claude and Perplexity are not being told to stay out.
llms.txt and llms-full.txt are not published
Neither guide file was available at the site root when tested. Prepared versions can point to official menu, location, catering and brand pages.
Example: the door is open, but there is no welcome guide showing AI systems which Tahini's information is official. Adding the files is useful technical hygiene, not a guarantee of rankings or citations.
Recommended action:
Keep crawler access open. Review and publish the prepared files at /llms.txt and /llms-full.txt, then confirm that both URLs load correctly.
OpenAI
49/100
Developing visibility. The report recognized the brand, but found limited market presence and share of voice.
Recognition7/20
Market score3/10
Presence9/20
Sentiment29/40
Share of voice1/10
Perplexity
68/100
The strongest overall result. Recognition and presence were higher here than in the other two providers.
Recognition14/20
Market score8/10
Presence15/20
Sentiment31/40
Share of voice0/10
Gemini
56/100
Developing visibility. Sentiment was strong, while recognition and market presence remained limited.
Recognition9/20
Market score3/10
Presence10/20
Sentiment33/40
Share of voice1/10
What AI already understands
The report consistently associates Tahini's with Mediterranean fast-casual food, shawarma, bowls, catering, fusion flavours and North American expansion.
The strongest signal: positive perception
Brand sentiment scored between 29 and 33 out of 40 across the three providers. When the brand appears, it is generally described positively.
The clearest gap: share of voice
Share of voice scored 1/10, 0/10 and 1/10. In simple terms, Tahini's is described reasonably well when found, but it is not appearing often enough across the market questions tested.
Why providers disagree
OpenAI, Perplexity and Gemini use different sources and knowledge windows. A strong result in one provider does not guarantee the same representation in another.
AI competitive picture
Which brands appear beside Tahini's?
HubSpot produced a separate market-position chart for each AI provider. The percentages describe that provider's test results, not restaurant sales or real-world market share, so each column should be read independently.
12%Tahini's slice
in this chart
in this chart
Examples shown nearby: Pita Pit 16%, Freshii 14% and CAVA 13%.
2.5%Tahini's slice
in this chart
in this chart
Examples shown nearby: Al Fresco 14.5%, Cafe Alondra 12% and Zaytouna 9%.
12.5%Tahini's slice
in this chart
in this chart
Examples shown nearby: Osmow's 22%, CAVA 18.5% and Paramount 14%.
ChatGPTMarket score: 3/10
PerplexityMarket score: 8/10
GeminiMarket score: 3/10
