# Does a more precise question give you a different brand recommendation? _We asked 11 AI models for Airtable agencies, then for Airtable implementation agencies. The answers were practically the same. What stood out instead: with so little written about the category, the models named 242 different brands, and the leaders stand out through their own content, not the Airtable partner directory._ Field Notes 03 · 2 Oct 2026 · 3 min read · https://www.brandranking.ai/field-notes/does-a-more-precise-question-give-a-different-brand-recommendation --- ## The short answer 1. **No.** "Airtable agency" and "Airtable implementation agency" got the same top four, in the same order, with scores within four points. 2. **The models scatter.** 1,080 picks went to 242 different brands, 163 of them named only once. A CRM question gets 9. 3. **The partner directory gets you listed, not ranked.** It's cited as often for brands named once as for the leaders. Each leader has its own channel: YouTube for GAP Consulting, the Airtable community for On2Air. _The test_ ## Two questions, one word apart We asked 11 models each question 10 times, answering from training with no web search: - "Which Airtable agency would you recommend?" - "Which Airtable **implementation** agency would you recommend?" **Brand score, both questions** | Brand | Agency | Implementation | | - | - | - | | GAP Consulting | 42.9 | 41.5 | | Openside | 33.5 | 29.6 | | On2Air | 30.7 | 29.1 | | XRay.Tech | 17.3 | 17.5 | _Brand score: 100 for rank 1 down to 20 for rank 5, 0 when not named, averaged over the 11 models._ The extra word didn't even clear out the software. About one pick in five went to tools like Softr and Stacker in both studies (20% and 19%). [Open the agency study](https://www.brandranking.ai/category-ranking/airtable-agencies) --- _Finding_ ## Thin data, lots of brands Few people write comparisons of Airtable agencies, so a model has little to go on, and it fills the gap with whatever it can recall. The top four take a third of all picks. The rest is spread so thin that most names appear once. **1,080 picks, both questions** | Name | Share | | - | - | | GAP Consulting, Openside, On2Air, XRay.Tech | 32.6% | | 32 brands named 6+ times | 41.8% | | 43 brands named 2–5 times | 10.6% | | 163 brands named once | 15.1% | _Only 53 of the 242 brands appear in answers to both questions._ The models say as much when they get to the smaller names. qwen3.8-max ranks one fifth because "its public footprint feels more boutique"; claude-haiku-4.5 describes another as having "a smaller market presence compared to top-tier agencies". Compared with categories where the internet has plenty of rankings, it is an outlier: **Different brands named in one study** | Name | Share | | - | - | | Airtable agencies | 153 | | Airtable implementation agencies | 142 | | MBA programs | 39 | | Business podcasts | 31 | | No-code platforms | 23 | | UK neobanks | 18 | | EVs | 14 | | CRM software | 9 | _Each study has 540 to 800 picks, counted the same way._ --- _Finding_ ## The partner directory isn't what makes a leader Every pick comes with up to three sources the model says it relied on. The Airtable partner directory is the most cited one, but it's cited just as often for brands named once as for the top four. It's the generic reason any firm gets in, not what puts a firm on top. **Picks citing each source** | Source | Top 4 | Others | Named once | | - | - | - | - | | Airtable partner directory | 44% | 31% | 47% | | Own website | 79% | 73% | 52% | | Airtable community | 33% | 20% | 17% | | YouTube | 27% | 4% | 1% | | Clutch | 20% | 15% | 12% | | LinkedIn | 5% | 14% | 22% | _Share of a group's picks that cite at least one source of that kind. Sources are recalled by the model, not verified._ What sets each leader apart is a second channel of its own, one the rest barely have: **What each leader is cited for** | Brand | Stand-out source | Cited | Others | | - | - | - | - | | GAP Consulting | YouTube | 65% | 4% | | On2Air | Airtable community | 61% | 20% | | Openside | Partner directory | 72% | 31% | | XRay.Tech | Own website | 93% | 73% | Openside is the one leader the directory does carry, with Clutch reviews (35%) behind it. But directory citations are spread across 127 different brands. A firm gets in by being listed; it gets to the top through content the model can remember it by. The reasons the models give line up with those sources: **What the models say about each leader** | Brand | Reason | Leader | Others | | - | - | - | - | | GAP Consulting | Teaching content | 86% | 9% | | On2Air | Airtable apps | 76% | 36% | | Openside | Certified partner | 43% | 15% | | XRay.Tech | Integrations | 91% | 47% | _Share of the brand's picks whose reasoning mentions it, against all brands outside the top four. Teaching content: tutorials, YouTube, courses, training. Airtable apps: extensions, add-ons, backups._ In their own words: - **GAP Consulting** has "extensive tutorials, courses, and case studies that demonstrate deep platform expertise" (kimi-k3) and an "extensive library of tutorials" (grok-4.6). - **On2Air** is "best known for Airtable backup and integration apps, but the team also delivers consulting" (claude-opus-5). gemini-3.8-flash ranks it fifth "because their primary focus is pre-built SaaS extensions rather than full-scale custom agency builds". - **Openside** is "one of the most established Airtable Services Partners" (claude-opus-5), and "the team behind the BuiltOnAir podcast and the On2Air app suite" (kimi-k3). - **XRay.Tech** is "a strong choice if Airtable is part of a broader operations stack" (kimi-k3), but "less exclusively known for Airtable implementation than the other four" (deepseek-v4.1-flash). > **gpt-6-astra, ranking GAP Consulting first** > > "Its extensive educational material makes its approach relatively easy to > evaluate before committing." --- ## How we measured - **Questions.** The two above, run on 24 September 2026, 10 times per model each, 220 successful runs. - **Models.** claude-opus-5, claude-haiku-4.5, gpt-6-astra, gpt-5.6-luna, gemini-3.8-flash, gemini-3.5-flash-lite, grok-4.6, deepseek-v4.1-flash, kimi-k3, qwen3.8-max and mistral-medium-3.5. - **Brands.** Spellings of the same company are merged before counting. Other categories are counted the same way. - **Sources.** Grouped by site, with and without a URL: "Airtable Partner Directory" and airtable.com/partners count as one. They are what the model says it relied on, not a verified citation. _BrandRanking.AI © 2026 · Study data: [brandranking.ai/categories](https://www.brandranking.ai/categories)_