Zero of fourteen models named Census first on the direct prompt; zero named Airbyte. Both were named by all fourteen models and Census carries 83 labels and Airbyte 41, so the shares are not directly comparable.
By Fivetran, United States. Named in five categories this edition.
Named in two categories this edition.
Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all fourteen models, for a mid-market B2B company; rank is within the category; every quote names the model and the prompt it came from. Both figures come from the Data warehouse and reverse ETL for marketing page.
Across every category in the November 2026 Edition, Census and Airbyte were named in the same answer 107 times, of the 291 answers naming Census and the 126 naming Airbyte. In those answers Airbyte took the first choice six times and Census fifteen.
| Model | Direct | Paraphrase | Comparative | Budget-constrained | Scale-constrained | Negative |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | ||||||
| GPT-5.4 mini | ||||||
| Gemini 3.5 Flash | ||||||
| Perplexity Sonar | ||||||
| Grok 4.20 | ||||||
| Mistral Small | ||||||
| DeepSeek V4 Flash | ||||||
| Llama 4 Maverick | ||||||
| Qwen 3.7 Flash | ||||||
| Kimi K2 | ||||||
| GLM 4.7 FlashX | ||||||
| MiniMax M2.5 | ||||||
| GPT-6 Luna | ||||||
| Muse Glimmer 30B |
Bold names in an answer are the products the judge labeled a first choice; a model naming several gives each of them that label. The full answer text for every row is in the record.
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“avoid Census post-acquisition pricing” DeepSeek V4 Flash · negative prompt · hard negative
“Census has been integrated into Fivetran's activation offering, so ask about the current product, roadmap, support, and contract terms rather than evaluating it as an independent vendor.” GPT-6 Luna · scale prompt · soft negative
“User-friendly but limited depth for complex marketing logic; post-acquisition changes ... have raised concerns about feature bloat or pricing shifts” Grok 4.20 · negative prompt · soft negative
“Commercial platforms like Census and Hightouch deliver faster time-to-value for most SaaS teams in North America.” Claude Haiku 4.5 · scale prompt · first choice
“Shortlist Hightouch and Fivetran Activations (the product associated with Census)” GPT-6 Luna · scale prompt · first choice
“Choose a tool that offers a self-service interface (like Hightouch or Census)” GLM 4.7 FlashX · negative prompt · first choice
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“Opting for self-hosted tools (like the self-hosted version of Airbyte or Castled) thinking it saves money... Self-hosting Reverse ETL is a massive engineering undertaking” Qwen 3.7 Flash · negative prompt · soft negative
“their reverse ETL capabilities are still developing and may not have the performance or reliability needed” Mistral Small · negative prompt · soft negative
“Avoid building your own (e.g., custom Airbyte + dbt + cron) unless your team is engineering-heavy” Grok 4.20 · paraphrase prompt · soft negative
“ELT / Ingestion | Airbyte (open-source) or Meltano | Open-source ELT with 300+ connectors” Grok 4.20 · budget prompt · first choice
“Airbyte (open-source, self-hosted) is a top pick for budget-conscious teams.” Mistral Small · budget prompt · first choice
“ELT/ETL (Data Ingestion): Airbyte — *$0–$50/month* ... Completely free” DeepSeek V4 Flash · budget prompt · first choice
Comparisons are drawn for the top eight products in each category, each against each. The output is the models' output; nothing here is a recommendation by the index.