Last updated: 11 August 2026
General assistants now dominate the questions finance teams ask about modelling, checking spreadsheets and running deal work. The specialist tools still have room to win, but product quality alone does not decide which names appear.
Methodology
We track 55 buyer questions across ChatGPT, Gemini and Google AI Overview. The figures below draw on 405 answers collected on 11 August 2026.
We record what the engine returns, not what it is asked. No brand gets credit for appearing in a question.
Named in answers shows the share of answers that mention a brand. Share of the answer shows the brand’s share of all measured brand mentions. Average position shows where the brand appears when it is named, with a lower number meaning a higher place.
The 2026 leaderboard
| Rank | Brand | Named in answers | Share of the answer | Average position |
|---|---|---|---|---|
| 1 | ChatGPT | 35.6% | 33.4% | 2.9 |
| 2 | Claude | 29.9% | 21.9% | 3.6 |
| 3 | Microsoft Copilot | 17.3% | 9.3% | 3.0 |
| 4 | Hebbia | 11.6% | 10.2% | 2.7 |
| 5 | Daloopa | 9.1% | 7.0% | 3.1 |
| 6 | Rogo | 8.4% | 7.2% | 2.1 |
| 7 | Shortcut | 4.4% | 1.9% | 3.6 |
| 8 | Macabacus | 3.5% | 2.5% | 3.1 |
| 9 | Endex | 3.2% | 2.1% | 1.5 |
| 10 | Numerous | 2.7% | 1.0% | 4.0 |
| 11 | TraceLight | 2.0% | 1.5% | 1.3 |
| 12 | Rockhopper | 1.7% | 1.0% | 3.4 |
| 13 | Formula Audit | 1.0% | 0.7% | 1.7 |
| 14 | Intrinsic | 0.5% | 0.1% | 5.5 |
| 15 | BlueFlame AI | 0.3% | 0.3% | 4.0 |
General assistants own 64% of the answer
ChatGPT, Claude and Microsoft Copilot take 64.5% of all measured brand mentions. Together, they appear far more often than every specialist tool combined.
That does not mean buyers prefer a general assistant for every finance task. It means those brands enter the answer first because they already own the language around Excel, agents, model building and data analysis.
Specialist tools split the remaining market
Hebbia leads the specialist set, appearing in 11.6% of answers and taking 10.2% of all measured mentions. Daloopa and Rogo form the next group. No other specialist tool appears in more than 5% of answers.
The long tail is not flat. Endex and TraceLight both reach a strong position once named, at 1.5 and 1.3 respectively, but neither appears often enough to turn that strength into category share.
Position and presence tell different stories
TraceLight has the best average position in the table at 1.3. It also ranks 11th for how often it is named.
This pattern matters. A tool can fit the question well when the engine already knows it, yet lose most shortlists because too few trusted pages connect the brand to the category in the first place.
Video and practitioner discussion shape the shortlist
YouTube appears in 21.2% of measured answers, LinkedIn in 15.1%, and Reddit in 12.6%.
The pages that travel furthest are direct comparisons, worked examples and tests on the same model. Product tours have less value than evidence a finance buyer can inspect and repeat.
ChatGPT advertising has reached the category
We saw 51 advertisers across 40 of the tracked questions. The leading advertiser held 15% of measured ad exposure, followed by Workday at 8% and CloudZero at 6%.
The current advertising field comes from adjacent software, consulting and finance providers. The specialist modelling brands in this table have not yet bought a measured placement. Organic recommendation still decides their category position.
What changes the next index
The category leaders publish across four layers:
- Product pages that name the exact finance workflow.
- Comparison and how-to pages that answer a buyer’s question before making the sale.
- Independent coverage from finance publications and creators.
- Worked examples in the practitioner communities engines already retrieve.
The next winners will connect benchmark proof to all four. A stronger model matters. A stronger body of public evidence makes sure the model gets considered.
If you want the same category view, book 20 minutes with Marco.
