Asked the same question twice, the models moved their own first choice 67% of the time. The leader still held in 27 of 30 readings (10 categories at three buyer sizes).
Ten categories were asked again three days after the edition run, with nothing changed: the same prompts, the same model versions, the same settings, one buyer segment, six framings, fourteen models, 840 answers. Whatever differs between the two runs is noise, and the noise is what this page measures. That is the reason the index asks every question six ways of every model rather than once: on its own, one ask moved 67% of the time; eighty-four, read together, are what the index reports.
Model by model
Mistral Small changed its first choice most often, 89% of its 38 pairs; GPT-6 Luna least, 46% of 37. Pooled over every model and question, 67%.
A flip is the same model, the same question, a different first choice a few days apart. It says how much one answer can be trusted on its own, and nothing about why the model answered as it did.
Category by category
| Category | Leader in the edition run | Share, run one | Share, repeat | Move | Leader |
|---|---|---|---|---|---|
| Cap table management · Mid-market |  Carta | 33% | 50% | +17 points | held |
| Entity management · Enterprise |  Diligent Entities | 50% | 35% | -15 points | held |
| Sales tax automation · Mid-market |  Avalara AvaTax | 32% | 45% | +13 points | held |
| Financial close manageme · Enterprise |  BlackLine | 67% | 78% | +10 points | held |
| SaaS metrics · Small business |  ChartMogul | 53% | 61% | +8 points | held |
| Business banking · Small business |  Mercury | 35% | 43% | +7 points | held |
| Account reconciliation · Mid-market |  FloQast | 33% | 26% | -7 points | held |
| Accounts payable automat · Enterprise |  Coupa | 20% | 27% | +7 points | held |
| SaaS metrics · Mid-market | ChartMogul | 49% | 55% | +6 points | held |
| Board · Mid-market |  OnBoard | 26% | 30% | +4 points | held |
| Financial close manageme · Mid-market | FloQast | 55% | 59% | +4 points | held |
| Sales tax automation · Enterprise |  Vertex | 48% | 52% | +4 points | held |
| Accounts receivable auto · Enterprise |  HighRadius | 78% | 74% | -3 points | held |
| Cap table management · Small business |  Eqvista | 29% | 25% | -3 points | changed: Pulley |
| Accounts receivable auto · Small business |  BILL Accounts Receivable | 17% | 20% | +3 points | held |
| Accounts payable automat · Small business |  BILL Accounts Payable | 41% | 44% | +3 points | held |
| Account reconciliation · Small business |  QuickBooks Online | 49% | 46% | -3 points | held |
| Cap table management · Enterprise |  Shareworks | 38% | 40% | +3 points | held |
| Board · Enterprise |  Diligent Boards | 30% | 32% | +2 points | held |
| SaaS metrics · Enterprise | ChartMogul | 25% | 26% | +1 point | held |
| Entity management · Small business |  EntityKeeper | 23% | 25% | +1 point | held |
| Business banking · Mid-market | Mercury | 23% | 22% | -1 points | held |
| Sales tax automation · Small business |  TaxJar | 33% | 34% | +1 point | held |
| Business banking · Enterprise |  Rho | 19% | 18% | -1 points | held |
| Accounts payable automat · Mid-market |  Stampli | 22% | 21% | -1 points | changed: Ramp |
| Account reconciliation · Enterprise | BlackLine | 63% | 63% | 0 points | held |
| Financial close manageme · Small business |  Numeric | 25% | 25% | 0 points | changed: FloQast |
| Board · Small business |  BoardPro | 29% | 29% | 0 points | held |
| Accounts receivable auto · Mid-market |  Upflow | 17% | 17% | +0 points | held |
| Entity management · Mid-market |  Athennian | 41% | 41% | +0 points | held |
In 27 of the 30 readings (10 categories at three buyer sizes) the same product led both runs. Where the leader changed, the two products were within 1 point of each other in the edition run.
The share floor
From the next edition on, a product's change in share counts as movement only when it is larger than 10 points, and a new leader is reported only when it clears the old one by more than that. Nine repeats in ten move a leader less. The floor is measured again with every edition and the method page carries the rule: how the floor is measured.
Cite this
Finance AI Recommendation Index, October 2026 Edition: repeat measurement. finance-ai-index.com/research/repeat-measurement/. Published under CC BY 4.0. Every figure on this page is computed from the published edition and changes with it; the edition and its date are the citation.
The output is the models' output. Nothing here says the models can be steered, and nothing here is a recommendation by the index.