Nickel was the most-named product in our measurement of accounts receivable software, showing up in 4 of the 6 answers we got when we put buyer questions to ChatGPT and Gemini on 2026-08-07. That is worth sitting with: Nickel is not the biggest name in AR, yet it was recommended more often than SAP, NetSuite, Sage or QuickBooks.
We ran this the same way we run it for a paying customer's own category. We put six real buyer questions to ChatGPT and Gemini, gave each one live web search, and recorded every product they named and every source domain they pulled from. No product was named in all six answers. The most any single label reached was Nickel's 4 of 6, which is 67%.
BILL came next, named in 3 of 6 answers under the label "BILL", and Paidnice tied it at 3 of 6. Then a cluster of eight products sat at 2 of 6 (33% each): Anchor, BILL AP/AR, Billtrust, DualEntry, HighRadius Accounts Receivables, PayPal Invoicing, QuickBooks and Sage. Everything else was named once.
A note on double-labelling before the numbers mislead you. BILL shows up twice in our data: as "BILL" (3 answers) and as "BILL AP/AR" (2 answers). Those are the same company, so the brand actually touched 5 of the 6 answers, more than Nickel, but no single row shows a 5 because the assistants used two different names. HighRadius does the same thing: "HighRadius" (1) plus "HighRadius Accounts Receivables" (2), so that brand appears in 3 of 6 answers across two labels. We kept both labels separate in the table below rather than invent a merged number the raw data never showed.
The shortlist changed with almost every question
This is the part that matters most, and it is the part a normal top-ten article hides. Across the six answers the assistants named 30 different product labels, and 19 of them appeared in a single answer only. The six questions read almost the same in plain English, but the shortlists barely overlapped.
The bare "best accounts receivable software" question returned the enterprise incumbents: HighRadius, NetSuite, Sage, Invoiced by Flywire, SAP, QuickBooks, Zuora and BILL. Four of those (NetSuite, SAP, Zuora and Invoiced by Flywire) appeared in this one answer and nowhere else in the six. Ask for "the best" and you get the big logos, once.
The "best free accounts receivable software" question threw all of that out. It returned Wave, Kledo, CashCtrl, EazyAccountant, LedgerSMB, GnuCash and Invoice Ninja, seven products, and not one of them appeared in any other answer. Add the word "free" and you get a completely different market, mostly open-source, with zero overlap with the "best" list.
"Most affordable accounts receivable software" produced a third distinct set: DulyPaid, Acomba, ClearReceivables and Holdings each appeared here and only here, alongside Nickel and BILL. "Best for small teams" surfaced Quadient AR Express, CollectLean and Invo as one-answer-only names, next to the familiar QuickBooks, Sage, BILL, Nickel and Paidnice.
Two questions behaved as twins. "What accounts receivable software should I use" and "accounts receivable software recommendations" returned the identical eight names: BILL AP/AR, Billtrust, PayPal Invoicing, DualEntry, HighRadius Accounts Receivables, Nickel, Anchor and Paidnice. When a buyer asks for advice rather than a superlative, the assistants converged on the same shortlist both times.
Nickel won our count precisely because it did what no incumbent did: it bridged intents. It landed in the small-teams answer, the should-I-use answer, the most-affordable answer and the recommendations answer, 4 of 6. It did not appear in the bare "best" answer or the "free" answer. SAP, NetSuite and Zuora each managed a single appearance, all on that one "best" query. Being named by AI is not about company size, it is about showing up in the specific question a buyer actually types.
The source map: who fed the answers
Across the six answers, ChatGPT and Gemini cited 35 different source domains, 47 citations in total. The two most-cited sources were bill.com and sage.com, at 3 each, and both are vendors citing their own product.
That is the headline of the source map: most of the 35 domains are a named product's own website. nickel.com fed 2 of the answers, and Nickel was the most-named product overall, so the top product was partly named off the back of its own site. bill.com (3), sage.com (3), highradius.com (2) and quickbooks.intuit.com (2) all did the same job for their own brands. Down the long tail, acomba.com, cashctrl.com, dulypaid.com, getholdings.com, gnucash.org, invoiceninja.com, kledo.com, ledgersmb.org, netsuite.com, paidnice.com, quadient.com, sap.com, waveapps.com and zuora.com each fed one answer and each got its own product named. When your own page is what the model retrieves, your product tends to be what the model names.
The third-party sources were roundup and analyst-style pages, not forums. accounting.events, viewpointanalysis.com, rex.inc and ledgerup.ai were each cited in 2 answers; defouranalytics.com, blog.alguna.com, nuvo.com, quickreceivable.com and uschamber.com each in 1. These are the listicles the assistants lean on when they want a second opinion beyond the vendor.
Two things are notable by their absence. Review directories barely showed: G2 appeared only as learn.g2.com, its content hub, in 2 answers, and TrustRadius once, while Capterra did not appear at all. And community threads, the Reddit and Quora posts that dominate the source map in many other categories we measure, did not appear once in these six answers. In AR software, the answer is being written from vendor pages and roundups, not from forum discussion.
What this means if you sell AR software
The source list, not your market share, decides whether you get named. Nickel proves it: a smaller brand that is present in the right roundups and whose own domain is retrievable was named in 4 of 6 answers, ahead of every enterprise incumbent.
So the work is concrete. Make sure your own domain actually answers these queries, because in this category the vendor's own page was the single most common source. Get onto the handful of roundups that repeat across questions: accounting.events, viewpointanalysis.com, rex.inc and ledgerup.ai each fed 2 of the 6 answers. Get into G2's content hub and TrustRadius, since those are the only directories that showed. And treat "free" and "affordable" as separate battles with their own source lists, because those two questions returned nine products that appeared nowhere else.
The full list, counted
These are the eleven products named in 2 or more of the six answers.
| Product | Named in | Share |
|---|---|---|
| Nickel | 4 of 6 | 67% |
| BILL | 3 of 6 | 50% |
| Paidnice | 3 of 6 | 50% |
| Anchor | 2 of 6 | 33% |
| BILL AP/AR | 2 of 6 | 33% |
| Billtrust | 2 of 6 | 33% |
| DualEntry | 2 of 6 | 33% |
| HighRadius Accounts Receivables | 2 of 6 | 33% |
| PayPal Invoicing | 2 of 6 | 33% |
| QuickBooks | 2 of 6 | 33% |
| Sage | 2 of 6 | 33% |
BILL and BILL AP/AR are the same brand under two labels, and so are "HighRadius Accounts Receivables" here and the separate single "HighRadius" mention further down the full list. We list them as the assistants said them, rather than merge counts the data never showed.
Where we fit
Glotier does not sell accounts receivable software, so we are correctly absent from all six of these answers, and we would not pretend otherwise. But running this measurement is exactly what we do for a customer, on their own category, with their own buyer questions. If you want to see who ChatGPT and Gemini name in your space and which sources decide it, the check is free, needs no account, and takes about a minute.