GEO for B2B software and FinTech
A buyer asks ChatGPT which ERP fits a machine builder. The answer names a few vendors, the rest do not appear in that research. For software vendors and FinTechs we measure which buyer questions they show up for in ChatGPT, Perplexity, Google AI Overviews and Claude, and work on making it more.
Own measurement
For two AI rankings we measured 15 vendors each with 10 buyer questions in ChatGPT, Perplexity, Google AI Overviews and Claude, three runs each. That is 120 answers per market. LegalTech on 25 September 2026, ERP for the German mid-market on 27 and 29 September 2026. Both rankings are in German.
Asked for an ERP for machine builders with one-off and small-batch production, all 12 answers name Proalpha and abas, but only 5 name APplus and 4 name ams.erp. In the LegalTech ranking, 8 of Lawlift's 12 mentions come from a single question, the one about drafting contracts from templates.
Perplexity names weclapp in 24 of 30 ERP answers. Actaport appears in 12 of 30 LegalTech answers in Google AI Overviews, more often than in any other system.
C.H.Beck is named in 31 of 120 LegalTech answers but explicitly recommended in only 5.
Outside our measured field, Odoo is named in 61 of 120 ERP answers, more often than twelve of the 15 vendors in the field. Your competitor analysis does not decide whom AI names as an alternative.
What follows: “we are visible in ChatGPT” is too coarse a statement. The useful question is which buyer questions you are named for, in which system, and who appears instead of you.
Buyer questions
No brand names, their own problem in the sentence. Four of the ten questions from our ERP ranking, translated from German:
For your measurement we write the questions for your market and your target customers, not from a template.
Reference
Goodson Softwaresolutions, service questions
14.2 % to 22.5 %
AI visibility
mediaSolution3, product questions
6.7 % to 23.3 %
AI visibility
Own measurement: 10 buyer questions each in ChatGPT, Perplexity, Google AI Overviews and Claude, three runs each. Before on 10 July, after on 12 and 13 August 2026. One engine went down for each entity, so this is not proof of causation.
Video, transcript and all figuresProcess
01
Free. Three buyer questions from your market, three runs each in ChatGPT, Perplexity and Google AI Overviews. Result as a one-pager by email: where you are named and who is named instead.
02
Diagnosis, strategy and implementation based on a measurement in four AI systems, followed by four weeks of accompanying monitoring. Then you decide whether to continue.
03
Monthly citation tracking, ongoing schema and content work, quarterly reviews.
FAQ
Software is bought through comparison: which system fits our case, what does implementation cost, which alternatives exist. These are exactly the questions we ask in our measurements. In both of our rankings, which vendors are named depends heavily on the individual question. A vendor can be in every answer to one question and missing from the next.
There is no published FinTech case on our website yet. Our published reference comes from a software company. The method is the same for FinTech: buyer questions from your market, measured in four AI systems, before and after. The free mini-audit shows you where you stand today.
Per question and per system: whether you are named, whether you are explicitly recommended and which competitors appear instead. Every question runs three times because AI answers vary from run to run. Method and date are stated in every result.
The mini-audit is free. GEO setup and GEO retainer are listed with prices on our services page.
Not automatically. According to Ahrefs (April 2026, 863,000 keywords) only 38 % of citations in Google AI Overviews come from the organic top 10 results. Only a measurement shows whether AI systems name you for your buyer questions.
Send us your website and your industry. We measure three buyer questions and email you the result as a one-pager.