Strategy & Planning · 9 min read · July 15, 2026
What questions clients really ask the AI about tax advisors
More and more clients no longer ask Google but ChatGPT: "Which tax advisor near me knows about setting up a GmbH?" The AI answers with concrete recommendations, and thereby decides who even makes it onto the shortlist. Whoever doesn't appear in these answers simply doesn't exist for a growing share of searchers. This is exactly where Generative Engine Optimization for tax firms comes in.
Why clients ask the AI today instead of the firm
The first contact with your firm increasingly no longer takes place on your website, but in a chat window. An entrepreneur considering switching from a sole proprietorship to a GmbH no longer types "tax advisor GmbH conversion" into Google in the evening and clicks through ten results. They ask ChatGPT directly: "As a freelancer with 90,000 euros of profit, is a GmbH worth it for me and which tax advisor helps with that?" The answer comes instantly, pre-formulated and seemingly objective. That changes the entire first contact.
This shifts the rules of the game fundamentally. In the past you fought for position one on Google. Today you fight to even be named in the single answer the AI spits out. There's no longer a second results page to fall back on. Either your firm appears in the generated text or it doesn't. There's hardly anything in between, and the client doesn't even notice whom the AI left out.
For tax advisors that's especially delicate, because trust and professional competence are the currency of the sector. If the AI recommends three firms and yours isn't among them, it comes across like a silent verdict on your relevance. The client contacts the named firms and you never learn that you were even in the running. Visibility in AI answers is therefore no nice-to-have, but the new basis of client acquisition.
The real questions: what clients actually type into the AI
The questions to the AI sound different from classic search terms. They're longer, more personal and often tied to a concrete situation. Typical are: "I have an online shop with sales in several EU countries, which tax advisor knows about the OSS procedure?" or "My tax advisor never gets back to me, what do I watch out for when switching?" or "What does a tax advisor cost for a small GmbH with two employees?". These are complete advisory enquiries, not keywords.
It's striking how often clients ask about specialisation. "Tax advisor for doctors", "tax advisor for influencers and content creators", "tax advisor for cryptocurrencies", "tax advisor with experience in tax audits in hospitality". The AI loves clear niches, because it can pin relevance on them. A firm that promises only "comprehensive tax advice for all industries" on its website delivers the AI not a single signal it could match to a concrete question.
Equally common are questions about process and costs: "How does a first consultation with a tax advisor go?", "Can I also submit my annual accounts digitally?", "Which documents does the tax advisor need for the income tax return?". Whoever answers these questions honestly and fully on their own website is drawn on by the AI as a source. These everyday questions are exactly the raw material from which AI systems build their recommendations.
How the AI even decides which firm to name
AI models draw on two things: what they learned during training, and what they load live from the web. For a local recommendation like "tax advisor in Regensburg for trades businesses", systems like ChatGPT's web-search mode or Google Gemini pull in current sources: your website, your Google Business Profile, industry directories, review portals and specialist articles. If you're missing from these sources, the AI simply has no data with which to suggest you.
The decisive factor is the match between the question and your publicly findable statements. If a client asks for "tax advisor start-up funding" and you have a detailed, clearly structured article on exactly that on your website, the AI recognises the fit. It then often even cites your name or your firm name directly. Vague marketing phrases, by contrast, are unreadable to the machine, because they answer no concrete question.
On top of that comes consistency. If your firm name, your address and your range of services stand the same everywhere, website, Google profile, chamber of tax advisors directory, LinkedIn, then the AI gains trust in the data. Contradictory details lead the systems, in case of doubt, to prefer naming another, more clearly identifiable firm. Clarity beats loudness here.
From keywords to real answers: GEO for tax firms
Generative Engine Optimization means preparing your content so an AI can use it as an answer building block. The practical core: phrase the actual client questions as headings and answer them directly beneath in two or three clear sentences. Instead of a "Services" page, you build pages like "What does bookkeeping cost for a small GmbH?" or "How do I switch tax advisors without hassle?". That's exactly the structure the AI cites from.
The honesty of the answers matters. Tax advisors can't communicate entirely freely about prices, because the Tax Advisor Remuneration Regulation sets the framework. But you can explain how fees are made up, which factors drive the effort and what a first consultation includes. Such solid, specific statements are preferred by the AI over advertising language, because they seem verifiable and offer the asker real value.
Complement this with professional depth that shows your positioning. An article about VAT on intra-community supplies, written in understandable language with a concrete example, signals competence in a niche. This is exactly what the AI pins its recommendation on when a client asks with exactly this problem. Depth beats breadth, because depth gives the machine a clear matching signal.
Reviews, mentions and external signals
No AI system relies only on your own website. It checks what others say about you. Google reviews, entries in tax-advisor directories, mentions in local business media and specialist portals flow into the picture the machine paints of your firm. A firm with many recent, substantively concrete reviews is named considerably more often than one without any trace on the web.
Actively ask satisfied clients for reviews that get concrete. "Accompanied our GmbH founding from A to Z" is more valuable to the AI than a mere "very good, recommendable", because it names a concrete service that fits a concrete question. Also respond visibly to reviews, because that too is a public signal of reliability and reachability.
You have to earn some mentions actively. A guest article in the local business association, an interview in the industry newsletter, a specialist article on a tax portal, all of that generates mentions of your name in connection with your expertise. For the AI, these external confirmations are like independent witnesses. The more of them exist, the more confidently it names you as an answer.
The practical roadmap for your own firm
Start with an honest stocktake: put the ten questions your clients really have to your firm in ChatGPT, Gemini and Perplexity. Are you named? Are competitors named? This test research shows you in fifteen minutes, in black and white, where you stand. Note which firms show up and what they do differently on their website than you.
Then you systematically build answer content. Choose the five most important client questions of your target group and answer each on its own clearly structured page. Use the question as a heading, give the core answer in the first two sentences and then add details, an example and your concrete relevance. Keep the language understandable, because the AI prefers texts a person understands without jargon.
Finally, ensure consistency and upkeep. Unify your details across all platforms, keep your Google profile current, collect reviews continuously and update your content when tax law changes. GEO isn't a one-off project but a routine. Whoever invests two hours every three to four weeks builds, over a year, a head start competitors can only catch up on with difficulty.
What you'd better leave out
The biggest mistake is trying to convince the AI with volume. Twenty superficial blog posts full of empty phrases bring less than five precise answer pages. AI systems recognise thin content and ignore it. Just as damaging is having texts generated entirely by an AI and published unchecked, because faulty tax statements cost you not only visibility but, in the worst case, your neck in terms of liability.
Also refrain from exaggerated promises. "We'll guarantee to save you thousands of euros in tax" is professionally dicey and, for the AI, a warning sign, because it's unsubstantiated. Reputable, verifiable statements come across as more credible to machine and client alike. The AI prefers sources that sound sober and fact-based, because they lower the risk of a wrong recommendation.
And finally: don't ignore the development in the hope that clients will come via referral anyway. Referrals remain important, but even the referred client checks you additionally in the AI today before calling. If the machine finds nothing about you there or plays out contradictions, even a good referral can come to nothing.
Conclusion: visibility is the new reachability
The way clients look for a tax advisor has changed quietly but fundamentally. The decisive question is no longer whether you have a pretty website, but whether the AI knows your firm and names it at the right moment. That's no passing fad, but the logical continuation of what began with Google search, only faster, more direct and with less room for the runner-up.
The good news: tax advisors have a real advantage here. Your work is full of concrete questions, clear answers and demonstrable professional competence. Exactly what AI systems need to recommend someone. Whoever makes this substance visible instead of hiding it behind marketing phrases is favoured by the machine. It's not about getting louder, but clearer and more honest.
The best time to start is now, while most firms are still sleeping through the topic. Every answered client question, every consistent detail, every concrete review is a building block of your AI visibility. Start small, keep at it and measure your progress by regularly asking the AI yourself. That way you make your firm the answer the machine gives.
Common questions
How do I find out whether ChatGPT even knows my tax firm?
Put your clients' questions to the AI, for example "Which tax advisor in my town helps with setting up a GmbH?" or name your firm directly and ask what the AI knows about you. Use ChatGPT with web search, Gemini and Perplexity, because they answer differently. That way you see within a few minutes whether and how you're named and which competitors show up.
As a tax advisor, am I even allowed to state prices online so the AI picks them up?
You're subject to the Tax Advisor Remuneration Regulation and may not freely advertise dumping prices. But it's permitted and sensible to explain transparently how fees are made up, which factors determine the effort and what a first consultation covers. The AI prefers such comprehensible explanations over vague statements, without you crossing professional boundaries.
Is my Google Business Profile enough or do I have to do more?
The Google profile is an important source for local AI recommendations, but on its own it isn't enough. AI systems combine your profile with your website, reviews, directory entries and specialist articles. Most effective is the combination of a well-maintained Google profile, consistent details everywhere and several concrete answer pages on your target group's real questions. Only this interplay makes you clearly recommendable to the machine.
Read on