Why your moving enquiries will soon come via AI instead of Google

The enquiry now starts with a question, not a search box

Take the Berger family, moving a four-room flat from Stuttgart to Hamburg. Two years ago they opened Google, typed "moving company Stuttgart", worked down a page of ads and listings and had six tabs open by the end of it. Increasingly, people skip all of that and open ChatGPT instead, typing out the whole situation: "we are moving a four-room home from Stuttgart to Hamburg, which companies are experienced with long-distance moves and roughly what should it cost?" The answer arrives in one block: two or three names, a price range, a reason for each.

That is the shift, and it is measurable. Pew Research Center watched real browsing across 68,879 Google searches: when an AI summary appeared, people clicked through to a website in 8% of visits, against 15% when it didn't. ChatGPT was at 900 million weekly users in February 2026; the Gemini app passed a billion monthly users in August. The ten blue links still exist. They are just read less.

For you, that moves the contest somewhere else. It is no longer only about where you sit on a results page; it is about whether you are one of the names inside the answer. The work of getting there has a name: Generative Engine Optimization, GEO for short. It decides whether you show up in the AI recommendation or never come up at all.

Why moving companies get hit first

Moving is an early indicator because every enquiry is a one-off, high-stakes decision made by someone with no practice at it. Nobody moves often enough to know what a four-room flat across the country should cost, whether a no-parking zone has to be applied for, who is allowed to carry a grand piano down three flights, or what changes when the destination is Vienna. Those are open, advice-shaped questions, and people would rather ask them once, in full sentences, than reconstruct the answer from ten websites.

Then there is the distrust. Almost everyone can name a horror story: the quote that doubled on the day, the crew that went quiet with a van full of furniture. That is why the word people put into the prompt is so often "reputable" rather than "cheap". They are asking the model to vouch for you before they will pick up the phone.

A move is also local and time-boxed at once. Fixed date, fixed address, a couple of evenings to sort it out. The customer wants three credible options tonight, not a research project. Urgency, complexity and a trust question in one enquiry is exactly the combination an AI answer serves better than a results page, which is why your sector feels this before most others do.

How an answer engine picks which movers to name

Unlike a ranked list of ten links, a generative AI works by retrieval: for a query like this it pulls a handful of sources, then writes a judgement out of them. For a mover, that pool is your Google Business Profile and its reviews, Trustpilot and ProvenExpert, directories, local press, forum threads, portals like Umzugsauktion, and your own site. None of the engines publishes how a retrieved page becomes a cited source, so exact weightings are guesswork. What you can see is what exists about you, and whether it agrees with itself.

Consistency is the part you control. If your name, address, services and pricing read the same on every profile, the model has one story to tell. An old address here, a different trading name there, opening hours that disagree, and it has several, so it hedges or names someone simpler. Ahrefs, across roughly 75,000 brands, found web mentions correlate with AI citation rate at about 0.66, some three times the figure for backlinks (0.22).

Models quote. If the numbers are genuinely yours, "family-run since 1998, 12,000 moves completed, AMÖ member, 4.8 stars from 340 reviews" is quotable, checkable, and lands in the answer. "Your reliable partner for all things moving" goes nowhere. The 2024 GEO paper found that adding quotations, statistics and cited sources lifted a source's visibility by roughly a quarter while keyword stuffing did nothing, though it ran on a simulated engine with the authors' own metrics, so treat that as direction rather than a promised lift.

What your customers actually type now

To show up in the answer you have to know how the question gets typed, and it is no longer two words. It is "what does it cost to move a 3-room flat within Munich?", "which movers also do furniture assembly and packing?", "who can organise a move to Vienna at two weeks' notice?" Full sentences, with the constraint included: size, route, deadline, extras.

Every one of those is a page you either have or you don't. Answer the exact question with a number and a process and you are a usable source; say only that you do moves and the model finds whoever went further and names them instead. So it is not enough to offer moves. Show that you handle the packing service, the piano, the no-parking permit paperwork and the customs side of a move to Austria.

Collect the real questions, not the ones you imagine. Keep a note by the phone for a month, read back through the enquiry inbox, ask whoever writes your quotes which three things come up every time. Anything a customer asks out loud is being typed into a model by someone else, and your job is to have a specific answer sitting somewhere public for each one.

What to change on your site and your profiles

Start with your own site and organise it around questions instead of keywords. Give every real service and area its own page: "Moving within Berlin", "Long-distance move to Bavaria", "Office relocation". Answer what people actually ask, from what drives the price to what a realistic range looks like and what happens on the day. Google's guidance is blunt about the rest: its AI features run on the normal Search ranking systems, no AI-specific markup or file is required, and content written separately "for AI" risks being treated as scaled abuse. Write for the customer; the machine reads the same page.

Then put the evidence within reach: founding year, how many trucks you run, AMÖ membership, insured sum per load, certifications, moves completed last year. Add a real FAQ, question as the heading and answer directly beneath, because that shape is trivial to extract. Use actual jobs as examples wherever the customer is happy for you to, because a described move is more citable than a claim about moves.

• Ask for a review after every move: one sentence at handover, with the link. Reviews are quotable third-party evidence, and mentions away from your own site track AI citations far more closely than links do.
• Keep your details identical everywhere: name, address and phone the same on the Google Business Profile, in every directory and in your imprint. A stale address or disagreeing hours is the commonest reason an answer hedges.
• Finish the profiles that already rank for your city: Umzugsauktion, MyHammer, ProvenExpert and the local directories, complete and current and matching your site.
• Mark up your facts with Schema.org: MovingCompany, opening hours, service area, reviews. Not a ranking lever, but it makes your details unambiguous, which is how you stop a model inventing your prices.
• Write up real reference jobs, such as a long-distance move of a medical practice from Freiburg to Leipzig, with rooms, distance, timeframe and what made it awkward. Specifics get quoted; adjectives don't.

Trust is the one thing you can't fake

Few trades live on trust the way this one does: the customer hands a stranger every object they own. But the signals a model can see are all second-hand. How many genuine reviews you have and how recent they are, whether an association vouches for you, whether your prices are public, whether complaints turn up next to your name. It reads what others say about you, not what you say about yourself.

Which is why you cannot fake your AI visibility for long. Bought reviews get spotted, and a model that finds signals pointing in different directions turns vague, which the customer reads as "probably not this one". The honest route is also the efficient one: do good moves, ask the satisfied customers to say so publicly, and make your real strengths easy to find and easy to quote.

Price transparency sits in the same place. Queries like "moving company with fixed prices and no hidden costs" exist because so many people got burned once. A page that explains honestly how your quote is built, from volume and distance to floor level, lift or no lift, packing and storage, answers a question your vaguer competitors leave open.

What the slow version of losing looks like

Moving runs on strangers. Almost nobody hires you twice, so there is no book of returning clients to coast on and you need a fresh stream of enquiries every month. If a growing share of that stream is routed through AI answers and you are not in them, the channel doesn't break, it thins. Nothing arrives to tell you that you were left out of an answer.

The delay is the dangerous part. Your rankings can look steady while a slice of your market has stopped opening Google for this at all. SparkToro and Similarweb measured US Google searches from January to April 2026 and found 68% ended with no click to any website, up from 60% in their 2024 study. By the time a thin quarter shows up in your calendar, the movers who started early are already the default answer, and catching up costs more than being early would have.

The opening is still there. Most moving companies have never heard the term, and most of those who have skipped the unglamorous parts: consistent data, real reviews, pages that answer the actual question. Do those now and you are in the answer before your competitors work out what the answer is.

A 90-day plan you can actually run

Start small, but start. Weeks one to four: write down the 20 questions you get asked most and answer each properly on your site, with numbers, ranges and what happens on the day. In parallel, build the handover habit, so every finished move ends with one sentence asking for a review and a link to leave it. Those two things are the foundation everything else sits on.

Weeks five to eight, get the evidence straight: identical company details everywhere, portal profiles finished rather than half-filled, AMÖ membership and insurance cover visible, three or four real jobs written up. Then find out where you stand by asking ChatGPT, Gemini and Perplexity: "which moving companies in Cologne are reputable for a long-distance move?", with your own city in place of Cologne. Note who gets named and which sources the answer leans on.

Do that monthly and both the gaps and the progress stop being guesswork. Expect two things: cited pages skew older, with Ahrefs putting the median age of a ChatGPT-cited URL at roughly 500 days, so nothing you publish today shows up tomorrow; and answers move around, so judge the trend over months, not one query. GEO is upkeep of a reputation, not a project with an end date. Keep it up and you are the name the model gives the next Berger family.