Technical & Structure · 9 min read · July 15, 2026

Maintaining your assortment and opening hours so the AI states them correctly

When a customer asks ChatGPT whether your shop is still open today or whether you carry a particular brand, it's not your gut feeling that decides the answer, but the data traces you leave on the web. This guide shows you how to maintain assortment and opening hours in retail so that generative AI reproduces them correctly, currently and completely.

Why AI so often gets opening hours and assortment wrong

Imagine someone standing in the pedestrian zone on a Saturday afternoon asking ChatGPT: "Which stationery shop nearby is still open now and carries Faber-Castell Polychromos?" The AI doesn't answer out of nowhere. It draws its statement from what it finds about you: Google Business Profile, your website, trade directories, review portals. If these sources are contradictory or outdated, the AI guesses or simply names your competitor whose data is cleaner.

In retail this is exactly the norm. The website says "Mon to Fri 9 a.m. to 6 p.m.," the Google profile says "until 6:30," and the old Yellow Pages entry still has the hours from three years ago. For a human that's annoying, for an AI it's a signal that it can't trust your details. Inconsistent data leads the AI to become cautious and recommend you concretely less often.

With the assortment it's even trickier. Hardly any retailer publicly maintains which brands and product groups they actually carry. The AI then only knows "fashion shop" or "sporting goods," but not that you're the only one in the area who stocks running shoes in width XL and trail-running models. But these very specifics are what customers ask the AI about.

Opening hours: one truth, identical everywhere

The most important rule first: your opening hours have to read exactly the same, character for character, everywhere on the web. Define a master source, usually your Google Business Profile and your own website, and align everything else to it. Enter in Google exactly the same hours as in the legal notice, on the contact page, at Apple Business Connect, at Bing Places and in the directories that list you. Every deviation is a point at which the AI starts to doubt.

Especially underestimated are special opening hours. Sundays with shops open, holidays, inventory days, extended hours in Advent: maintain these actively in your Google profile as "special opening hours." If someone asks the AI on December 24 whether you're open, it increasingly draws on this structured holiday data. If it's missing, the AI names the regular hours and may be completely wrong, which means a disappointed customer in front of your locked door.

A practical tip from retail: set yourself a fixed rhythm, say every first Monday of the month, on which you briefly check all entries. Five minutes are enough. Whoever sets up the hours cleanly once and then maintains them with discipline gets named noticeably more reliably by the AI than the neighbor who hasn't touched their entry since the shop opened.

Structured data: how the machine reads your hours

AI systems and search engines love machine-readable formats. On your website you should record your opening hours as structured data, concretely with the schema type "LocalBusiness" and the field "openingHoursSpecification." That's a small code building block in the source code that your web developer, or many shop and website builders, add with a few clicks. For the AI this is the difference between "I suppose" and "I know."

The advantage: while a running text like "we're basically always there" is worthless to the machine, the schema delivers exact weekdays, times and even exceptions in unambiguous form. Be sure to add address, phone number and location details in the same structured block. This way the AI can assign your shop unambiguously to a location, which with common names like "Modehaus Muller" is what tips the scales on whether you or a shop of the same name gets recommended.

You can check this for free with Google's Rich Results Test or a schema validator. If you enter your URL there, you see whether the hours are cleanly recognized. If you find errors here, the AI finds them too. This half hour of effort pays off, because structured data is the most stable foundation for correct AI answers.

Making the assortment visible: writing out brands and product groups

The AI can only name what stands somewhere in text form. If you want it to describe you as a "wine shop with a large selection of natural wines and Austrian winemakers," then exactly that has to stand on your website, in clear sentences, not just as an image of the shelf. Set up a page or a detailed section that describes your assortment in words: the product groups, the most important brands, the price ranges, the particularities.

Think about how your customers ask. Nobody types "specialist retailer for stationery." People ask: "Where can I get refillable fountain pens and replacement cartridges in Regensburg?" or "Which shop carries vegan running shoes?" Take up these real phrasings. List concrete brand names, concrete product categories and typical occasions. The closer your text is to real customer language, the sooner the AI matches the question with your offering.

Beware of exaggeration: write only what you really carry. If the AI recommends you for a brand you don't have at all, the customer stands disappointed in the shop and you lose trust doubly. Honest, precise assortment details are more valuable long-term than bold promises the reality on the shelf doesn't back up.

Using the Google Business Profile as an assortment showcase

Your Google Business Profile is one of the most important sources for the AI, and it can do much more than address and hours. Use the "Products" section to set up central product groups with image, name and short description. Set the fitting main category and several additional categories, for instance "shoe shop," "sporting goods shop," "running-shoe specialist." These categories are strong signals for the AI about what it should recommend you for.

The attributes are worthwhile too: "accessible," "fitting possible," "click and collect," "gift vouchers." Customers filter their AI queries by exactly such features, for instance "toy shop with gift wrapping nearby." Also answer the questions in your profile's Q&A section yourself before others do. These answers likewise flow into what the AI knows and reproduces about you.

Post short updates regularly about new arrivals in the assortment, seasonal goods or freshly received collections. That keeps your profile fresh and signals currency. A profile that lives is classified by Google and the AI systems connected to it as more trustworthy than a dead file that hasn't changed in two years.

Reviews and external sources: the second voice about you

The AI doesn't just believe you, it cross-checks. What stands about you in reviews, local blogs, city portals and trade directories often carries more weight than your own self-presentation. When ten customers write that you have a "great selection of maternity fashion," this assortment feature anchors itself in the AI's training and search data, even if you yourself never phrased it that way.

Use this actively. Ask satisfied regulars for reviews and respond to reviews in a way that naturally mentions your assortment and your hours, for instance "Glad you liked our selection of regional cheese, we're here for you on Fridays until 6:30." Such responses are public, machine-readable text that connects your core statements with external credibility.

Watch for consistency here too: if a portal lists you with an outdated address or a wrong category, correct it. Many directories take over data from one another, and a single wrong entry can multiply. The effort of cleaning up old entries is invisible work, but it co-decides whether the AI paints a coherent or a contradictory picture of you.

Seasonal goods and fast changes: the tempo problem in retail

Retail lives from currency: end-of-summer sale, new autumn collection, the sold-out toy before Christmas. AI systems are naturally slower here, because they draw on data states that aren't accurate to the second. You can't fully resolve this, but you can cushion it. Communicate changes first where the AI accesses quickly: in the Google profile and in clearly dated website sections.

Avoid writing time-critical statements permanently into your running text. If your homepage still says "Now the Christmas assortment" in July, that confuses customers and AI alike. Work instead with clearly dated posts and a fixed, timeless assortment framework that describes the permanent goods. The framework gives the AI stability, the posts deliver the freshness.

Be honest about availability. If you don't show live stock online, then don't promise day-current availability either. Better phrase it "we carry regularly" instead of "currently in stock." This way you avoid the AI assuring an availability that doesn't exist on the shelf, and you save yourself disappointed customers who came especially for it.

Cross-check: ask the AI itself what it knows about you

The most honest test costs nothing. Open ChatGPT, Gemini and Google and put the questions your customers would ask: "Is [your shop name] open on Saturday?", "Where can I get [your core product] in [your city]?", "Does [your shop name] carry brand X?" Note what the AI answers. This cross-check shows you in black and white where your data trace is incomplete or wrong.

If you notice the AI doesn't name you at all even though you carry exactly that, then the text is missing at the source. If it names wrong hours, there's an outdated entry somewhere. If it confuses you with another shop, your location isn't recorded unambiguously enough. Each wrong answer is a concrete task on your upkeep list, not an abstract problem.

Repeat this test every few months, especially after changes. This way you measure whether your upkeep works. AI visibility is not a one-time project but a routine like restocking the shelves. Whoever takes it seriously gets recommended at the very moment the customer, phone in hand, decides which shop to go to.

Common questions

As a small retailer, do I really need structured data on my website?

It's not mandatory, but it's the single most effective lever. Structured data (Schema.org "LocalBusiness" with opening hours) gives the AI unambiguous, machine-readable details instead of interpretable running text. Most website builders and shop systems offer this via a setting or plugin. If you implement only one thing, then this in combination with a well-maintained Google Business Profile.

How often does the AI even name individual shops instead of big chains?

More often than you think, especially with local and specific queries like "owner-run wine shop with natural wines near me." This is exactly where small retailers have an advantage, because they have a clear profile. The precondition is that your assortment and your particularities are findable in text form. Chains dominate generic queries, but with specifics whoever writes out their niche cleanly wins.

What do I do if the AI names a wrong assortment or wrong hours for my shop?

First find the source. Usually behind a wrong AI answer sits an outdated entry in a directory, a contradictory Google profile or missing text on your website. Correct the master sources (Google profile, own website) and old third-party entries. Changes take a few weeks until they reach the AI. Afterward check again via cross-check whether the answer is now correct.

Read on

Local & Industries