Title Tag: Grounding Drift: AI Search Volatility for Trades Meta Description: AI search volatility means your trades business can appear in ChatGPT today and vanish tomorrow. Here is what causes grounding drift and how to fix it. URL: https://firstcallvisibility.com/blog/grounding-drift-ai-search-volatility-trades Primary Keyword: AI search volatility Secondary Keywords: grounding drift, AI visibility changes, ChatGPT vs Gemini local search, AI citation stability Date: 2026-08-28 Word Count: ~1,800

<meta property="og:title" content="Grounding Drift: Why Your Trades Business Shows Up in AI Answers Today and Disappears Tomorrow"> <meta property="og:description" content="AI search volatility means your trades business can appear in ChatGPT today and vanish tomorrow. Here is what causes grounding drift and how to fix it."> <meta property="og:type" content="article"> <meta property="og:url" content="https://firstcallvisibility.com/blog/grounding-drift-ai-search-volatility-trades"> <meta property="og:image" content="https://firstcallvisibility.com/og-images/grounding-drift-trades.jpg"> <meta name="twitter:card" content="summary_large_image"> <meta name="twitter:title" content="Grounding Drift: Why Your Trades Business Shows Up in AI Answers Today and Disappears Tomorrow"> <meta name="twitter:description" content="AI search volatility means your trades business can appear in ChatGPT today and vanish tomorrow. Here is what causes grounding drift and how to fix it.">

Grounding Drift: Why Your Trades Business Shows Up in AI Answers Today and Disappears Tomorrow

A homeowner asks ChatGPT for a plumber on Monday. Your business shows up in the answer. They ask the same question on Wednesday, and your name is gone. Nothing changed about your website, your reviews, or your Google Business Profile. The AI just picked different sources the second time.

This is not a glitch. It is a documented pattern called grounding drift, and it is the single biggest reason AI visibility feels impossible to hold onto.

Steady Demand, a local SEO agency, ran 1,487 local search queries across 50 U.S. metro areas and 10 service categories through Gemini in August 2026. They collected 14,472 citations and published the results as the AI Citation Ledger. The findings were covered by Search Engine Land on August 19.

The headline number: repeating the same Gemini query produced overlapping cited sources only about 40% of the time. Google's local pack returned the same top listing about 90% of the time. Gemini recommended the same top business just 7% of the time.

For a trades business owner, that means a single AI search is a snapshot. It is not a reliable measure of whether your business is visible.

What Causes Grounding Drift in AI Search

AI engines do not work like Google's map pack. Google ranks businesses using a relatively stable set of signals: proximity, relevance, and prominence. The top three spots tend to stay the same week to week. AI engines re-evaluate from scratch on every single query, which is a fundamentally different process.

When ChatGPT or Gemini answers a question like "best plumber near me," the model pulls from a mix of sources. It might cite your website, a Yelp listing, a Reddit thread, an Angi profile, or a local newspaper article. Which sources it picks depends on the model's current training data, whether it is browsing the web in real time, and which sources it weighted most heavily on that particular query.

Gemini has direct access to the Google ecosystem: Google Business Profile data, Maps signals, Google reviews, and location relevance. ChatGPT pulls through Bing's index and relies more heavily on Reddit and business directories. Perplexity does live web searches and weighs content freshness more than the other two.

The problem is that all three engines re-evaluate on every query. There is no cached ranking. The model re-reads its sources, re-weights signals, and produces a fresh answer each time. Small changes in how it interprets the query can shift which businesses make the cut.

Steady Demand co-founder Ben Fisher put it plainly: a single AI search should be treated as a snapshot, not a reliable measure of long-term visibility.

Your Website Matters More Than Directories for Gemini

One finding from the Steady Demand study cuts against what most trades business owners assume. Nearly 60% of Gemini's local citations pointed directly to business websites. That was more than directories, review platforms, and forums combined.

Reddit ranked second, accounting for 13.7% of citations. That beat Angi, Thumbtack, and HomeAdvisor put together.

For years, the local SEO playbook told trades businesses to focus on directory listings: claim your Angi profile, fill out your HomeAdvisor page, get on Thumbtack. Those still matter. But the Steady Demand data suggests that for Gemini at least, your own website content carries more weight than any single directory.

This makes sense when you think about how Gemini works. It is grounded in Google's index. Google has been crawling business websites for years. It has rich entity data: your business name, your services, your service area, your hours, your reviews. When Gemini needs to recommend a local business, it pulls from that well of structured data. A business with a clean, well-structured website with LocalBusiness schema gives Gemini more to work with than a thin Angi listing.

ChatGPT and Gemini Rarely Agree

Steady Demand ran the same 1,487 queries through ChatGPT to compare. The results showed how different these engines are.

Both platforms cited the same domains only 8% of the time. They recommended the same top business in just 4.2% of queries.

Gemini mainly cited business websites. ChatGPT relied much more on Reddit and business directories. A plumbing company that shows up consistently in Gemini's answers might be completely absent from ChatGPT's responses to the same query, and vice versa.

This is the core problem with monitoring only one platform. A trades business owner who checks ChatGPT and sees their name might assume they are visible across AI search, but visibility on one engine, on one query, on one day does not carry over to the next query.

Why This Matters for Plumbers, HVAC, and Electricians

Trades businesses are already at a disadvantage in AI search. The 5WPR HVAC and Plumbing AI Visibility Index found that 87% of independent contractors have effectively zero AI citation share. National franchises dominate because they have consistent entity data across hundreds of locations, schema markup on every page, and mentions in trade publications.

An independent plumber with a decent Google ranking and a handful of reviews can show up in AI answers one week and disappear the next. The volatility is not a sign that something is broken. It is a sign that AI search is inherently less stable than traditional search.

But volatility also creates opportunity. If the top business changes 85% of the time, then the number one slot is not permanently locked. A business that builds strong entity signals, consistent directory listings, and answer-shaped content on its own website can move into that slot on any given query. The barrier to entry is lower than the barrier to staying.

How to Build AI Visibility That Survives Grounding Drift

You cannot eliminate volatility. AI engines re-evaluate on every query, and model updates shift the ground without warning. But you can build signals that make your business more likely to appear across queries, across platforms, and across model updates.

1. Build your website for entity richness. Gemini cites business websites more than any other source type. Your website needs structured data: schema markup, consistent name-address-phone information, service area pages, and clear service descriptions. A page that says "plumbing services" with no detail gives Gemini nothing to cite. A page that says "emergency drain cleaning in Bergen County, NJ, available 24/7, licensed master plumber #12345" gives it everything.

2. Get mentioned in places AI engines read. ChatGPT relies on Reddit and directories more than Gemini does. Being mentioned in a Reddit thread about "best plumber in [city]" carries real weight. So does a consistent profile across Bing Places, Yelp, and the Better Business Bureau. F9XR's analysis of 267,000 AI citations found that third-party mentions matter as much as your own website for ChatGPT.

3. Keep content fresh. Gemini's model swaps have increased the weight of freshness. Pages updated within 30 days gained an edge in citation share after the Gemini 3 swap. A blog post about common plumbing problems in your service area, updated quarterly, signals that your business is active. A page that has not changed since 2023 does not.

4. Monitor across all three platforms. Checking ChatGPT alone gives you a partial picture. You need to run the same queries across ChatGPT, Gemini, and Perplexity to understand where you appear and where you do not. Track results over time, not just once. A single snapshot tells you nothing.

5. Track separately from Google rankings. A strong Google Maps position does not guarantee AI visibility. SOCi's 2026 Local Visibility Index found that fewer than half of the brands with top traditional local search visibility also ranked among the brands AI recommends most often. Treat AI visibility as its own channel with its own metrics.

What Grounding Drift Means for Your Daily Operations

If you are running a plumbing, HVAC, or electrical business, grounding drift means you cannot set up your online presence once and expect it to hold. The businesses that maintain AI visibility are the ones that treat it as ongoing maintenance, not a one-time project.

That means periodically updating your website content, checking that your business information is consistent across every directory, asking satisfied customers for reviews on the platforms AI engines read, and running test queries across ChatGPT, Gemini, and Perplexity to see where you stand.

FirstCall's free AI visibility audit runs a 12-point check across all three platforms and tells you exactly where your business is visible, where it is missing, and what to fix first. It takes the guesswork out of a process that changes every time a model gets updated.

FAQ

What is grounding drift in AI search? Grounding drift is a pattern where repeating the same AI search query produces different recommended businesses each time. Steady Demand's August 2026 study found that repeating the same Gemini query produced overlapping cited sources only about 40% of the time, compared to 90% consistency for Google's local pack.

How often do AI search results change for local businesses? AI recommended businesses show roughly 85% volatility, meaning most of the businesses AI recommends in a given month are different from the ones it recommended the previous month. This varies by platform: Gemini recommended the same top business only 7% of the time in repeated queries.

Do directories like Angi and HomeAdvisor matter for AI search? They matter less than most trades business owners assume. Steady Demand found that business websites accounted for nearly 60% of Gemini's local citations, while Reddit alone accounted for 13.7%. The entire local-service directory category, including Angi, Thumbtack, and HomeAdvisor, trailed both. Your own website content carries more weight than any single directory for Gemini.

Should I track AI visibility on just one platform? No. Steady Demand found that Gemini and ChatGPT cited the same domains only 8% of the time and recommended the same top business just 4.2% of the time. Monitoring only one platform gives a misleading picture. Run the same queries across ChatGPT, Gemini, and Perplexity.

How can a trades business maintain AI visibility despite volatility? Build entity-rich website content with schema markup, maintain consistent listings across directories and Reddit, keep content updated within 30 days, and monitor across all three major AI platforms. Volatility means the top slot is not permanently locked. Strong entity signals give your business a better chance of appearing on any given query.