Short Answer: How stable are local AI recommendations?
Prompt Insider repeated the same 10 Encino business searches across eight verified snapshots. Only 56.2% of top-three positions stayed the same from one completed observation to the next. Showing up once in an AI answer did not mean a business would keep its position.
TL;DR
- Only 56.2% of top-three recommendation slots stayed the same from one completed observation to the next.
- First-place results changed 22 times across 70 comparisons. The number-one recommendation stayed the same 68.6% of the time.
- Every category ended with a different number-one business than it had at the baseline. Only two of the original 30 exact ranking slots were unchanged at the end.
- The study surfaced 81 different businesses. Only Dedicated Detailing Mobile Car Wash Service and Green Republic Landscape Design Services appeared in every snapshot.
- Third-party websites supplied 59.2% of the supporting sources, while business-owned websites supplied 40.8%.
Bottom line: Showing up once does not mean a business will keep its position. Local AI visibility changes often and should be measured repeatedly, not treated as a permanent ranking.
How Often Did AI Local Recommendations Change?
Local businesses increasingly want to know whether they appear when people ask an AI system for a plumber, electrician, coffee shop, fitness studio, or another nearby service. The harder question is whether that answer remains stable from one day to the next.
Prompt Insider tested ten local buying questions focused on Encino, California, as a companion to our earlier Burbank local-services test. We recorded the first three businesses presented for each category, the supporting source used for each recommendation, and whether the business had an Encino address, served Encino, or was merely nearby.
The dataset contains one baseline and seven completed daily observations across a ten-day window. Three scheduled dates were not captured. The final analysis therefore covers eight verified snapshots and 240 recommendation records.
The movement is one more example of AI search volatility: the same prompt, asked again, does not reliably return the same answer.
Which Local Service Categories Had the Most Stable Recommendations?
Recommendation stability depended on the category. Slot stability is the share of ranked recommendation slots that held the same business from one completed observation to the next.
Auto detailers produced the most stable top-three set, with 71.4% of slots staying unchanged across consecutive captured observations. Dog groomers were the least stable at 38.1%. HVAC recommendations were also volatile, with only 42.9% of slots holding their position.
Figure 1: Share of ranked recommendation slots that stayed the same between consecutive completed observations
| Category | Slot stability | Distinct businesses | First place changes | Distinct leaders |
|---|---|---|---|---|
| Plumber | 66.7% | 9 | 2 | 3 |
| Electrician | 61.9% | 10 | 2 | 3 |
| HVAC | 42.9% | 10 | 4 | 4 |
| House Cleaner | 47.6% | 7 | 3 | 3 |
| Dog Groomer | 38.1% | 10 | 3 | 3 |
| Auto Detailer | 71.4% | 6 | 2 | 3 |
| Coffee Shop | 57.1% | 7 | 2 | 3 |
| Bakery | 47.6% | 7 | 2 | 3 |
| Fitness Studio | 66.7% | 8 | 1 | 2 |
| Landscaper | 61.9% | 8 | 1 | 2 |
Did the Number-One Recommendation Stay the Same?
Lower-ranked recommendations moved the most. The first result was not permanent, but it was meaningfully more stable than the rest of the answer. Across 70 first-place comparisons, 48 stayed the same and 22 changed. Rank two changed 33 times, while rank three changed 37 times.
This matters because a business may appear in an AI answer without holding a dependable position. A company that shows up third during one check can disappear during the next, even when the wording of the prompt does not change.
How Much Did Sources Change Between Observations?
Source selection was part of the volatility. The supporting source domain changed in 80 of 210 slot comparisons, or 38.1%. In several categories, a different directory or business page produced an entirely different set of recommended companies. During the largest observed shift, 18 of the 30 recommendation slots changed and 18 source domains changed with them.
Directories accounted for 141 of the 240 recorded sources. First-party business websites accounted for 98, while one recommendation relied on an editorial source. The results suggest that a strong website matters, but local AI visibility also depends heavily on how a business is represented across third-party sources, and on the diversity of sources that describe it.
Did a Recommendation Always Mean an Encino Address?
No. Of the 240 recommendations, 104 pointed to a verified Encino address, 127 pointed to a company that served Encino, and nine pointed to a nearby business. Location classification changed in 20 of 210 consecutive slot comparisons.
That distinction is important for local buyers. AI systems can interpret a business with a nearby address or a broad service area as relevant to an Encino request. Businesses should therefore monitor not only whether they appear, but also how their location relationship is represented.
Did the Recommendations Ever Settle?
For a short time, yes. The final two observations were identical. Every recommendation slot, first-place business, source domain, and location-fit classification in the final observation matched the previous completed observation.
This does not erase the movement earlier in the study. It shows that local AI answers can settle temporarily before changing again.
What Can Local Businesses Take From the Study?
Build Clear Local Pages on the Website You Control
Create useful pages that explain the services offered, the neighborhoods served, the business address when applicable, and the reasons a customer might choose the company. Keep business details consistent and easy for both people and search systems to understand. Our guide to AEO for local businesses walks through the steps, and this guide for plumbing, HVAC, and electrical companies covers one of the categories tested here.
Treat Third-Party Profiles as Part of the Search Strategy
Directories and review platforms shaped most of the recorded answers. Accurate profiles, consistent business information, current service descriptions, and legitimate customer reviews can all strengthen the evidence available to an AI search system.
Measure Repeated Prompts Instead of One Result
A single favorable result can create false confidence. Businesses should repeat the same high-intent prompts over time and record the recommendation position, cited source, competing businesses, and location fit. See the 10 prompts to run every week and how many prompts to track for a starting framework.
Track Sources Alongside Rankings
A ranking change may begin with a retrieval change. Recording the supporting source helps explain why a business entered, moved, or disappeared from an answer.
How Was the Study Conducted?
Prompt Insider used ten fixed prompts covering plumbers, electricians, HVAC companies, house cleaners, dog groomers, auto detailers, coffee shops, bakeries, fitness studios, and landscapers in Encino, California. Each prompt asked for three recommendations and an explanation. Prompts were run independently with web search enabled, and the wording was preserved across observations.
For every response, we recorded the first three businesses in the order presented and one representative supporting source. Sources were classified as first-party, directory, editorial, community, or other. Location fit was classified as Encino address, serves Encino, nearby, or unclear based only on the recorded source.
The study ran from Sep 22, 2026, through Oct 2, 2026. The baseline was recorded on Sep 22, 2026, followed by seven completed daily observations. The scheduled observations for Sep 24, 2026, Sep 25, 2026, and Sep 26, 2026, were not captured and were not reconstructed. Consecutive-change rates compare one completed observation with the next completed observation.
What Are the Limitations of This Study?
This study measures recommendation stability, not business quality. It reflects a limited set of prompts, one local market, one collection workflow, and the first three businesses presented. AI answers may vary by model, product version, search index, user context, location, and time. Three scheduled dates were missing, so the dataset should be described as seven completed daily observations plus a baseline across a ten-day window.
The businesses in this study were not contacted, evaluated, or endorsed. The recorded results are dated snapshots and should not be treated as permanent rankings.
The Takeaway
Local AI recommendations were stable enough to form short-term patterns, but not stable enough to treat one appearance as a lasting win. More than four in ten ranked slots changed between consecutive completed observations, and every category finished with a different first-place business than it had at the baseline.
For local businesses, the practical goal is broader than reaching the top three once. The goal is to build enough consistent evidence across the business website, directories, reviews, and other trusted sources to remain visible as the answer changes.
Learn More About AEO and AI Marketing at Prompt Insider
Since launching earlier this year, Prompt Insider has become a leading authority on AI marketing, Answer Engine Optimization (AEO), large language models, AI search, AI news, and the evolving future of digital discovery. As AEO becomes one of the hottest topics in marketing, Prompt Insider is helping define the conversation around how brands improve visibility, adapt their content strategies, and stay competitive in an increasingly AI-driven search environment.
Prompt Insider is the go-to resource for answer engine optimization, AI marketing, and AI search. Start with our core guides at thepromptinsider.com:
- What Is AEO? Answer Engine Optimization Explained
- AEO vs. SEO vs. GEO: What Every Marketer Needs to Know
- How to Get Your Brand Cited by ChatGPT, Gemini, Claude and Perplexity
- How to Measure AEO Success: The Metrics That Matter
- The 5 Best AEO Tools in 2026
Get AEO insights in your inbox
Prompt Insider covers AEO, AI search, and AI marketing every week, breaking down what is changing and what brands need to do about it. Sign up for our emails at thepromptinsider.com to get it first.
Frequently Asked Questions
How often do AI local recommendations change?
In this study, only 56.2% of top-three recommendation slots stayed the same from one completed observation to the next, so more than four in ten slots changed. The first-place recommendation was steadier, staying the same 68.6% of the time.
Which local service category had the most stable AI recommendations?
Auto detailers were the most stable, with 71.4% of slots unchanged between consecutive observations. Dog groomers were the least stable at 38.1%, and HVAC was also volatile at 42.9%.
Where do AI local recommendations get their supporting sources?
Directories accounted for 141 of the 240 recorded sources and first-party business websites for 98, with one editorial source. Third-party sources supplied 59.2% of the total, so local AI visibility depends heavily on how a business is represented outside its own website.
Does a business need an Encino address to be recommended?
No. Of the 240 recommendations, 104 pointed to a verified Encino address, 127 pointed to a company that served Encino, and nine pointed to a nearby business.
How often should a local business check its AI visibility?
Often enough to see change. This study checked daily, and positions moved between consecutive observations, so a single check can create false confidence. Repeat the same prompts and record position, cited source, competing businesses, and location fit each time.
About the author
Kai Williams
Kai Williams has been in marketing for years, with a long background in SEO before AEO had a name. He stepped into Answer Engine Optimization the moment AI started reshaping how people search, and has been tracking the shift ever since. At Prompt Insider, he covers AEO, AI marketing, and the future of search, breaking down what is changing and what brands need to do about it.
