The location test: can an AI assistant actually find your locations?
Every multi-location agency will offer an audit. This one you can run yourself in twenty minutes, before you pay for anything, and it will tell you which kind of agency you need.
Pick five locations, not all of them
Choose your best-performing location, your worst, your newest, one in a dense metro, and one in a small market. The pattern across those five tells you almost everything; running all eighty tells you the same thing more slowly.
Ask four questions about each, in a logged-out session
Use ChatGPT, Perplexity, and Google's AI answers. Log the responses with the date.
- Existence: “Does [brand] have a location in [city]?”
- Detail: “What are the hours and address of [brand] in [city]?”
- Category: “Who does [service] in [city]?” — do you appear unprompted?
- Comparison: “[brand] vs [local competitor] in [city]” — what does it say about you, and where did that come from?
Read the results as a diagnosis
| What you see | What it means | Who fixes it |
|---|
| The location is not known to exist | No indexed location page, or one that cannot be rendered or crawled | Technical and templating work — an SEO team, not a listings vendor |
| It exists but the hours or address are wrong | Your sources disagree with each other; the model picked one | Listings and data-consistency work |
| It exists, details are right, but you never appear for category questions | No third-party corroboration — you are absent from the directories and articles being cited | Digital PR, review platforms, local citations |
| Details are right for corporate locations and wrong for franchised ones | Entity fragmentation — the classic franchise failure | A franchise specialist, and a governance conversation before any marketing work |
Then audit where your location data actually lives
For one location, write down what each of these says. They are almost never identical, and every inconsistency is a chance for a model to hedge or omit.
- The corporate site's location page
- The franchisee's or branch's own site or social page, if one exists
- The Google Business Profile
- The major directories and review platforms for your category
- Whatever your booking, ordering, or scheduling system publishes
Name, address, phone, hours, services offered. Where two sources conflict, decide which one is authoritative — that decision is the actual deliverable, and it is a governance decision that no agency can make on your behalf.
The arithmetic that decides whether to hire at all
Multi-location retainers are quoted as one number, but the number that matters is cost per location per month. Divide the proposed fee by your location count, then compare it to the margin a single additional customer produces at one location.
- A $6,000 monthly retainer across 40 locations is $150 per location per month. If one extra customer a month per location clears that, the programme is straightforward to justify.
- The same $6,000 across 6 locations is $1,000 each, and now the maths is closer to a single-location engagement and should be judged like one.
- Below roughly ten locations, the templating and governance advantage that justifies a multi-location specialist mostly disappears — you are paying for machinery you do not yet need.
Then ask the question nobody asks: is this fee split with franchisees out of a co-op fund, or paid entirely by corporate? If it is co-op funded, adoption is voluntary in practice regardless of what the agreement says, and the agency's real job is persuasion. Firms that have worked in franchise systems will raise this before you do. Firms that have not will discover it in month three.
