Tracking Locations & Franchises

Spyglasses can track AI visibility for each market a brand operates in, not just the brand as a whole. Use it when a single brand has multiple locations — a dental group with three offices, a national restaurant chain, an agency with regional desks, a retailer expanding into a new country — and you want to know whether AI assistants surface the brand in each city or country on its own.

What You'll Learn

• When to track by location (and when not to bother)
• How to add a location and pick the right granularity (City vs. Country)
• What's inherited from the parent brand vs. owned by each location
• How metrics roll up, compare across markets, and feed historical trends
• What a location report looks like compared to a brand report

When to Use Location Tracking

The core question to ask: does AI surface the brand differently in different markets? If yes, location tracking is the right tool. Typical cases:

• Multi-location professional services — a dental, legal, or medical practice with offices in three cities. Prospects searching from each city get city-flavored answers, and the brand's share of voice may be very different in Charleston vs. San Francisco.
• Franchise networks — a national chain wants to know where the franchise is winning the AI conversation and where local competitors are dominating.
• Regional retail or hospitality — a hotel group, a restaurant chain, a regional grocery brand. Local results are heavily influenced by review aggregators and city-specific sources.
• Geographic expansion — baselining AI visibility in a new market before launching, then watching it climb as the brand invests in local content and press.
• National vs. international — a brand that's strong in the US but unknown in the UK or Germany. Country-level tracking helps you measure each national rollout independently.

If your brand is B2B, B2C, a single-location, or doesn't have a geography to compare against, a regular brand property is enough. The parent's brand snapshot and metrics tell the whole story.

How Locations Relate to the Parent Brand

A location is a sub-property of a parent brand, not a stand-alone property. The parent owns the brand identity; the location is a geographic slice of it.

This matters because of what each side owns:

The brand-identity rule is the most important: even on a location's runs, brand mentions are always matched against the parent's company name and aliases — never against the location's label. A location labeled "Palm Beach Gardens" still counts mentions of Anchor Bank, not mentions of Palm Beach Gardens. The label is a navigation aid for you; it's not a brand the LLM is being asked to recognize.

This is what makes share of voice comparable across markets: the same brand is being tested in each one, just with the location context injected into the prompts.

Adding a Location

From the parent brand's sidebar, click Locations, then Add location. The dialog asks for two things:

• Location (required) — search and pick a place. Toggle City ↔ Country to switch the granularity:
  • Use City for local businesses where prospects search with neighborhood- or city-level intent.
  • Use Country when local results vary little within the country and you mostly care about national prompts ("the best banks in the UK").
• Label — defaults to the city or country name; customize it if you want something more meaningful (e.g. "Downtown" or "HQ" instead of the city). The label is shown in lists and on the location's dashboard; it does not affect brand-mention matching.

Once created, the location is wired into the nightly prompt-run pipeline automatically. The first metric data points appear after the next nightly run completes — usually within 24 hours — or you can trigger an AI Visibility Report on the location to populate data immediately.

For details on the dashboard layout, the per-row SOV and mention counts, and where the management actions live, see the Locations dashboard guide.

How Prompts Run for Each Location

Each nightly run fans out the parent's canonical prompts once per location. The same prompt text runs for each market, with the location injected two ways:

• The location's city / region / country is passed as runtime context to every LLM call, so the answer is shaped by local relevance.
• Any explicit {location} placeholder in a prompt's text is substituted with the location string before the prompt is sent.

This is what makes the comparison fair. You're not asking different questions of different markets — you're asking the same question of each, and seeing how the answer changes. Resist the urge to write location-specific prompts; keep the canonical set on the parent and let the injection do the work.

A brand with five locations and ten active discovery prompts will run (5 × 10) + 10 = 60 prompts per night — fifty location-scoped runs plus the parent's brand-wide rollup. Each generates its own metric row, so the daily storage cost scales linearly with locations. At dozens of locations this is fine; at hundreds it's worth a conversation about which prompts really need per-location runs.

Reading Location Metrics

Location data shows up in three places:

1. The parent brand's main dashboard

When a brand has at least one location, the parent's dashboard replaces the brand-wide Share of Voice widget with a Share of Voice by Location widget. Three at-a-glance tiles plus a ranked list:

• Average SOV — equal-weighted across all locations with data ("typical location" — not skewed by big-traffic markets)
• Total Mentions — summed across locations
• Total Citations — summed across locations
• Ranked location list — each market's latest SOV, mentions, and citations; sortable by Highest SOV, Lowest SOV, Most Mentions, or Name

This view is what most teams use day-to-day: a quick scan to see which markets are winning, which are lagging, and which haven't started accruing data yet.

2. The Locations dashboard

The dedicated Locations dashboard is both the management interface (add, open, delete locations) and an at-a-glance list with SOV and Mentions displayed inline for each row.

3. Each location's own dashboard

Drilling into a single location gives you a focused view scoped to that market — its own SOV widget, its own competitor list, its own Historical Metrics, and the ability to run a dedicated AI Visibility Report for that location.

The historical metrics dashboard on a location shows only that location's series, so a sustained downward trend on Boston really does mean Boston is slipping — not that the brand as a whole is slipping. See the Historical Metrics docs for a deeper read on how the data is scoped.

How a Location Report Differs from a Brand Report

You can run a regular AI Visibility Report on any location, just like on a brand. A location report:

• Uses the parent's brand snapshot for matching — the snapshot card on the report shows the parent brand's identity, with the location label appended in the report title (e.g. "Anchor Bank: Palm Beach Gardens AI Visibility Report").
• Uses the parent's discovery prompts — the canonical set, executed with the location injected.
• Uses the location's own competitors — not the parent's. This is intentional: local SOV is interesting precisely because the competitive set differs.
• Skips the brand consistency block — location reports run in discovery-only mode (no separate consistency analysis), and the Brand Consistency tile is hidden on the report.
• Skips the Grounding Searches block — SERP gap analysis is keyed to a brand domain and isn't run per-location.
• Stores metrics that roll up to the parent's per-location widget — so an ad-hoc report on a location feeds the same dashboards as the nightly pipeline.

Everything else — share of voice, brand mentions, citations, the per-platform tabs, the public report link — works exactly like a brand report.

Practical Tips

• Start with City for local businesses, Country for national chains. City-level produces the most actionable insights when prospects search with local intent. Country-level reduces noise when local results don't vary much within a country.
• Always set competitors per location. The default empty list won't surface useful SOV comparisons. Take ten minutes when you add a market to research the local players that AI is actually mentioning, and add them.
• Use the parent for prompt management. Locations don't have their own prompt sets — you manage one canonical set on the parent and every location runs them. If you find yourself wanting a different prompt for one market, ask whether it's really different or just phrased differently; if it's the latter, tag the prompt and slice by tag in Historical Metrics instead.
• Baseline before you launch. If you're about to enter a new market, add the location before the launch and let nightly runs collect a few days of pre-launch data. Then you'll have a real baseline to measure post-launch lift against.
• Use the rank order to focus attention. The Share of Voice by Location widget is sorted by SOV; the bottom of the list is where intervention opportunity is. Sort by Most Mentions to find markets that should be winning given the chatter, but aren't converting mentions into SOV (often a sign of strong local competitors capturing the recommendation).
• Watch the historical chart on a struggling location, not the parent's. Brand-wide trends can mask divergence between markets — a flat parent average can hide one market falling while another rises.

Where to Go Next

• Locations dashboard — the dashboard-level guide: page layout, adding locations, what each column means
• Historical Metrics — how trend data is scoped for locations
• Competitors — managing each location's local competitor list
• Discovery Queries — the canonical prompt set every location inherits from the parent