Reading Google Business Profile Insights: The Numbers That Matter
A café owner opens her profile's performance tab, sees "1,200 interactions this month", feels great, and closes the tab. Two problems: the number is a sum of unlike things (a phone call and a photo view counted as neighbors), and buried in the same report — unread — was the fact that direction requests have been flat for four months while a competitor's queue grows.
Google Business Profile Insights is the free analytics of your Maps presence: what people did after finding you on Google. It's genuinely valuable — and genuinely misread, because the dashboard leads with vanity aggregates and hides the four numbers that tell a story. This post is about reading it the way a practitioner reads it.
Google Business Profile Insights shows what customers did after finding your business on Google: the four metrics that matter are phone calls, direction requests, website clicks, and bookings/conversations — each indicating intent, from research (clicks) to planning a visit (directions) to acting (calls). Read them month-over-month rather than as isolated totals, always alongside the search-phrases report that shows which queries found you. The "total interactions" headline mixes photo views with calls and hides the trend; behavior data pairs with grid-scan visibility data to form the full picture. :::
The Four Numbers That Actually Matter
Everything in insights is behavior after discovery. The four metrics form an intent ladder:
| Metric | What the customer is doing | Intent stage |
|---|---|---|
| Website clicks | Researching you on your own site | Curious |
| Direction requests | Planning the trip to you | Deciding |
| Phone calls | Asking something specific — hours, price, availability | Acting |
| Bookings / conversations | Committing through Google's tools directly | Committed |
The ladder is the analysis. Healthy patterns look like a funnel: plenty of clicks, a solid share of directions, calls converting at the bottom. Unhealthy patterns diagnose themselves:
- Many clicks, few direction requests — people look and don't come. Usually a profile-to-reality gap: photos overselling, reviews mentioning problems, prices hidden. Your website's job is converting curious into visitors; check what page they land on.
- Directions but no calls — normal for walk-in businesses (cafés, convenience), a warning for appointment businesses (clinics booking nothing but mapping plenty).
- Calls but no directions — service-area businesses answering by phone; healthy. For a restaurant, it means people are asking questions your profile should answer (hours, menu, parking).
- Everything tiny — a visibility problem, not a conversion problem. That's the grid-scan's territory (see below).
The Search-Phrases Report: Free Keyword Data You're Ignoring
The most under-read section of insights: which queries people used when your profile appeared. It splits into discovery searches (generic — "coffee shop Ari") and direct searches (your name or number).
Three honest uses:
- Know your real keywords. The phrases that actually surface you are your proven market. If "ร้านกาแฟ อารีย์" appears weekly and "coffee ari" never does, you've learned which language your discovery runs in — for free, no tool.
- Find missing phrases. The phrases you WISH appeared and don't are your to-do list: they're what customers search and competitors answer. Cross-reference against your keyword research.
- Watch for brand-name dominance. If most of your "discovery" is actually your own name in slight variants, your reach is existing customers — new-customer acquisition is the weak flank.
Monthly 5-minute read:
1. Four metrics, month vs previous month
2. Top 10 discovery phrases — any new entries? any missing?
3. Direction requests by day — when do people plan to visit?
One caution: search-phrase data is directional, not exhaustive — Google samples and anonymizes. Treat it as signposts, not census.
Trends Beat Snapshots (and the Dashboard Knows It)
The dashboard's default views are cherry-pickers: short ranges, flattering comparisons, aggregates. Defend yourself with discipline:
- Always compare month-over-month, and for seasonality, the same month last year. A Bangkok restaurant's October is not its December.
- Ignore "total interactions" as a headline. A photo-view spike (you uploaded an album) can double it while calls fall. The aggregate hides the composition.
- Beware week-vs-week noise. The local pack wobbles; so does behavior. Monthly cadence filters the weather out.
- Screenshot or export monthly. The dashboard keeps limited history; your archive is the real dataset. A simple monthly note — four numbers, top phrases — becomes your business's local-search ledger.
What Insights Can't Tell You
Honest limits, so you don't over-read:
- Whether you ranked well. Insights shows behavior AFTER appearance, not how often or where you appeared. Two profiles with identical insights can have very different visibility. That's what grid scans measure.
- Whether behavior converted. A direction request isn't a sale; a call isn't a booking. The last mile lives in your POS, your booking system, or a counter question.
- Why anything happened. Insights reports what, never why. Attribution is your job: pair each movement with what changed that month (new photos, review surge, competitor closure, category edit).
The Pairing: Insights + Grid Scan
The full local picture needs both instruments:
Visibility (grid scan) Behavior (insights) Diagnosis
───────────────────── ─────────────────── ─────────
strong footprint strong behavior healthy — grow outward
strong footprint weak behavior conversion problem — fix photos/reviews/first impressions
weak footprint strong behavior beloved local — expand reach; you're invisible where you don't rank
weak footprint weak behavior fundamentals first — categories, NAP, reviews
That 2×2 is the entire monthly review for most businesses: one grid scan, one insights read, one decision about what to work on this month. The strong-visibility/weak-behavior quadrant is the one people skip — they keep chasing rank when the profile itself is leaking customers.
The newer signals: bookings, conversations, and device
If you've enabled booking links or chat through Google, insights adds bookings started and conversations — the most literal conversion numbers in the dashboard, and worth enabling if your business takes appointments. And glance at the device split occasionally: in Bangkok, a profile whose direction requests come overwhelmingly from mobile (nearly all of them, for most businesses) reminds you where to spend effort — tap-to-call working, hours legible on a small screen, photos that read at thumbnail size. A beautiful desktop experience nobody reaches from Maps is effort spent on the wrong screen.
A Workable Monthly Ritual
- First of the month, 15 minutes, calendar-recurring.
- Record the four metrics + top 10 discovery phrases (your ledger).
- Run the same grid scan as always — same points, same phrases, both languages.
- Place yourself in the 2×2. Pick ONE workstream for the month based on the quadrant.
- Note what you shipped last month next to the movement — over a quarter, the cause-effect becomes visible.
Frequently Asked Questions
Are photo views and "interactions" worth tracking at all?
Individually, rarely — they're the most inflated numbers in the dashboard. The one honest use: photo views as a leading indicator after you upload new photos, and as a comparison against businesses with similar foot traffic. If the dashboard shows a view surge with no movement in the four real metrics, treat it as decoration, not progress.
Why did my calls drop while everything else held?
Check the phone number itself first — a misformatted or outdated number on the profile breaks the tap-to-call silently. Then check patterns: drops clustered on certain days often mean hours or staffing changes, not search changes. Only after ruling out the mechanical should you look at search-side explanations like a new competitor absorbing calls.
How accurate is the direction-requests metric?
It counts people who requested directions to your pin — a strong visit-intent signal, but not proof of arrival. Treat it as the best available proxy for "planned to come." Its trend matters more than its level: a 30% sustained rise in directions with stable everything-else is usually real foot-traffic movement.
Can I see insights for multiple locations together?
Each location's profile reports its own insights; there's no honest aggregate across branches — and per the multi-location rules, you shouldn't want one. Read each branch separately on the same day each month; the comparison between branches is for learning patterns, not averaging results.
Insights is the "what did customers do" half of your local search picture; the grid scan is the "where could they find us" half. Read both monthly, decide once, and you're running local SEO with better data than most agencies report.
Start the visibility half right now: run a free grid scan at gbppeak.com/free-maps — no signup — then pair it with this month's insights read and place yourself in the 2×2.
Related posts
One Brand, Ten Locations: Managing Multiple Google Business Profiles
Three branches sharing one phone number — and Sathorn's reviews landing on the Silom profile. Multi-location local SEO fails structurally; here's the structure that works.
NAP Consistency: Why Your Shop's Phone Number Is Quietly Costing You Rankings
Old address on Facebook, 2019 phone number in a dozen Thai directories — conflicting listings quietly cap your local rankings. The one-hour audit that fixes them.
Local Keyword Research for Bangkok: Finding the Words Customers Actually Type
"Dental clinic Bangkok" is a knife fight with hospital groups. The customers you want type service + neighborhood, in Thai AND English — here's how to find and win those phrases.



