GBP Insights / Performance

What are GBP insights?

GBP insights — renamed "Performance" in the updated GBP interface, though most people still call it Insights and search for "GBP dashboard" when they mean the same screen — is Google Business Profile's built-in analytics section. It shows how customers find and interact with a business's profile: search impressions, calls, direction requests, website clicks, photo views, and a handful of other engagement metrics.

The mechanism: where the numbers actually come from

GBP insights pulls its data from the Google Business Profile Performance API, the same interface Google itself uses to populate the dashboard a business owner sees. That's worth knowing because it means any third-party tool showing "GBP insights" data is reading the same underlying Google numbers, not generating an independent measurement — the API is the single source, and every dashboard downstream of it, including Google's own, inherits the same lag and the same sampling behaviour described below.

Key metrics and what each one actually tells you

Search impressions count how many times the profile appeared in search results across Google Search and Maps combined, split into "direct" (someone searched for the business by name) and "discovery" (someone searched for a category or product and the profile appeared as a match). The ratio between the two tells you whether growth is coming from brand recognition or from category visibility.

Calls count phone calls initiated directly from the GBP — the call button on mobile, or the listed phone number on desktop.

Direction requests track how many times a user tapped "Get directions." This is arguably the strongest footfall-intent signal in the whole dataset; a spike here often shows up as a footfall increase in the following days.

Website clicks measure click-throughs to the website from the GBP panel. Usually lower volume than calls or directions, but still useful for tracking earlier-stage research behaviour rather than immediate-intent actions.

Photo views count how many times a profile's photos were viewed inside the GBP panel — a real engagement signal, and one that correlates with photo quantity and quality more directly than most other metrics do.

Booking clicks apply only to profiles with a booking link set up, tracking how often that link gets clicked from the GBP.

Reading each metric without overreacting to it

Every metric in GBP insights has a different amount of natural week-to-week noise, and treating all of them as equally sensitive is where a lot of businesses talk themselves into false alarms. Search impressions swing the most on their own, because impression volume is sensitive to broader seasonal search demand that has nothing to do with the listing itself — a 15-20% week-to-week move in impressions alone, with nothing else changing, is well within normal variance and not worth investigating on its own.

Calls and direction requests are steadier and more worth trusting when they move, precisely because they represent an actual decision a person made rather than a passive impression. A single day's spike or dip in either is still noise; a sustained move across two or three consecutive weeks is the point where it becomes worth asking why. Photo views tend to move in step with how recently new photos were uploaded rather than with anything competitive happening, so a dip right after a busy upload month is expected decay, not a warning sign. Website clicks are the metric most worth reading in combination with something else rather than alone — a click-through dip paired with flat impressions and flat calls usually just means the website link itself, or its placement, changed, not that overall interest in the business dropped.

The general rule: single-metric, single-day movements are almost always noise. Multi-metric, multi-week movements in the same direction are the signal worth acting on.

Why this data is not real-time, and why that matters

It's worth stating plainly: GBP insights is not a real-time analytics feed. Google's own performance data lags by roughly a month before it's fully reliable, and any platform, including Angryturtle's, that surfaces GBP insights is working within that same limitation, because the constraint sits in Google's API, not in any individual tool built on top of it. A business checking insights the day after a change won't see that change reflected yet; the honest expectation is a monthly view of trend, not a live dashboard a business owner watches change minute to minute.

What the reporting lag means for deciding whether a change worked

The practical consequence of the roughly month-long lag is that "did this work" is never a same-week question for anything measured through insights. A business that updates its categories, refreshes its products tab, or starts a posting cadence and checks insights three days later isn't seeing a null result because the change failed — it's seeing a null result because the data behind that view hasn't caught up yet. Drawing a conclusion that early, in either direction, is a mistake in the same way judging a single noisy metric movement is: the sample simply isn't there yet.

The workable discipline is to fix a measurement window before making the change, not after: decide in advance to compare the full month before a change against the full month starting roughly four to six weeks after it, giving the lag room to clear and giving the change itself time to actually influence behaviour. A business that instead checks daily and reacts to every wiggle in that gap ends up making decisions on data that isn't finished forming, which is a common way a genuinely working change gets abandoned too early because an impatient check in week two looked flat.

Why the numbers should be read with some caution

GBP insights data has known inaccuracies and relies on sample-based estimation for some metrics. Call data undercounts significantly whenever customers dial the listed number directly rather than tapping through the GBP call button — a common pattern once a customer has visited before and saved the number. That means insights is genuinely useful for trend analysis and relative month-to-month comparison, but it shouldn't be treated as an absolute count of every customer interaction that actually happened.

What's confirmed vs. what's assumed

Google documents the metric definitions and the general fact that some figures are estimated rather than exact counts; that part is confirmed in Google's own help documentation. What isn't published anywhere is the precise size of the undercount for calls, or an exact formula for how "impressions" get sampled at high volume. Any specific percentage claiming to quantify the gap between GBP insights numbers and real-world interactions is an estimate, not a Google-confirmed figure — the safe, honest framing is that the data is directionally reliable and numerically approximate.

Reading insights as a diagnostic, not just a report

The real value of insights isn't the raw numbers in isolation — it's what a sudden change signals. A steady month-over-month decline in direction requests, with impressions holding flat, usually points to something specific: a competitor pulling ahead in the local pack, a review-score drop changing which listing users trust enough to act on, or a profile issue Google hasn't surfaced clearly. Read in isolation, one month's numbers rarely tell you much. Read as a trend against the previous five or six months, they often flag a problem before it shows up anywhere else.

Worked example: a Hyderabad gym's insights trend

A multi-branch gym chain in Hyderabad noticed direction requests for one branch had fallen 30% over three months while impressions stayed roughly flat. On its own, that single-month number wouldn't have raised a flag. Trended against the prior six months, the decline was steady and specific to one branch out of five, which pointed away from a seasonal or citywide explanation and toward something local to that branch. The actual cause turned out to be a new competitor gym that had opened two streets away and quickly accumulated reviews — insights didn't diagnose the competitor directly, but the isolated, branch-specific pattern is what prompted the team to check nearby competitor activity in the first place. That kind of comparison is exactly what competitor benchmarking → is built to surface once insights has flagged where to look.

How to use insights to measure real progress

Treat any single month's numbers as a data point, not a verdict, and build a rolling six-month view instead. A good reading shows direction requests and calls tracking upward alongside a rising review count and improving rank — several signals moving together. A bad reading is a spike in impressions with flat or falling calls and directions, which usually means the profile is being seen more but converting less, often a sign that something in the listing itself — photos, reviews, category, hours — isn't matching what searchers expect once they look closer.

Common mistakes

The most common mistake is treating a single month's dip as an emergency without checking whether it's seasonal — many categories in India see real, recurring dips around specific festival periods that have nothing to do with GBP health. The second is ignoring photo views entirely because they seem like a soft metric; in practice they're one of the more reliable proxies for whether a profile's visual content is actually doing its job. A third is comparing raw numbers month to month without adjusting for known reporting lag, drawing conclusions from the most recent partial month of data before Google has finished counting it. A fourth, specific to the lag problem, is checking results too early after a change and concluding it failed, when the honest answer at that point is simply that there isn't enough settled data yet to say either way.

How Angryturtle uses GBP insights: monthly insights data is included in every client report, trended over six months rather than shown as a single snapshot. Unusual drops in call or direction volume are flagged for investigation as potential rank or profile issues before they compound into a bigger problem, within the same reporting-lag limits described above.

Related terms: GBP attributes → · GBP categories → · GBP API → · Local pack → · GBP messaging → · GBP posts → · Google Business Profile → · Rank OS →

More on this: Reading GBP performance data (learning centre) → · Local SEO measurement (learning centre) → · Product: Insights & Performance → · Product: Competitors → · Bulk operations & the GBP API (learning centre) → · GBP Management Services →