Store Visit Conversions in Google Ads 2026: Measuring the Offline Effect of Ads

For any business with physical locations, the Google Ads report understates reality. Someone sees an ad on their phone, walks into the shop two days later and pays at the till — in the account, that is a click with no conversion. As long as a meaningful share of revenue lives offline, every bidding decision is being made on half the data — and store visit conversions are the report built to close that gap.

Store visit conversions close part of that gap: Google models how many people who interacted with your ads subsequently walked into your location. This guide covers how the model works, what an account needs before the report appears at all, how to read the numbers without fooling yourself, and when it is safe to bid on foot traffic.

What store visit conversions are and how they are measured

The mechanism has three links:

1. A signed-in user with location history enabled interacts with your ad — a click or an impression.
2. Some time later they physically enter your location.
3. Google connects the two events and extrapolates from anonymous, aggregated statistics to represent the broader population.

That third step is the important one. Store visits are not a count of individual people. They are an estimate built from the subset of users whose location data is available, then scaled up. Google validates the model with survey panels, including its Opinion Rewards audience.

Store visits are a statistical estimate, not a visitor log. Trust them as a trend and as a basis for comparing campaigns against each other — not as a customer list you can reconcile against the till, line by line.

Two practical rules follow from that. First, recent days are unreliable: the model recalculates continuously, and the last five days or so may show zero or heavily understated figures. Second, a visit is attributed to the date of the ad interaction, not the date of the visit. That is standard Google Ads conversion logic, but it reliably causes arguments with offline managers.

What your account needs before the report appears

Google does not publish exact thresholds and states plainly that it cannot guarantee eligibility for any specific account. The requirements themselves are known:

• Physical locations in countries where store visits reporting is offered. Coverage is incomplete — check this first.
• Verified Business Profile locations — all of them, not a subset, or visits to uncovered addresses simply will not be counted.
• Location assets linked to campaigns (or affiliate location assets, if you are a manufacturer sold through retailers).
• Enough ad clicks and impressions, and enough foot traffic, to clear privacy thresholds.
• No sensitive locations — addresses connected with religion, health or children are excluded.
• You cannot bid on store visits and store sales simultaneously.

A useful detail: a disapproved location asset does not block attribution. Google continues to attribute visits even when the asset itself is not serving. Asset policy issues and reporting issues live in different places.

Once an account qualifies, a “Store visits” conversion action is created automatically and appears under “All conversions”, including view-through conversions. Nothing has to be switched on — which also means the number can appear unexpectedly and distort period comparisons you were used to.

Where visits show up, and how not to confuse them with everything else

The most common interpretation error is adding visits to online leads and celebrating a doubled conversion count. Two reasons not to: a visit is not a sale, and visits are modelled rather than recorded.

Question	Store visits	Online conversions
Nature of the data	Model and extrapolation	Recorded event
Date attribution	Date of click/impression	Date of click (standard reporting)
Freshness	Roughly five-day lag	Minutes to hours
Value	You have to define it	Usually known
Default column	“All conversions”	Configurable

The practical approach: keep visits out of the “Conversions” column — the one bidding optimises toward — until you have done three things: valued a visit, confirmed the data is stable, and decided on a strategy. The mechanics of that split are covered in primary and secondary conversions.

Putting a value on a single visit

Without a monetary value, a visit is a pleasant number you cannot act on. The working formula:

Visit value = average transaction value × walk-in-to-buyer rate × margin

Where each input comes from:

• average transaction value — from the POS system for the same period;
• walk-in-to-buyer rate — from door counters if you have them; if not, use a conservative estimate and record it explicitly as an assumption;
• margin — from finance, not from memory.

An illustrative example: £60 average basket, 30% of entrants buy something, 35% margin. Visit value = 60 × 0.3 × 0.35 = £6.30. That is your ceiling for what a visit is worth paying for in advertising.

One caveat: not all of that value is created by advertising. Some of those people would have walked in anyway. Establishing the real contribution requires comparative measurement — the strategic framing is in local advertising and geotargeting as strategy, and the rigorous method is geo-experimentation.

When it is safe to bid on visits

Smart Bidding supports store visits across several campaign types:

Campaign type	Supported strategies
Search	Multiple strategies, including enhanced CPC
Performance Max	Target ROAS, Target CPA, Maximise conversions
Demand Gen	Maximise conversions, Target CPA
Shopping	Target ROAS, Maximise conversion value

The conditions that make it worth doing:

1. Visit volume has been stable for at least two months. The model should produce predictable numbers, not swing threefold week to week.
2. Visit value is calculated and agreed with finance. Otherwise the algorithm optimises an abstraction.
3. Visits are a material share of total outcome. If offline represents 5% of results, the risk of destabilising working campaigns outweighs the gain.
4. You are prepared for a learning period. Adding a new conversion action to optimisation is a material edit — see the Smart Bidding learning period.

One warning specific to this data: with roughly a five-day lag, the strategy is learning from an incomplete picture of recent days. That is another reason not to react to apparent dips — the same logic as in conversion windows and conversion lag.

Setup, in order

Step 1. Clean up the Business Profile

Every location verified, addresses and hours current, duplicates removed. A single unverified location is a hole in your reporting that you will never be told about.

Step 2. Link the profile and enable location assets

Assets must be attached to the campaigns whose effect you want to measure. If the asset sits on one campaign only, visits driven by the others stay invisible.

Step 3. Check location targeting

More campaigns break here than anyone expects. “Presence or interest” brings in people who are physically far from your location and cannot walk in. For offline objectives, “presence” is almost always correct. The difference is unpacked in presence versus presence-or-interest targeting.

Step 4. Wait for the conversion action to appear

It is created automatically once thresholds are met. If it has not shown up after several weeks, the cause is usually one of three: insufficient click volume, unverified locations, or an unsupported country.

Step 5. Segment before drawing conclusions

Look at visits by device, hour of day and targeting radius. That is where the insights live: mobile within 3 km drives visits while desktop at 20 km does not. Scheduling decisions follow from there — see ad scheduling and dayparting.

Step 6. Reconcile with the till monthly

Compare the trend in visits against the trend in store revenue. You are not looking for matching absolute numbers — you will not get them — but for correlated direction. If spend rises, visits rise, and revenue does not, the problem is in the store or the range, not the advertising.

Local actions: the signal every account can have

Store visits have a younger sibling — local actions. These are interactions a user takes in the ad or the business listing, recorded directly with no modelling involved:

• clicking the phone number;
• requesting directions;
• clicking the address or opening the location page;
• clicking through to the site from the location listing;
• menu, order or booking clicks where those are enabled.

Three reasons to start here:

1. No thresholds. Local actions are available to small businesses that will never see store visits reporting.
2. Data arrives immediately. No five-day lag, so decisions can be made while the week is still live.
3. They are recorded, not estimated. A directions click either happened or it did not.

The limitation is equally obvious: intent is not arrival. Someone can request directions and never set off. The sensible construction is to treat local actions as the operational indicator and visits as the check on what that intent converts into. If direction requests climb and visits do not, the question is no longer about advertising — look at parking, opening hours, queues, and whether the listed address is even right.

Calls deserve separate mention. For many offline businesses a call is worth more than a visit, and it is also the thing most often missing from reporting. The measurement mechanics are covered in call tracking and call campaigns in Google Ads.

Radius, device and time: where the money hides in visit data

An aggregate visit count is nearly useless for management. The data becomes actionable in three cuts.

Radius

Plot visits against targeting radius. The typical retail pattern: most visits fall within 3–5 km, and beyond 10 km the cost per visit multiplies. The conclusion is not “switch off distant zones” but “split them into separate campaigns with their own bids and their own creative” — because someone 15 km away needs a different reason to travel than someone three streets over.

Device

Mobile almost always produces more visits per click: a higher share of near-me intent and people already in motion. Just calibrate any bid-adjustment conclusions against how those adjustments actually function in modern strategies — see device bid adjustments.

Time

Visits tolerate out-of-hours impressions poorly: an 11pm click rarely becomes a next-day visit. Hour-of-day and day-of-week data translate directly into a schedule — one of the few cases where dayparting delivers a measurable result rather than a feeling of control.

One cut people routinely skip is the individual location. A network average conceals the situation where three stores take half the visits and five produce nothing. Where that pattern exists, the issue is not advertising but the business: siting, signage, range, opening hours.

What visits will never tell you

• Who the customers are. It is an aggregate model, not a CRM.
• Revenue. That requires uploading actual offline sales — the general pattern is covered in offline conversion import.
• Causation. Visits correlate with advertising; they do not prove the person would have stayed home without it.
• Anything about small locations. Privacy thresholds exclude low-traffic sites, so those report nothing even when visits happen.

If offline revenue matters to your business, the honest stack is: visits for day-to-day bidding and campaign comparison, offline sales upload for the money, geo-experiments for proof of incrementality. Three tools answering three different questions.

Common mistakes

1. Presenting visits as sales. The most expensive mistake in any board report.
2. Comparing the last few days to earlier periods. The model lag makes recent days look like a collapse.
3. Adding visits to core optimisation immediately. Observe and value first; bid second.
4. Never segmenting by location. A network average hides that three stores out of twenty produce half the visits.
5. Ignoring seasonality. Foot traffic depends on weather, holidays and local events far more than online traffic does.
6. Leaving “presence or interest” on. A direct route to expensive clicks with no chance of a visit.
7. Forgetting product data. For retail, visits are tightly coupled to local inventory — stale availability data sends people to an empty shelf. See local inventory ads on by default in Shopping.

Fitting visits into the wider measurement stack

Visits are one of four layers of offline measurement; each is more accurate and more expensive to implement than the last.

Layer	What it gives you	Implementation cost
Local actions (calls, directions, address clicks)	Intent signal, available immediately	Low
Store visits	Modelled estimate of arrivals, automatic	Low, once thresholds are met
Offline sales upload	Actual revenue per transaction	Medium: needs CRM/POS data
Geo-experiments	Proven incremental lift	High: requires withholding traffic

Implement top-down. Start with local actions and calls (see call tracking and call campaigns), add visits, then feed real sales back, and only then invest in proving incrementality. The reverse order usually produces an elaborate project nobody uses.

One more context: if you run promotions and demand spikes, visits are a good way to see the offline echo that online conversions miss entirely. Planning those windows in the account is covered in promotion mode in Google Ads — and before any peak it is worth confirming your landing pages are reachable, which is the subject of AdsBot and landing page crawlability.

If account infrastructure for this kind of volume is still an open question, see Google Ads agency account rental and the other PPC Rebels services.

FAQ

Why does my account have no store visits reporting?

Usually one of three reasons: an unsupported country, unverified Business Profile locations, or click and visit volume below privacy thresholds. Google states explicitly that it cannot guarantee eligibility for any given account.

Can I switch the report on manually?

No. The “Store visits” conversion action is created automatically when an account qualifies. You can only influence the inputs: verify locations, attach location assets, grow click volume.

Why do the last few days show zero visits?

That is the normal model lag — roughly the past five days are still being processed. Judge the period only after it has matured.

Which date does a visit belong to?

The date of the ad interaction — the click or impression — not the date of the visit itself. Standard Google Ads attribution behaviour.

Are visits deduplicated across accounts in a manager structure?

If conversion tracking is configured at the manager account level, store visits are deduplicated across client accounts. Each sub-account still has to meet the eligibility requirements independently.

Can I bid on store visits and store sales at the same time?

No. They are mutually exclusive — an account has to choose one.

Does a disapproved location asset break visit attribution?

No. A disapproved asset will not serve in ads, but it does not affect Google’s ability to attribute visits.

How accurate are the numbers?

They are a modelled estimate built from users with available location data and extrapolated to the wider population, validated through survey panels. Use them for comparison and trend, not as an exact headcount.

What if my locations are too small to report?

Privacy thresholds exclude low volumes and there is no way around them. For small sites, lean on local actions instead — calls, direction requests, address clicks are recorded directly.

What value should I assign to a visit?

The one your own formula produces — average basket × purchase rate × margin — not a market average. Recalculate it at least quarterly; when the basket changes, so does the threshold.

Does this work if I do not own stores and sell through retailers?

Yes — affiliate location assets exist for exactly that case, letting a manufacturer measure visits to the retail locations that stock its products.

How do I prove to leadership that ads drive footfall?

Visits alone will not do it, because they show correlation. The proof is a geo-experiment: hold advertising out of some regions temporarily and compare the difference against a control group.