How to Find Newly Opened Businesses Without an Opening Date
A business that has just opened is the best prospect a local service provider can get. No suppliers locked in, the owner still answers the phone, and they are spending on everything at once.
The problem is that you cannot filter by opening date, because that data does not exist.
Google does not record when a business opened
The Places API does have a field called openingDate. It looks like the answer until you read the small print: according to Google's own documentation, it is only populated when the business status is FUTURE_OPENING and the anticipated opening is fewer than ninety days away.
In other words, it is a "has not opened yet" field. The day the shop raises its shutters, the status flips to operational and the date disappears. Nothing records when it happened.
This is worth knowing before you buy anything, because any tool selling you a "businesses opened in the last six months" filter is using a proxy, not a data field. The useful question is not which one has the data. It is which proxy each one uses and how that proxy breaks.
The most common proxy is the one that works worst
The usual shortcut is "few or no reviews". It sounds reasonable. A place that opened last month has not had time to accumulate any.
Here is what happens when you check it against real data. These are the shares of listings with no reviews at all, across a sample of 3,147 local businesses in twenty countries.
| Country | Sample | No reviews at all | Median reviews |
|---|---|---|---|
| Argentina | 155 | 45.2% | 2 |
| Germany | 150 | 44.0% | 9.5 |
| Canada | 165 | 39.4% | 33 |
| United Kingdom | 150 | 32.7% | 106 |
| Poland | 159 | 28.9% | 36 |
| United Arab Emirates | 198 | 28.8% | 177.5 |
| Italy | 138 | 23.2% | 66.5 |
| South Africa | 300 | 23.0% | 13 |
| Spain | 111 | 22.5% | 79 |
| Portugal | 149 | 16.8% | 124 |
| Türkiye | 196 | 13.3% | 157 |
| Mexico | 143 | 11.2% | 20 |
| United States | 175 | 9.7% | 319 |
| India | 150 | 4.7% | 176.5 |
| France | 158 | 4.4% | 64.5 |
| Brazil | 146 | 4.1% | 148.5 |
Across the whole sample, 20.7% of listings carry no reviews. But the range runs from 4.1% in Brazil to 45.2% in Argentina. Eleven times.
No business formation rate looks like that. Germany does not open 44.0% of its businesses every season. What the signal measures is something else: the local habit of leaving reviews. Where people review rarely, half the market looks freshly opened.
The consequence is simple and it changes the method: the threshold has to be relative to your market, not absolute. In Argentina, zero reviews is the default state and tells you nothing. In Brazil or France, zero reviews is a genuine anomaly worth a visit.
You can check the numbers in the review benchmark data by country.
The second trap: old listings have zero reviews too
Even after you correct for the local baseline, one problem remains that almost nobody mentions. The zero-review band does not only contain recent openings. It also contains abandoned listings, duplicates, and businesses that closed without anyone updating the record.
We saw it clearly in a sample of 975 Madrid restaurants. There the no-website share falls cleanly as reviews rise, and the zero band breaks the staircase:
| Reviews | Places | No website | 95% CI | Has own domain |
|---|---|---|---|---|
| None | 11 | 36.4% | 15.2 – 64.6 | 54.5% |
| 1 to 49 | 56 | 75.0% | 62.3 – 84.5 | 19.6% |
| 50 to 199 | 124 | 52.4% | 43.7 – 61.0 | 37.1% |
| 200 to 999 | 385 | 33.8% | 29.2 – 38.6 | 57.4% |
| 1000 or more | 399 | 8.5% | 6.2 – 11.7 | 89.5% |
Look at the last column of the first row. 54.5% of the places with no reviews already had their own domain, more than the 1 to 49 band and more than the 50 to 199 band. A restaurant that opened last week rarely has a registered, configured domain already.
That band is a mixture, and its confidence interval says so: 15.2% to 64.6%. Nothing can be claimed at that width. The zero-review band is not the band of new businesses. It is the band of businesses we know nothing about.
The method that works: compare two runs
The only honest approximation of "newly opened" is "newly appeared in the index". And that one can be measured.
- Run today's extraction. Fixed category and city, with the parameters you intend to repeat.
- Keep the
place_idcolumn. It is the stable identifier of each listing. It ships in the REST API and MCP connector results. It is not in the CSV or the Excel file, which carry twenty-two columns aimed at a CRM import, so this part is an API workflow. - Repeat the same search a month later. Same parameters, nothing touched.
- Diff the two sets. The
place_idvalues present in the second run and absent from the first are index additions. That is your list. - Filter the result. Drop the ones that already carry a lot of reviews, because those are not new: they are listings your first search failed to reach. Keep the ones below your country's baseline.
- Check
business_status. It is useful for removing closures, not for finding openings.
The monthly cadence is not arbitrary. The shared result pool refreshes listings inside a freshness window, so at a month's distance you are comparing re-read data rather than a stored copy. At a year's distance the noise of openings, closures and renames swallows the signal.
If you would rather stay on the review side without building the comparison, the approach and its limits are in how to find businesses with few reviews. The starting point for the extraction itself is how to get a list of businesses in any area.
What a newly opened business is actually worth
It pays to be realistic before building the process.
Go back to the Madrid table. The 1 to 49 review band, the closest thing to a young venue, was 75.0% without a website. That is the best number in the whole table for anyone selling digital presence.
But the same band has the worst contactability in the table: 25.0% had no phone number on the listing, against 1.0% in the band above a thousand reviews. One in four prospects in your best segment cannot be reached at all.
And there is a third factor that appears in no column: a business with a median of twenty-six reviews is not turning over much. Maximum need, minimum budget. Which is why the band that usually closes is not the newest one but the next one up, 50 to 199 reviews, where there is cash coming in and still nobody running the digital side.
The full analysis of that relationship is in restaurants without websites, and the sales script for that profile is in how to sell web design to local businesses.
A process that holds up
Putting it together, the recipe that works has three conditions at once, not one:
- It appeared in the index since the last pass. That is the only genuine novelty signal you can measure.
- It has fewer reviews than the median for its country. Use the table above, not a round number.
- It has a phone. Without one there is no conversation and it does not belong on the list.
With those three, a monthly pass over one category and one city gives you a short, defensible list that your competitors do not have, because most of them stopped at the review filter. To decide which category and city are worth setting this up for, local market research with Google Maps is the step before.
FAQ
Does Google Maps store a business opening date?
Not the one you need. The Places API has an openingDate field, but it is only populated when the business status is FUTURE_OPENING and the opening is fewer than ninety days away. It is a "has not opened yet" field. The moment the place opens, the value disappears and no record of the actual opening date remains.
Is the review count a reliable sign of a new business?
Only relative to the local baseline. Across our sample of 3,147 listings, the share with no reviews at all runs from 4.1% in Brazil to 45.2% in Argentina. That is an elevenfold spread, and no business formation rate explains it. The signal measures local reviewing habits, not the age of the business.
So how do you actually find them?
By running the same search twice a month apart and comparing the place_id field. Anything present in the second run and absent from the first is new to the index, which is the closest honest approximation of newly opened. The place_id ships in the REST API and MCP connector results, not in the CSV.
How often should the extraction be repeated?
Once a month. The shared result pool refreshes listings inside a freshness window, so a monthly comparison works on re-read data rather than a stored copy. An annual comparison is useless: at that distance the noise of openings and closures swallows the signal.
Is a newly opened business a good prospect?
It is the group that lacks the most and the one that is hardest to reach, at the same time. In our sample of 975 Madrid restaurants, the 1 to 49 review band was 75.0% without a website, but also 25.0% without a phone number on the listing, against 1.0% in the band above a thousand reviews. Maximum need, minimum contactability.
Getting started
Extract one category and one city today, keep the identifiers, and set a reminder for thirty days out. The second pass will give you something no review filter can: the difference.
Create an account and build it with 100 free credits every month, no card required.