Google Ads Customer Match now accepts IP addresses. Google's help page on creating a Customer Match list by uploading a data file now lists two new columns, "User IP address" and "User interaction timestamp", alongside email, phone and mailing address. Search Engine Land reported the change on September 25. If you upload customer lists to Google Ads, this could mean bigger audiences from data you already hold. It also means sending Google something it has never taken through this screen before: raw, unhashed IP addresses. That needs a decision, not a default.
What Google actually changed
Here is what the help page confirms, in plain terms:
- Two new match keys. You can add a customer's IP address and the time of their interaction to a Customer Match file. You can send an IP address with or without email or phone data.
- They go in unhashed. Email, phone and names still get SHA256 hashed before they reach Google. IP addresses and timestamps must be passed as plain text. Google's instruction is literally "don't hash".
- No matching for Europe. Google says IP matching "isn't supported for end users located in the EEA, UK, or Switzerland."
- A timestamp needs an IP. You cannot send a timestamp on its own. And if you send an IP address without a timestamp, Google says the system "defaults to the 'latest known' user of that IP."
That last point is the one most people will miss. Home and office IP addresses get reassigned and shared. An address you captured six months ago, uploaded without a time, gets matched to whoever Google last saw on it. That may not be your customer at all.
This is not brand new plumbing. PPC Land reports the same capability arrived in Google's Data Manager API in May, with Google saying IP data would help lift Customer Match match rates from the third quarter of 2026. What changed now is that anyone with a CSV can use it from the Audience manager screen.
The documentation is messy, so test before you trust it
Google's page says column headers must use "the exact names specified in English." Then it names the timestamp column three different ways on the same page: "User interaction timestamp", "User Interaction timestamp" and "User Engagement timestamp". It also describes an earliest and a last timestamp but gives only one column. PPC Land flagged both issues. Google has not clarified which spelling the uploader reads or how two timestamps fit in one column.
We are not going to guess. The practical answer is a small test upload before you rebuild any real list around this.
Why this matters for your ad account
Customer Match lists do real work in a Google Ads account. They let you exclude existing customers from prospecting, bid differently on past buyers, and give Performance Max and Demand Gen a signal about who your best customers look like. Small or poorly matched lists weaken all of that.
So a better match rate is worth having. Two cautions, though.
A higher match rate is not a better list. Google's own help page says a match rate "isn't an indicator of list performance." Matching more rows only helps if those rows are the right people. Bare IP addresses with no timestamp are the likeliest to match the wrong person.
IP data is a privacy question before it is a targeting question. Search Engine Land's read is that advertisers using the feature need to exclude EEA, UK and Swiss users and make sure their data collection "include[s] the appropriate disclosures and consent." Google's upload screen also asks you to confirm the data was collected in line with its Customer Match policies. Whether your privacy policy covers sharing IP addresses with Google is a question for whoever owns your legal and privacy setup. We are not lawyers, and the answer depends on where your customers are.
Who should bother
If you do not upload customer lists to Google Ads today, this changes nothing for you. Start with email and phone lists first. That is where most of the value is.
If you already upload lists and your match rates are low, or your list is mostly form leads with thin contact data, this is worth a careful test. That is common for service businesses and SaaS trials, where you often hold an IP and timestamp from a form submission but little else.
If a meaningful share of your customers are in Europe or the UK, the IP columns will not help for those customers and add work to keep them out.
What to do this week
1. Check what you already collect. Do your forms, CRM or store actually record the customer's IP address and the time of the interaction? Many setups store the IP but not a reliable timestamp. Without the timestamp, we would not upload the IP.
2. Check your privacy policy and consent flow. Before anything leaves your systems, confirm your policy says customer data, including IP addresses, may be shared with advertising partners like Google. If it does not, fix that first or skip the feature.
3. Filter out Europe, the UK and Switzerland. Google has not said how it handles rows for those users. Do not rely on it. Remove them from any file that carries IP data.
4. Run a small side-by-side test. Take one existing list. Upload a copy with email and phone only, and a copy with the same people plus IP and timestamp. Compare the matched size after processing, which Google says can take up to 48 hours. If the file errors on the timestamp header, try the other spellings from the help page and note which one works.
5. Judge it by performance, not match rate. If the bigger list goes into an exclusion or a bid adjustment, watch whether new customer acquisition cost actually improves over the next few weeks. If nothing moves, the extra data is not worth the privacy exposure.
6. Keep feeding the lists. A list uploaded once and forgotten decays. If you are going to use Customer Match properly, set up a regular refresh, or look at Google's own conversion-based customer lists, which build from your conversion data automatically.
The honest summary
This is a genuine new lever for accounts that upload customer data, and it could close a real gap for lead-gen businesses with thin records. It is also the first time Google has asked for raw IP addresses in the everyday upload screen, with documentation that contradicts itself in places. Use it on purpose, with timestamps, with Europe filtered out and your privacy policy checked. Do not use it just because the column is there.
Whether your first-party data is actually helping your bidding, or just sitting in a list nobody refreshes, is one of the things we check in our free 15-point Profit Audit: tracking accuracy, audience and exclusion setup, and where your reported numbers drift from real orders. Thirty minutes, and you keep the findings either way.