Ads Warmup: Warm Up Your Ads Before You Spend

The Ads Warmup guide: when to use it and how to get a good match

Every ad platform learns from conversions. When a campaign launches on a new account, a new Pixel or a new offer, the platform has very little to learn from. It shows your ads to a wide mix of people, watches who converts, and slowly narrows in. You pay for that learning. Ads Warmup gives the platform a head start by sending it people who already bought from you, signed up, or booked a call, so the first conversions it sees are real ones.

This guide covers when Ads Warmup is worth using, how to prepare your file, how to read the match score, what happens on each platform, and the honest limits you should know before you send anything.

When Ads Warmup makes sense

Ads Warmup is most useful at the start of something. A brand new ad account has no history. A new Pixel on a rebuilt website has no events yet. An agency taking over a client often inherits an account with little clean conversion data. In each case you probably already have a customer list sitting in a spreadsheet, a store export or a CRM. That list is exactly what the platform is missing.

It also helps when you launch a new event. If you start optimising for Schedule because you now sell booked calls, the platform has never seen a Schedule conversion from you. Sending a list of people who booked in the past gives it examples to work from.

It is less useful when an account already sends plenty of live conversions every day. The platform has signal already, and a one time upload adds less on top. Ads Warmup is a starting push, not a replacement for tracking live conversions as they happen.

How to prepare your CSV

The only column you must have is email. DataCops reads your column names for you, so you do not need to rename anything or follow a template. You can export straight from your store, your CRM or your booking tool and upload the file as it is. Each upload can hold up to 20,000 rows. If your list is bigger, split it into more than one file.

Every extra column that identifies a person makes it easier for the platforms to find them. The columns that raise the match score are:

• Phone number. Gives the platform a second way to find the person besides email.
• First and last name. Helps the platform confirm a match it is unsure about.
• Country, city or postcode. Narrows the match to the right person when names are common.
• Click ID. If your export kept an ad click ID, it links the person straight to a past ad click.
• Your customer ID. Sent as your own ID for the person, which adds to the match.

Price does not change the match score. If your file has an order value, it is sent with a Purchase so the platform learns what your buyers are worth. If your file has no value, none is sent. DataCops never makes up a price or a location to fill a gap.

Reading the match score

After upload, every person gets a score from 0 to 10 before anything is sent. The score is an estimate of how well the platforms can match that person, based on how much identifying data the row carries. A row with only an email scores lower than a row with email, phone, name and postcode.

Use the score to decide whether your file is ready. If most rows score low, it is usually worth going back to your source and exporting phone numbers or addresses too. If most rows score well, send. The preview and the scores are free, so you can check a list before you commit to a paid plan.

The score is an estimate, not a promise. Each platform runs its own matching on its side and does not report back exactly who it matched. Treat the score as a guide to the quality of your file.

Choosing the right event

You can send a list as Lead, Purchase, Complete registration, Add to cart or Schedule. Pick the event that matches what these people actually did, and ideally the event your new campaign optimises for.

• Purchase for past buyers, with order values if you have them.
• Lead for people who filled in a form or asked for a quote.
• Complete registration for sign ups, trial users or account holders.
• Schedule for people who booked a call, a demo or an appointment.
• Add to cart for shoppers who showed intent but did not buy.

Sending the wrong event teaches the platform the wrong lesson. A list of newsletter sign ups sent as Purchase tells the algorithm to find people like them as buyers, which is not what you want.

What happens on each platform

Meta. Rows arrive on your Pixel as server-side website events, the same way live conversions from your site arrive. Meta can use them for optimisation and audiences on that Pixel.

Google Ads. Rows go through offline click upload. Google only credits a conversion to a person who clicked one of your Google ads, so a list of people who never saw a Google ad will mostly not count there. If your list includes click IDs, those rows have the best chance.

TikTok. Rows are sent to your connected TikTok account as well.

Only the platforms you have connected receive the list. Microsoft Ads, Reddit, Pinterest and X take the list too, sent with today's date.

Limits to know before you send

Every row is sent with the time you press send as its event time. The original purchase or sign up date in your file is not kept. For warming up a new campaign this is usually what you want, since the platforms weight recent events, but it means Ads Warmup is not a tool for rebuilding historic reports.

The same person with the same event on the same day is skipped. If you upload a list twice by mistake, nobody is counted twice. Sending the same list again on a later day is treated as a new send.

If a row fails on a platform, it is saved and you get an email. One click on Resend failed rows sends them again for free, and only the platforms that failed are retried. There is no automatic retry, so check your results after a send.

Uploading, previewing and scoring are free. Sending to the platforms needs a paid plan.

A simple launch plan

Connect Meta, Google Ads or TikTok. Export your best customer list with as many identifying columns as you have. Upload it and look at the match scores. Fix the file if the scores are low. Pick the event your campaign will optimise for, send, and check the result for each person. Then launch the campaign. From there, live conversions from your site keep the platforms learning.