Nielsen Bought DoubleVerify Own Your Own Data

Measurement is consolidating and most teams do not own theirs

Nielsen is acquiring DoubleVerify for roughly two billion dollars, with the close expected in the first quarter of 2027. It follows Integral Ad Science being taken private last year. The independent verification layer of the advertising industry, which existed specifically in order to be independent, is being absorbed into fewer and larger owners.

At the same time the platforms are reporting less. Google's move to AI Max reduces visible control and visible reporting. Meta replaced placement exclusions with value rules. Google Search Console has been logging impressions incorrectly in its generative AI report since mid August.

The pattern is familiar from other industries. A category fragments, dozens of specialists appear, buyers get tired of managing dozens of relationships, and the category consolidates into a handful of platforms that do most things adequately. That is usually good for buyers on price and convenience. It is less good for anyone whose strategy depended on a specific vendor behaving a specific way.

None of those facts is alarming on its own. Together they describe a direction, and the direction is that the numbers you use to run your business are increasingly produced by people who are not you.

Rented numbers are a dependency

Every growth team runs on a stack of measurement it did not build. Platform reporting, a verification vendor, an analytics suite, an attribution model. It works, right up until an ownership change, a methodology update or a product decision changes what the numbers mean, and then a year of trend data becomes uncomfortable to interpret.

That is not a scandal. It is the normal life cycle of any vendor relationship. The mistake is treating rented measurement as though it were owned infrastructure, and building decisions on top of it as though the definition were permanent.

There is a second reason this matters more than it used to. As more of the buying decision moves into automated systems, the measurement you feed those systems becomes the steering. A conversion definition that is slightly wrong no longer just produces a misleading report, it actively teaches the machine to find the wrong customers. Bad measurement used to cost you clarity. Now it costs you spend.

This publication has made the argument in a different context. Growth problems are usually infrastructure problems. Measurement is infrastructure, and most companies have been renting theirs without ever writing down what happens if the landlord changes.

First party data is the cheapest advantage available

The alternative is not to build a measurement company. It is to make sure you own a spine of your own that does not move when the vendors do.

That means a customer record you control, keyed to something durable. It means server side capture of the events that matter to your business, defined by you, in your own words, stored where you can query them in five years. It means at least one number per funnel stage that you calculate yourself from your own source, so you have a constant to compare vendor numbers against.

It is unglamorous work and it competes for resources with things that show results this quarter. It is also the cheapest durable advantage available to most businesses, because it compounds. A company with four years of consistently defined first party data can answer questions a competitor with better tools and no history cannot answer at all.

What to build before you need it

None of this argues against vendors. Verification, panel data and third party measurement do things no internal team should try to replicate. The argument is only about dependency. Use the vendors for what they are good at, and keep enough of your own record that you would survive losing any one of them without losing your ability to answer a basic question about your business.

Define your own metrics in writing. What counts as a lead, a customer, a repeat, a churn, stated precisely enough that two people in the company would calculate them the same way. Most organizations have never done this, which is why the argument about whose number is right never ends.

Capture the events yourself. Server side, in your own store, alongside whatever the platforms record. The volume is smaller than people expect and the cost is mostly one time.

Keep a manual baseline. One number per stage, calculated from your own data, tracked monthly. When a vendor changes methodology, that baseline is what tells you whether the business moved or the definition did.

Consolidation in the measurement industry is not a crisis. Nielsen buying DoubleVerify will probably produce a more capable product than either had alone. But every consolidation is a reminder of the same thing, which is that the only measurement you can be certain will still mean what it means today is the measurement you built.

← All stories