DV measures the impact of ChatGPT ads and tests brand safety

What's new

DoubleVerify, abbreviated DV, is a US company that measures and verifies digital ads for major brands. Now the company has launched attribution measurement for ads in ChatGPT. Attribution means figuring out which ads and channels actually got the customer to buy.

Advertisers can now count ChatGPT ads as one of several touchpoints on the customer's journey, alongside, for example, search ads, social media and TV. This is meant to show what ChatGPT contributes to the result and to the overall return on ad spend (ROAS).

In addition, DV and OpenAI are developing a pilot on brand suitability. Here, independent parties are allowed to test and assess which rules and safeguards should apply to where ads may appear. It takes place in a controlled test environment and, according to the sources, without access to users' private conversations.

What's clever

A conversation with an AI is not like a website. What the user asks, what the answer says and how the conversation develops can change along the way. That is why you cannot simply reuse the old rules for where an ad may appear.

With attribution measurement, companies can see ChatGPT ads in the context of the rest of their marketing instead of looking at them in isolation. That makes it easier to determine whether the money is well spent.

The pilot is meant to help figure out when an ad fits a conversation and when it does not. That is important for brands that do not want to appear next to sensitive topics.

Cheaper or better?

This is not a new AI model, so there is no price to compare. The price for DV's new measurement is not stated in the material I have found. Ordinary ChatGPT users will not notice anything either, as the tool is for advertisers.

An early example comes from Weight Watchers. In the test, the attributed cost per acquired customer (CPA) on ChatGPT was 15.3 percent below the company's own benchmark in the measured period. That is a single case from a press release and not general proof.

The brand suitability part is still a pilot and therefore not a finished product. How DV compares with competitors such as IAS is not stated in the sources. OpenAI's head of ads has previously said that independent verification is a natural next step, but without naming vendors.

What it's good for

The news is first and foremost for marketers and agencies that are considering spending money on ads in ChatGPT, or that already do. They can better document what the ads deliver and keep track alongside other channels.

For Danish companies, it is worth knowing that the sources do not state when or whether the tools are available in Denmark and the EU. Ordinary ChatGPT users can see it as a sign that ads in chatbots are becoming a permanent part of the internet and that the industry is working on the rules of the game.