Paramount Made CTV Measurable. Ask Who Is Doing the Measuring.

On 22 October 2025, Taboola and Paramount Advertising announced Performance Multiplier, a Paramount-branded product built into Paramount Ads Manager and aimed at small and mid-sized advertisers. It uses Taboola’s Realize technology to extend a CTV campaign to matched and lookalike viewers across Taboola’s network of more than 9,000 publisher partners, then reports post-view outcomes such as clicks, sign-ups and purchases inside the Paramount dashboard (Taboola’s announcement).

It is a reasonable product aimed at a real gap. It is also a useful case for a question that will outlast this particular partnership, because a dozen more like it are coming: when a media seller also supplies the proof that the media worked, what exactly have you been shown?

There are two products in one box

Separating them is the first analytical move, because they are evaluated differently.

The first is a media extension. After someone sees your spot on Paramount’s streaming inventory, you can reach that household, and households that resemble it, across a large network of publisher sites. That is a straightforward retargeting and prospecting product, and its value depends on the usual things: inventory quality, price, and whether the audience match is any good.

The second is an attribution readout. The dashboard shows what happened after exposure, so a channel that produces no click acquires a column of outcomes. That is genuinely new for an advertiser buying streaming through a self-service tool with no measurement function of their own.

The trouble is that the two are sold together, which means the thing being measured includes the thing doing the measuring.

Post-view is not incremental, and the gap between them is the whole argument

A post-view outcome is an observation: this person saw the ad, and later did something. It is not a claim about causation, and it cannot become one by being displayed next to a spend figure.

Follow the sequence here. A household sees a CTV spot. That household, or one that looks like it, is then served display units across publisher sites. Some of those people click, sign up or buy. The dashboard reports the outcome as a result of the campaign, and it is, in the sense that the campaign touched them twice. What it cannot tell you is how much of the outcome came from the television exposure, how much from the second paid touch, and how much would have happened without either. Three different questions, one number.

None of that makes the product dishonest. Post-view reporting is standard across the industry and it is more informative than the alternative, which for most small advertisers is nothing at all. It is simply a different class of evidence from a test, and the failure mode is treating the two as interchangeable when the budget conversation arrives.

The durable question is whether the answer could ever be no

Every vendor measurement product, from this one to whatever launches next quarter, can be assessed with the same short list. It is worth keeping.

• Who is counting? Measurement supplied by the seller is not third-party measurement, however sophisticated the technology.
• Is this post-view or incremental? If the method has no unexposed comparison group, it is observation.
• Does the measurement product also spend money? If attribution arrives bundled with a media extension, the reported outcome includes that extension.
• Could it ever return a negative? A method that cannot show a campaign did nothing is a reporting tool rather than a measurement one.
• Do you keep the data? Numbers that live only in one platform’s dashboard cannot be compared against anything else you run.
• Does it support a holdout? A vendor confident in the effect will let you withhold a matched group and show them the difference.

The last one is the fastest test of a sales conversation. Ask whether the platform will support a geographic holdout and watch what happens next.

What this genuinely improves

Being skeptical about the method is not the same as dismissing the product, and there are two things here worth taking seriously.

Access plus a directional read is a real improvement for an advertiser who previously had neither. A small business buying streaming at any budget through a self-service platform is not going to commission an incrementality study, and a post-view column is more useful to them than an impression count.

The second is directional. Streaming sellers competing on business outcomes rather than reach is the correct fight for the category, and it puts pressure on every other platform to explain what happens after the ad. That pressure is good for buyers regardless of which product wins, and it is the same effect search marketers see when a television flight runs, now packaged for people who cannot run the analysis themselves.

Use it as a supplement, never as the test

The practical position is straightforward. Take the vendor dashboard as a signal that the campaign is delivering and that people are behaving as expected downstream. Then design the actual test yourself: matched markets held out, total conversions compared, branded search watched in both. What the channel is and why it cannot be clicked sits in the difference between a screen and a delivery method, and what a readable first flight costs to run is in what a first CTV buy actually costs.

A vendor number and a holdout test answer different questions, and the second one is the one you will need when somebody asks whether to spend more. Fitting both into a plan that already includes search, social and everything else is a media program question, and if you want an outside read on whether the measurement you are being sold tells you anything your own account cannot, that is worth checking before renewal rather than after.

Frequently Asked Questions

What is Performance Multiplier?

A Paramount-branded product, built with Taboola and integrated into Paramount Ads Manager, that extends a CTV campaign to matched and lookalike viewers across Taboola’s publisher network and reports post-view outcomes such as clicks, sign-ups and purchases in the Paramount dashboard. It was announced in October 2025 and is aimed at small and mid-sized advertisers buying streaming through a self-service platform.

Is post-view attribution reliable?

It is reliable as a description and unreliable as a causal claim. Post-view reporting tells you accurately that a person who saw the ad later took an action. It cannot tell you whether the ad caused the action, because there is no comparison group of similar people who did not see it. Use it to confirm delivery and downstream behavior, not to decide whether the channel earned its budget.

Should we avoid vendor-supplied CTV measurement?

No, use it and understand what it is. Seller-supplied measurement is faster, cheaper and more granular than anything you would build, and for many advertisers it is the only option. The discipline is to keep one independent read alongside it, usually a geographic holdout, so that a disagreement between the two is visible rather than invisible.

What should we ask a CTV vendor about measurement?

Four questions cover most of it. Is this post-view or incremental? Does the attribution product also spend media on my behalf? Can the methodology ever show that a campaign produced nothing? And will you support a matched-market holdout so I can check? The answers tell you more about what you are buying than any case study will.