Celebrity ROI: What the Numbers Tell You

Measuring ROI on celebrity deals is one of the most contested problems in marketing, partly because the numbers are genuinely hard to isolate, and partly because a lot of people in the room have an incentive not to look too closely. The honest answer is that most celebrity partnerships are not measured rigorously, they are justified retrospectively, and the industry has quietly agreed to keep it that way.

That does not mean celebrity deals cannot work, or that ROI is unmeasurable. It means the measurement frameworks most brands use are built for comfort, not clarity. If you want to know whether a celebrity deal is actually earning its fee, you need a different set of questions.

Why Celebrity ROI Is So Consistently Overstated

I spent years running agency P&Ls and watching how brands reported on celebrity and influencer partnerships. The pattern was almost universal: the deal gets signed with vague success criteria, the campaign runs, and then someone builds a slide that makes the numbers look good enough to justify renewal. The rigour applied to measuring a paid search campaign is never applied to measuring whether the celebrity actually moved the business.

Part of this is structural. Celebrity deals sit at the intersection of brand and commercial, which means accountability is diffuse. The brand team owns the relationship, the performance team owns the numbers, and neither has full visibility of what the other is doing. In that gap, measurement gets soft.

Part of it is also the same problem I have seen in performance marketing more broadly. Earlier in my career, I overvalued lower-funnel signals. Clicks, conversions, cost-per-acquisition, all of it looked clean and causal. It took years of working across enough businesses to recognise that a lot of what performance channels get credited for was going to happen anyway. Someone who was already in-market, already searching, already close to a decision, converted. The channel got the credit. The celebrity version of this is someone who was already a fan of the brand, saw the campaign, and bought. The celebrity gets the credit.

The harder question, and the more commercially important one, is whether the deal brought in people who would not otherwise have come. That is the question most measurement frameworks are not built to answer.

What You Are Actually Trying to Measure

Before you can measure celebrity ROI, you need to be honest about what the deal was supposed to do. There are broadly three legitimate commercial objectives a celebrity partnership can serve, and each requires a different measurement approach.

The first is audience expansion. The celebrity brings their audience to your brand, people who did not previously consider you become aware of and interested in you. This is the highest-value outcome and the hardest to measure cleanly. The proxy metrics here are new customer acquisition rate, first-purchase data segmented by acquisition channel, and audience demographic shift over the campaign period.

The second is brand perception shift. The celebrity’s association changes how existing and potential customers feel about the brand, its quality, its relevance, its desirability. This requires brand tracking with proper baselines, not a post-campaign survey with no control group. If you did not measure brand perception before the deal, you cannot claim it shifted because of it.

The third is short-term sales activation. The celebrity appears in a promotion, a product launch, or a retail moment, and drives a measurable uplift in purchase intent or actual sales within a defined window. This is the easiest to measure and the most commonly abused, because it conflates the celebrity’s contribution with the promotional mechanic, the media spend, and the seasonal context.

If you cannot articulate which of these three your deal was designed to deliver before it launches, the measurement conversation is already lost. This kind of commercial clarity is exactly what separates strong go-to-market thinking from activity for its own sake, and it is something I write about regularly in the Go-To-Market and Growth Strategy hub on The Marketing Juice.

The Attribution Problem Nobody Wants to Talk About

Attribution in celebrity partnerships is broken in the same way it is broken everywhere else in marketing, just with more money at stake and more ego in the room.

The fundamental problem is that attribution models assign credit to touchpoints based on proximity to conversion, not based on causal influence. A customer who saw a celebrity post on Instagram, then searched the brand name, then clicked a paid search ad, then converted will almost certainly have that conversion credited to paid search. The celebrity post, which may have been the catalyst for the entire sequence, gets nothing.

This is why brands consistently underinvest in the upper funnel and then wonder why their performance channels stop working at scale. I have seen this play out across multiple turnaround situations. The business cuts brand spend, doubles down on performance, sees short-term efficiency gains, and then watches acquisition costs climb as they exhaust the pool of people who already knew and trusted them. The celebrity deal, the brand campaign, the awareness activity, all of it was creating the conditions for performance to work. When it stops, the system degrades slowly and then suddenly.

The more honest approach to celebrity attribution involves a few things most brands are reluctant to do. First, measure new customer acquisition as a primary KPI, not total conversions. If the deal is not bringing in people who were not already yours, it is not doing what celebrity deals are supposed to do. Second, use brand search volume as a leading indicator. A genuine celebrity-driven awareness spike almost always shows up in branded search before it shows up in sales. Third, run incrementality tests where the budget allows, with holdout regions or audience segments that did not receive the celebrity campaign, so you have something to compare against.

The Metrics That Actually Matter

Impressions and reach are reported on every celebrity deal debrief I have ever seen. They are also among the least useful numbers in the pack. Reach tells you how many people the content was served to. It tells you almost nothing about whether those people were the right people, whether they registered the message, or whether anything changed as a result.

The metrics worth tracking fall into three categories.

Audience quality metrics: what proportion of the celebrity’s audience overlaps with your target customer profile, what proportion is genuinely new to your brand, and what the demographic and psychographic composition looks like. Platforms like Later have written thoughtfully about how creator campaigns can be structured to reach the right audiences rather than just large ones. The size of a celebrity’s following is far less relevant than the fit between that following and your growth opportunity.

Commercial outcome metrics: new customer acquisition rate during and after the campaign period, first-purchase average order value from new customers acquired during the campaign, and customer lifetime value projections for that cohort. If you can track these, you can start to build a genuine fee-to-outcome ratio. A celebrity who costs £300,000 and delivers 20,000 new customers at a healthy LTV is a better deal than a celebrity who costs £80,000 and delivers a spike in engagement from your existing base.

Brand health metrics: unaided awareness among target segments, consideration and preference scores, and net sentiment. These require proper tracking infrastructure, which most brands do not have in place before they sign a deal. If you are serious about measuring celebrity ROI, brand tracking needs to be set up before the campaign launches, not commissioned afterward to validate a decision already made.

How to Set Up the Measurement Before the Deal Is Signed

The measurement conversation should happen before the contract is signed, not after the campaign has run. This sounds obvious. In practice, it almost never happens that way.

When I was running agencies, the deals that got measured properly were the ones where the client had defined success criteria upfront and built the measurement infrastructure before the campaign launched. The deals that got measured poorly were the ones where the celebrity was secured, the content was produced, the campaign went live, and then someone asked: how do we show this worked?

The pre-deal measurement checklist should include: a baseline brand health snapshot across your target segments, a clear definition of what new customer acquisition looks like and how it will be tracked, agreement on what the control condition is (what would have happened without the celebrity), and a defined window for measurement that accounts for the lag between awareness and purchase in your category.

That last point matters more than most people account for. In a fast-moving consumer category, the purchase cycle might be days. In a considered purchase category, it might be months. A celebrity campaign for a car brand that is measured over four weeks is not going to show the full commercial picture. The measurement window needs to match the purchase cycle, not the campaign calendar.

BCG’s work on commercial transformation and go-to-market strategy makes a related point about the importance of aligning measurement frameworks to commercial objectives rather than to what is easy to track. The same principle applies here. Easy-to-track metrics (impressions, engagements, hashtag mentions) are almost always the wrong metrics for evaluating whether a celebrity deal is earning its fee.

The Honest Conversation About Soft Value

There is a version of this conversation where someone says: celebrity deals have soft value that cannot be fully measured, and that is fine. I have some sympathy for this position, but it is frequently used as a shield against accountability rather than as a genuine argument about measurement complexity.

Soft value is real. A celebrity association can shift how a brand is perceived in ways that compound over time and do not show up cleanly in any single measurement window. I have judged the Effie Awards and seen campaigns where the long-term brand-building effect was clearly visible in the market data even when the short-term sales numbers were modest. The relationship between brand investment and commercial outcome is not always immediate, and penalising every piece of brand activity for failing to produce a clean short-term ROI number is a recipe for a brand that slowly stops meaning anything.

But soft value needs to be argued for explicitly, not invoked as a get-out clause when the hard numbers do not add up. If you are making the case that a celebrity deal is delivering brand value that will pay out over a longer horizon, you need to show the brand tracking data that supports that argument. You need to show that awareness, consideration, and preference are moving in the right direction among the right audience. Saying “this is hard to measure” is not the same as saying “we are measuring the right things and they are moving.”

The broader challenge of balancing short-term measurement with long-term brand investment is one of the defining tensions in modern marketing strategy. It comes up constantly in the growth strategy work I cover here, because it sits at the heart of how businesses allocate marketing budgets and how they think about what marketing is actually for.

When Celebrity Deals Are Worth the Money

For all the measurement complexity, there are situations where celebrity partnerships genuinely earn their fee and the ROI case is relatively straightforward to make.

The clearest case is audience expansion into a segment where the brand has low penetration and the celebrity has genuine credibility. If your brand needs to reach a younger demographic, or a different cultural community, or a geographic market where you have no existing presence, a celebrity with authentic standing in that community can open doors that media spend alone cannot. The measurement is relatively clean: did new customer acquisition from that segment increase, and by how much, relative to what you would have expected without the deal?

The second strong case is category creation or repositioning. When a brand is trying to change what it stands for, celebrity association can accelerate the perception shift in a way that advertising alone struggles to do. People update their beliefs about brands partly through the signals sent by who chooses to associate with them. A celebrity who is credible in a space you are trying to own sends a signal that advertising cannot replicate. The measurement here is brand tracking over a longer window, with specific focus on the perception dimensions you are trying to shift.

The weakest case, and the one most commonly made, is using a celebrity to drive short-term sales among an existing customer base that already knows and likes the brand. This is the equivalent of the clothes shop scenario I think about often: the customer who tries something on is already close to buying. The celebrity in this context is taking credit for a conversion that was already in motion. You might see a sales spike. You will almost certainly attribute too much of it to the celebrity and too little to the promotional mechanic, the media weight, and the existing brand equity that made the customer receptive in the first place.

A Framework for Honest Celebrity ROI Evaluation

If I were advising a brand on how to evaluate a celebrity deal honestly, I would structure the assessment around four questions.

First: what is the counterfactual? What would have happened to awareness, consideration, and sales if you had spent the same budget on media without the celebrity? This is uncomfortable to ask because it implies the celebrity might not be necessary, but it is the only way to isolate the celebrity’s actual contribution from the contribution of the media weight behind them.

Second: are we reaching new people? New customer acquisition data, broken down by whether those customers were previously known to the brand, is the single most important commercial signal. Growth requires reaching new audiences. Deals that only resonate with existing customers are brand maintenance at best, and expensive brand maintenance at that.

Third: is the brand health moving in the right direction among the right people? Not total awareness, which is a blunt instrument, but awareness, consideration, and preference among the specific audience segments the deal was designed to reach. Growth hacking tactics and performance tools like the ones covered at Crazy Egg can help diagnose on-site behaviour changes that follow a campaign, but they need to be read alongside brand data, not instead of it.

Fourth: what is the fee-to-outcome ratio relative to alternatives? This is the question that almost never gets asked, because it requires comparing the celebrity deal to what else the money could have done. If the same budget deployed into a different channel, a different creative approach, or a different market would have delivered better outcomes, the celebrity deal did not earn its fee regardless of what the post-campaign report says.

None of this is easy. But it is honest. And in my experience, the brands that measure celebrity deals honestly are the ones that get better at them over time, because they know what is actually working and why. The brands that measure them for comfort end up renewing deals that are not moving the business and passing on deals that would have.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions