This post started with a conversation with my good friend who is a psychotherapist and doesn’t have a clue about what I do. I was explaining my frustration with a recent report I’d seen that was claimed millions of dollars in lost potential value as a result of one campaign an agency organized. Her response was “that’s like one of my clients who is struggling to keep a job, saying that her mother owed her millions because had not her mother mistreated her as a child, she would have been a rocket scientist.” Yup, that’s the perfect analogy for so many of the value claims I’ve heard over the years.
I’ve spent four decades measuring media value. I’ve seen bad measurements. I’ve seen lazy measurements. I’ve seen well-intentioned measurement that just happened to be wrong. And I’ve seen intentionally inflated definitions of “value” that no one bothered to question. All of it boils down to seven fatal Credibility Killers that far too many people in our industry, including some of the major measurement providers, keep using. If this sounds familiar, here’s a mini master class in what not to do.
Credibility Killer #1 Claiming half the planet saw your press coverage (aka what is a “valid” impression?)
I can’t tell you how many agency and vendor reports I’ve seen that cite a total audience reach (aka impression counts) that, when I actually do the math, suggest that whatever they are promoting is the equivalent of a moon landing. Citing impression counts — assembled by stacking up publisher-reported Unique Visitors per Month figures from every outlet that mentioned the event or product — implies that more than half the humans on Planet Earth have been reached. (Never mind what the company might do if 4 billion people started demanding their product or service.) Numbers like that have damaged the credibility of PR for decades.
The lesson:
As Johna Burke, head of AMEC (that created the industry standard Barcelona Principles) might say: Impressions don’t necessarily impress anyone.” Impressions are only valuable if they help achieve the goal your campaign was trying to achieve. That means they reached people who can actually act on them and persuaded them to act in a way that helps achieve your goal.
Never mind that, unless you are trying to reach the AI army, industry data consistently shows that approximately 40-50% of digital impressions are non-human. The IAB and MRC have published Invalid Traffic Detection and Filtration Guidelines specifically because this problem is so pervasive and so rarely corrected for. When sophisticated advertisers buy digital media, they routinely demand bot-scrubbed impression data. PR measurement should hold itself to the same standard.
Credibility Killer #2 An impression outside your market that doesn’t advance your goals is worthless
This is the most overlooked adjustment in PR measurement and arguably the most important for niche market-focused clients. An impression only has commercial value if it reaches someone who can act on the message — buy the product, visit the store, attend the event, donate to the cause. In other words, a valid PR impression must reach someone with the ability and intent to act on the message. For a regional US retailer operating in 35 states, impressions reaching people in the other 15 states have no purchase value. Impressions reaching non-customers even within the operating footprint have limited value. Impressions outside the country have none at all. And we’re assuming all this coverage leaves someone more inclined to do whatever it is you want them to do. Coverage that leaves them less likely to achieve your goals is not only not valuable, but potentially damaging to the brand.
The Lesson:
Only count desirable content that appears in the agreed upon Top Tier Priority Media. Or at least ask yourself: What percentage of this outlet’s audience is actually in our (or our client’s) addressable market and benefits the brand?
Credibility Killer #3 Impression counts don’t mean what you think they mean
Publisher-reported impression figures, typically based on Unique Visitors per Month (UVM) represent traffic to the entire media website, not to the specific article that mentioned your product or your client’s product. When your coverage appears on MSN or Forbes, you are not reaching their entire monthly audience. You are reaching the people who happened to navigate to that specific page on that specific day.
A site with 5 million monthly visitors might deliver 5,000 actual page views to a single article — a ratio of 1,000:1. When you apply the site’s full UVM to your impression count, you are overstating reach by orders of magnitude. Multiply that across a campaign list of hundreds of articles and the headline impression figure becomes meaningless. The reality is that the odds of any particular post being seen are vanishingly small — but UVM-based measurement treats them as equal.
The Lesson:
If available, demand your vendors provide page-level traffic data, not site-wide UVM, when evaluating coverage value. When UVM is all you have, apply a conservative fraction (typically 5-10%) to estimate article-level reach. At the very least, document your methodology explicitly so it can be explained and defended and with luck maintain some credibility.
Credibility Killer #4 The Platform Problem
Too many vendors include platforms like Medium, MSN, and Yahoo News to boost the impression counts. They fail to mention that just because someone’s home page defaults to MSN or Yahoo, doesn’t mean they are reading about the company or product at hand. Including Medium is even worse. Anyone who pays for an account can post anything they want, which results in over 1.38 million unedited, unreviewed additional posts every month. The odds of your single post being seen by all the medium users are infinitesimal. But vendors routinely report that your one post or story receives Medium’s full site impression count. To imply that your event or product is the equivalent of the platform’s most-read article of the month not credible.
The Lesson:
If your product or event is local, or regional or not available outside a specific geographic area, you need to strip out international audiences and audiences outside the relevant geography. (Or just delete anything from MSN, Medium, and Yahoo.) Yes, two-thirds of the claimed impressions may evaporate, but at least you’re being honest. For national brands, this adjustment is modest. For regional retailers, healthcare systems, restaurant chains, event promoters, or nonprofits with defined service areas, this filter is essential — and will substantially reduce your headline numbers. The most precise source for this data is the client’s own loyalty or CRM data, which defines the actual customer base far more accurately than any third-party demographic estimate.
Credibility Killer #5 Applying a dollar value
By now the problems with using AVE should have been pounded into every communicator’s brain, but still they persist. (Unfortunately, Cision, Meltwater, MuckRack and Critical Mention are still ignoring the Barcelona Principles by providing them.) The principles make clear that paid advertising has different goals and objectives than earned media. When you pay for an ad, presumably you make sure it contains the images, messages and positioning needed to increase sales or persuade their audiences of whatever you need them to do. It seems insane that I have to reiterate this, but in earned media, the writer, editor, or publisher is in control. They decide what appears and what gets spiked. Thus, advertising is not the “equivalent” of an earned media editorial post unless it contains ALL the same images, messages and positioning that drives audience behavior. So, whether you use the AVE numbers off your vendors report, or some version of CPM numbers, you are basing your results on a fundamentally false equivalency.
CPM is another commonly used metric that reflects what publishers charge buyers to deliver 1000 eyeballs. This is useful if you are paying to reach the eyeballs you need with the right messages to achieve your goal. But in earned media, I’ve seen practioners apply single flat “average “rate applied to a broad range of media types, which simultaneously overvalues low-CPM digital outlets and undervalues high-CPM broadcast placements. In one campaign, the agency applied average CPM rates to all the media. When I examined the actual CPM rates of the publications, there was a more than a $45 million swing from low to high values.
The Lesson:
The industry-standard Barcelona Principles — adopted in 2010 and updated in 2015, 2020, and 2025, endorsed by PRSA, CIPR, and PR practitioners from 33 countries — explicitly reject AVE-based and CPM-based valuation as invalid measures of PR value. If you are still multiplying gross impressions by a flat CPM and presenting the result as media value, you are using a methodology your industry’s own standards body has formally rejected for 15 years. Anyone reporting ad equivalency in any form deserves to lose all credibility.
Credibility Killer #6 Attribution – What actually generated the coverage?
Agencies, this one’s for you. Far too many agencies claim credit for every piece of news that is generated, regardless of why a reporter or blogger or influencer chose to write about your client.
For an agency to claim total credit for all coverage that appeared in a particular monthe or quarter ignores all the other possible reasons coverage might have appeared, like an earnings report, proximity of other news to your client’s product or brand or appearing in a random rant by a political figure. Can you imagine Tylenol’s agency claiming credit for that all of its coverage in September 2025?
The Lesson:
In any report, separate the coverage that was actually generated by you or your agency from everything else, and insert a comment or a thought balloon in your charts so that the cause of the coverage is clear. Otherwise, your credibility will be at stake.
Credibility Killer #7 The Beneficiary: Who actually benefited from the coverage if anyone?
This is the question most media value reports never ask — and it may be the most strategically important one. There is a meaningful difference between estimating total media value and determining how much of that value accrued to your specific client.
When coverage is generated by a multi-party campaign or event, i.e., an event with multiple sponsors or a campaign with multiple agencies, the “value” does not flow exclusively to one party. Presumably, every participant who receives media mentions derives some value. A celebrity headliner, the organizers, and paying sponsors whose logos appears on stage all have claims on the value generated by that coverage.
I measured an event that featured a primary sponsor, my client, as well as numerous secondary sponsors, and a well-known entertainer. Coverage of the event generated hundreds of stories. But my analysis of the actual articles showed that less than half didn’t mention the primary sponsor at all. Most (55%) of the coverage focused on the celebrity. Any rigorous measurement must assess what share of coverage value actually served the client’s specific objectives — not simply what publicity the event generated in aggregate.
The Lesson: Analyze what percentage of articles mention your client specifically and assess whether coverage drove desirable outcomes for your client or primarily for co-participants.
Why this matters for the profession
The goal of communication is not to produce the largest number. It is to reach the right people with the right message that persuades them to act or believe in the way you need them to. The problem is that we consistently deliver inflated numbers, we train clients to expect them. When those numbers don’t connect to real business outcomes — store visits, sales, awareness shifts, loyalty program sign-ups — clients stop trusting PR measurement altogether. And when they stop trusting measurement, they stop investing in the work.
Six Things to Check Before You Deliver Any Report
- Has the audience been filtered for geographic relevance and ability to act on the message?
- Has both invalid traffic been deducted? (IAB/MRC benchmarks: 40–50% of digital impressions)
- Are the numbers I am reporting based on article-level data or inflated, site-wide UVM?
- Has attribution been assessed — what share of coverage value actually accrued to your client versus co-participants?
- Is all the coverage beneficial, and have I isolated the coverage accordingly?