Why Your PR Agency Can’t Measure ROI (And How AI Changes Everything)
Most PR reporting still looks like it’s stuck in 2006.
A clip book. A reach number. A handful of “top tier” logos. Maybe a sentiment chart that somehow always trends positive.
And then — because everyone can feel the emptiness — someone throws in a big, reassuring number:
“This campaign generated $500,000 in AVEs.”
Advertising Value Equivalents (AVEs) are the PR industry’s comfort blanket: tidy, impressive-looking, and fundamentally disconnected from reality.
Here’s the uncomfortable truth: most agencies can’t measure PR ROI — not because they’re lazy, but because until recently the measurement was genuinely hard and expensive.
But 2026 is different.
AI isn’t just making PR faster. It’s making outcome-based measurement accessible to teams who’ve never been able to afford it. And that changes what clients should demand — and what PR practitioners should sell.
This article is for two groups:
- Clients who are tired of paying for “coverage” and want evidence of impact.
- PR and comms people who know AVEs are nonsense but still need a credible way to prove value.
Let’s talk about why PR measurement got stuck, what actually matters, and how to build a measurement approach that makes PR a strategic function again.
The AVE problem (and why it lasted so long)
AVEs persisted for one reason: they were simple.
PR is messy. It influences decisions indirectly. It works through trust, reputation, and timing. It shapes perception and reduces friction in the sales process. None of that shows up neatly in a spreadsheet.
So the industry reached for a proxy:
- If advertising costs X per column inch
- and our story got Y column inches
- then… our PR is worth X × Y.
It was always a stretch.
It also created a perverse incentive: chase coverage, not outcomes.
The irony is that the PR profession has known for years that this was wrong. The Barcelona Principles pushed the industry away from AVEs and toward outcomes-based measurement. Plenty of senior people have written smart critiques.
And yet AVEs kept showing up in reports.
Because the alternative required three things that many teams didn’t have:
1. Instrumentation (tracking links, campaign landing pages, analytics discipline)
2. Integration (CRM, marketing automation, attribution)
3. Interpretation (someone who can connect comms activity to business outcomes)
Until recently, doing this well was enterprise-only.
Now it isn’t.
What PR actually does (that AVEs can’t see)
If you want to measure PR, you have to start with what PR truly drives.
In most businesses, PR creates value in four main ways:
1) It increases qualified awareness
Not “reach.” Not eyeballs. Awareness among the people who matter.
The result isn’t immediate sales — it’s that when your audience encounters you later (through ads, search, sales outreach, word of mouth), you feel familiar and credible.
2) It builds credibility (and changes conversion behaviour)
Good PR reduces uncertainty.
People don’t just buy products — they buy risk reduction. PR is often the layer that makes a brand feel “real” and safe.
This shows up as:
- higher conversion rates on key pages
- higher demo request completion
- shorter sales cycles
- less price sensitivity
3) It shapes reputation (which affects talent, partners, and investors)
PR isn’t only about customers.
It affects:
- who applies for roles
- who returns your calls
- whether partners take a meeting
- whether you get the benefit of the doubt when something goes wrong
4) It creates a narrative position in-market
This is the “crown jewel” effect.
If your category is crowded, the story you own matters. PR doesn’t just get attention — it sets context.
AVEs can’t capture any of this.
But we can.
Why PR ROI measurement is hard (and how AI changes the equation)
There are two classic blockers to measuring PR ROI:
Blocker #1: Attribution isn’t straightforward
PR works across time and touchpoints.
Someone might:
- see an article
- follow you on LinkedIn
- read a blog post
- watch a webinar
- then finally convert after a sales call
If you only measure “last click,” PR looks useless.
The fix is multi-touch thinking — something like:
- “PR influenced this opportunity”
- “PR accelerated this deal”
- “PR improved conversion at the consideration stage”
Blocker #2: The data is messy
Even when you can track outcomes, it requires discipline:
- UTM links
- consistent campaign naming
- a CRM that isn’t a junk drawer
- definitions that everyone agrees on (“qualified lead” means what?)
This is where AI is genuinely helpful.
AI can:
- classify inbound leads by intent
- summarise qualitative feedback from sales calls
- cluster themes across media coverage
- automate reporting and anomaly detection
- connect messy, partial signals into something decision-ready
The biggest shift is this:
AI makes measurement possible for SMBs that couldn’t justify an enterprise analytics stack.
Which means the bar is rising — for agencies and for clients.
A practical AI-enhanced PR measurement framework (that doesn’t require a data science team)
Here’s a simple four-layer model you can use whether you’re a comms lead, a founder, or an agency.
Layer 1: Output (the basics)
This is the stuff PR teams have always reported:
- placements
- share of voice
- message pull-through
- sentiment (carefully)
This layer is not useless — it’s just not enough.
AI helps here by automating monitoring, tagging, and summarising coverage so you spend less time counting.
Layer 2: Engagement (signals of real interest)
This is where you start connecting PR to audience behaviour:
- referral traffic from coverage
- branded search lift
- time on site from PR-driven visitors
- content downloads
- newsletter signups
- event/webinar registrations
Key point: you don’t need perfect attribution. You need directionally trustworthy signals.
Layer 3: Outcomes (what the business actually cares about)
Pick 2–3 outcomes that match the business goal:
- Pipeline influenced (opportunities that had PR touchpoints)
- Deals accelerated (PR activity correlated with shortened cycle)
- Revenue attributed (where practical)
- Recruitment (quality applicants, acceptance rate)
- Partnerships (partner inbound, partner deal value)
This is where PR becomes a board-level conversation.
Layer 4: Strategic impact (narrative + positioning)
This is the hardest layer — and the one most valuable when done well:
- Are the right messages being repeated in-market?
- Are we being described the way we want?
- Are we winning the “category story”?
AI is useful here for analysing language at scale: coverage, competitor positioning, and message resonance.
How to implement this without boiling the ocean
If you’re a PR practitioner, here’s the simplest starting point:
Step 1: Retire AVEs (today)
You don’t need a new framework to stop reporting the wrong metric.
Replace AVEs with:
- output metrics (for transparency)
- engagement metrics (to show movement)
- one outcome metric (to prove value)
Step 2: Pick 2–3 outcome metrics per client
Examples:
- B2B SaaS: demo requests, sales-qualified pipeline influenced
- eCommerce: conversion rate lift, email list growth, repeat purchase lift
- Professional services: consult calls booked, proposal requests
Step 3: Instrument the basics
- Use clean UTMs
- Use unique landing pages for major campaigns
- Make sure GA4 events cover the core conversion actions
- Keep naming consistent (campaigns, sources, mediums)
Step 4: Build a monthly report that answers one question
Not: “What did we do?”
But:
“What changed in the business that we can credibly connect to communications activity?”
If you can answer that consistently, you’ll stop being a vendor and start being a strategic partner.
If you’re hiring PR: the questions that separate agencies fast
Here are questions that quickly reveal whether an agency understands outcomes:
1) What business outcome will this work influence?
2) How will you track that outcome?
3) What would make us stop or change approach? (a real partner will define failure conditions)
4) What’s your view on AVEs? (if they defend them, run)
5) How will you integrate with marketing and sales?
PR can’t be measured in isolation. If an agency treats comms as a separate universe, the reporting will be theatre.
The future: output gets commoditised, outcomes become the premium
AI will continue to commoditise parts of PR:
- drafts
- media list building
- monitoring
- first-pass reporting
That’s not the threat.
The threat is agencies staying stuck selling “coverage” while clients get smarter about outcomes.
The opportunity is this:
Agencies (and comms leaders) who can connect narrative to measurable business outcomes will win better clients, charge higher retainers, and do better work.
Conclusion: retire one metric this quarter
I’ve worked in PR long enough to remember when AVEs felt like the only way to make the work tangible.
But I don’t think we have that excuse anymore.
The industry is at a crossroads:
- Keep selling what’s easy to count
- Or build measurement that reflects what PR actually does
If you’re a practitioner: retire AVEs and replace them with a simple outcomes framework.
If you’re a client: demand measurement that maps to your business — pipeline, revenue, recruitment, or reputation.
Because AI isn’t the threat.
Irrelevance is.