The right PR ROI framework tracks four things: share of voice against named competitors, tier-weighted placement volume, downstream search and inbound lift, and cost per credible placement. Impressions, reach, and clip counts are not metrics. They are the noise you report when you do not have the signal.
I run fractional PR for Web3, DePIN, AI, and cybersecurity founders, and every engagement eventually surfaces the same board question: what are we getting for this? It is a completely fair question, and the honest answer is that most PR measurement frameworks fail it badly. They count clips. They report impressions from a wire distribution no journalist read. They attach an AVE number, which is a fabricated dollar value that PR trade bodies have tried to retire for twenty years. None of that maps to pipeline. None of it tells a founder whether their share of voice is growing relative to the protocols they compete with. This playbook is the measurement system I actually use, from how to define your competitive SOV set to what you put in the quarterly board slide.
Why most PR metrics are wrong before you start
The standard clip report you get from a PR agency has three lines: number of placements, total reach, and advertising value equivalent. All three are misleading in different ways.
Placement count without tier weighting is useless. A CoinDesk feature and a repost on a DeFi aggregator blog do not carry the same weight in a founder's category position, with investors, or with AI engines indexing the web. Counting them equally is the same as treating a Series A lead and a $5K angel check as equivalent in your cap table because they are both "investors."
Total reach is borrowed from advertising, where it makes sense because you paid for the eyeballs. In earned media, the number is a modeled estimate built on outlet traffic data that is often six to eighteen months stale, and it tells you nothing about whether the right people saw the piece. An article in Blockworks read by 40,000 institutional crypto allocators is worth more than a wire pickup reaching a claimed 200 million "potential impressions."
AVE, advertising value equivalent, is the worst of the three. It converts editorial coverage into a dollar figure by asking what you would have paid to buy that space as an ad. The number is always large and always flattering. It is also meaningless, because editorial coverage and advertising are not substitutes, which is the entire reason you are running PR in the first place. If they were the same, you would just buy the ad.
The four metrics that actually matter
1. Share of voice against a named competitive set
Share of voice is the percentage of total media mentions in your category that name your protocol or founder, relative to a fixed set of competitors you define at the start of the engagement. You pick four to six competitors, pull monthly mention volume from a monitoring tool like Meltwater, Brandwatch, or even a well-configured Google Alerts setup for smaller budgets, and track the ratio over time. The goal is directional: are you gaining share, holding, or losing ground? In the launches I have run, a focused 90-day sprint on tier-1 placement typically moves SOV by 8 to 15 percentage points in a well-defined niche. Across a thin niche, it can move more. The deeper breakdown is in the dedicated share of voice guide for crypto.
2. Tier-weighted placement volume
Not all placements are equal, so you weight them. The tier system I use for crypto and Web3:
| Tier | Example outlets | Weight | Why it matters |
|---|---|---|---|
| Tier 1 | CoinDesk, Cointelegraph, The Block, Decrypt, Forbes, Bloomberg Crypto, TechCrunch | 10 points | Indexed by AI engines, read by investors and institutional buyers, sets category framing |
| Tier 2 | Blockworks, The Defiant, CryptoSlate, Benzinga Crypto, Protos | 5 points | Strong community reach, good for product and ecosystem depth, cited in SOV counts |
| Tier 3 | Regional outlets: BloomingBit, TokenPost, CryptoTimes JP, Inc42, Coincu | 2 points | Multilingual syndication, regional investor and community reach, supports local market entry |
| Tier 4 | Aggregators, wire republishes, crypto news blogs | 0.5 points | Minimal independent weight; only useful in volume for SEO link equity |
At the end of each month, you sum weighted scores and compare to the prior period and to the same period for named competitors where data is available. For a launch like RARI Chain's mainnet, which generated 11 tier-1 placements in 24 hours, a weighted score for that single day would be 110 points. That number is a meaningful signal. A clip count of 11 is just a number.
3. Search lift and branded query growth
Sustained PR coverage drives branded search volume. This is the channel where credibility compounds most visibly, and it is almost never reported in standard PR decks. Pull Google Search Console data for branded queries (your protocol name, your founder's name, your product name) and overlay it against your PR activity calendar. The lag is typically six to twelve weeks: a spike in tier-1 coverage in month one tends to show up as search lift in months two and three. When MANTRA Chain landed its CoinDesk exclusive on the $11M raise with a Middle East RWA angle, branded search for "MANTRA Chain" tracked upward for eleven weeks after publication. That is a compounding asset a press release can create once but an ongoing PR program sustains.
4. Inbound quality and attribution
The hardest metric to pull cleanly, but the most persuasive one for a board. Track inbound investor inquiries, partnership requests, and qualified sales leads, and ask every new inbound how they found you. In the months following a major PR push, "I read the CoinDesk piece" or "I saw you on The Block" is a directly attributable source. I track this in a simple log: date of inbound, source named, tier of coverage that drove it, outcome. Over six months it builds a defensible map from PR activity to pipeline. For the Gaia AI campaign, which placed a Forbes "Stripe for AI agents" piece alongside Decrypt and Benzinga coverage and a six-podcast tour, the founders tracked fourteen investor introductions in eight weeks that named the Forbes piece as the first touchpoint. That is pipeline, not impressions.
Share of voice: how to actually measure it
SOV calculation has three steps.
Step one: define the competitive set. Pick four to six protocols or companies that compete for the same investor attention, same community, and same journalist beat as you. Do not include every player in the broader sector. The set should be tight enough that a move in your SOV ratio is a meaningful signal, not just sector noise.
Step two: pull mention volume from a consistent source. Use one tool and stick with it. The absolute number matters less than the trend line. Pull monthly, not weekly, because weekly volume has too much event-driven noise. Include brand mentions, founder name mentions, and your primary product terms. Exclude your own owned media from the count unless you are specifically tracking content reach.
Step three: calculate ratio, not raw number. Your SOV for a given month is your mention count divided by total mentions across the entire set, expressed as a percentage. If your set has five competitors plus you, and total mentions across all six are 2,000 in a month, and 340 of those name you, your SOV is 17 percent. That is the number you track. If it is 22 percent the following month, that is a 5-point gain and worth reporting. If it dropped to 14 percent, something in the competitive landscape shifted and you need to know what.
What to report to a board
Board decks get one PR slide, two at most. The mistake founders make is treating it like a clip report with logos. Boards do not care which outlets you were in. They care whether the coverage is moving the metrics that matter to the business. Here is what goes on the slide.
- SOV trend line: three to six months, your percentage versus the average of your named competitors. One chart, clearly labeled. If you are gaining share, the line goes up. If not, explain why and what changes.
- Tier-1 placements in the period: count only, with a brief note on the most significant one and what it drove. "4 tier-1 placements including a CoinDesk feature, drove 3 named investor inbounds in the following six weeks" is one bullet that tells a board everything.
- Branded search lift: a before/after or trend line from Google Search Console. Simple and direct. This is the compounding asset metric and it resonates with operationally literate board members who understand SEO.
- Cost per credible placement: total PR spend in the period divided by tier-1 and tier-2 placements only. This is the number that benchmarks efficiency. If you are running fractional PR at $5,000 to $12,000 per month and landing four to six tier-1 placements, the math is defensible against a full agency at $15,000 to $45,000. The full cost comparison is in the crypto PR cost breakdown.
The vanity metrics founders keep asking for, and why to drop them
I get asked to include impression counts and clip totals in reporting more often than I would like. The pressure usually comes from investors who are used to digital marketing dashboards where impression volume is a meaningful signal because it is tied to spend. In earned media it is not. Here is the table I use to reframe the conversation.
| Metric | What it claims to measure | Why it misleads | Replace with |
|---|---|---|---|
| Total impressions | How many people saw the coverage | Based on modeled outlet traffic, not actual readers; does not weight audience quality | Tier-weighted placement score |
| Clip count | How much coverage you got | Treats a wire republish and a CoinDesk feature identically | Tier-1 + tier-2 count separately |
| AVE | Dollar value of coverage | Fabricated: editorial and advertising are not substitutes | Cost per credible placement |
| Social shares of coverage | Virality / engagement | Vanity metric with high noise; does not correlate with investor or buyer behaviour | Branded search lift |
| Number of syndications | Amplification of a release | Wire syndications mostly reach aggregators, not journalists or investors | Named journalist pickups from the release |
The AI-search layer: a new metric worth tracking
In 2026, there is a fifth metric that is just beginning to appear in serious PR reporting: AI citation share. This is whether your brand, founder name, or product appears in AI-generated answers when a buyer or investor asks a relevant question in ChatGPT, Perplexity, Google AI Mode, or Gemini. It is the generative-engine equivalent of organic search rank, and it is driven by the same underlying assets: tier-1 editorial coverage, bylined founder content, cited data and expertise, and entity consistency across the open web.
You cannot yet pull this from a single dashboard the way you pull Google Search Console data. The practical method is a weekly sample: pick ten to fifteen questions a potential investor or buyer would ask in your category, run them in two or three AI engines, and log whether your brand appears and in what context. Over three months you have a trend. The GEO primer covers the underlying mechanics. The short version: tier-1 editorial placements and bylined expert writing drive AI citation share the same way they drive organic authority. It is not a separate channel from PR, it is what PR was already building, now made measurable in a new surface.
Putting the framework together: a 90-day reporting cadence
The measurement system only works if you are consistent about cadence. Here is the rhythm I run for Web3 PR campaign clients.
Weekly (10 minutes): log new placements to the tracking sheet with tier, link, and any inbound attribution noted. Pull AI citation sample. Flag any notable competitor coverage.
Monthly (30 minutes): calculate SOV ratio for the month, update the trend line, sum weighted placement score, pull branded search data from Search Console, note inbound count and any attributable sources.
Quarterly (for board or investor update): build the one-slide summary: SOV trend, tier-1/2 placement count, branded search lift chart, cost per credible placement, and two to three named inbound attributions. One page. Clear numbers. No clip counts, no impressions, no AVE.
PR is narrative architecture. The architecture does not announce itself. It compounds quietly in search rankings, in investor memory, in AI engine citation lists, and in the SOV ratio that eventually tells you whether your category is yours or someone else's. The founders I work with who run this measurement system for six months stop asking whether PR is working and start asking how to scale what is already compounding. That is the right question, and the metrics above are what get you there.
Frequently asked questions
Ready to build a measurement system that holds up in a board room? Start with the SOV methodology for crypto, then the cost benchmarks to calibrate spend against outcomes. The full framework library lives in the playbook index.