A Clutch score of 4.9 means a PR agency collected reviews from happy clients. It does not mean those clients were running the kind of launch you are running, or that the agency earned coverage rather than bought it, or that the retainer will feel like what was sold to you on the discovery call. Here is how to read review-site ratings for Web3 PR agencies without being misled by curated testimonials, and which questions to ask before you sign.
I run fractional PR for Web3, AI, DePIN and cybersecurity founders, and one of the questions I get most often before a founder decides how to hire is some version of: "this agency has a 4.9 on Clutch and fifty reviews, that must mean something, right?" It means something narrow. It means the agency was organised enough to ask clients for reviews after engagements, and that those clients were willing to say good things publicly. Neither tells you whether the agency can place a story about your specific protocol in CoinDesk, Cointelegraph or The Block, or whether their retainer is built around earned coverage or around press-release wire fees dressed up as outreach. For the full picture on how to choose, the Web3 PR agency checklist is the place to start. This piece is specifically about how to read review sites critically, and what they structurally cannot tell you.
What Clutch actually measures
Clutch is a B2B directory that scores agencies on four weighted factors: the number of reviews, the recency of those reviews, the average star rating, and the verified client spend. Agencies with higher spend and more reviews rank higher in the directory. The review itself is a structured interview Clutch conducts by phone with the client, then publishes as a case study. The questions are standardised: overall satisfaction, would-you-recommend, specific results mentioned.
That structure has two built-in constraints that matter for anyone evaluating PR agencies in particular. First, the client who gets interviewed is almost always self-selected: the agency asks its happiest clients to participate. A client who was frustrated, or who achieved modest results, rarely appears in the directory. Second, the results language in reviews is almost never audited. If a client says "they got us great coverage," that goes into the published review without Clutch verifying what outlets, what placements, whether they were earned or sponsored, or whether the coverage moved any commercial needle.
The specific ways Web3 PR reviews mislead
The pattern I see most often when founders share their agency shortlists: five agencies, all with 4.7 or above on Clutch, all with reviews mentioning "blockchain," "crypto" and "Web3." The reviews look identical in quality. But the actual work those reviews describe is not comparable. Here are the most common gaps.
Earned versus paid placements
A review that says "they placed us in CoinDesk" tells you nothing about whether that placement was a paid sponsored article, a branded content arrangement, or earned editorial coverage from a pitch a journalist chose to act on. These are categorically different outcomes with different credibility, different longevity and different costs, but review sites do not distinguish between them. Founders who do not know to ask end up comparing agencies that run very different models under the same "media coverage" label. The full breakdown of what those models cost is in the Web3 PR campaigns service page.
Vertical depth
A Web3 PR agency that ran a successful NFT drop campaign in 2022 and a DeFi protocol launch in 2023 is not the same thing as one that has genuine relationships on the DePIN beat, or that understands the narrative architecture of a cybersecurity firm entering the enterprise market. "Web3" is not a vertical. It is an umbrella under which at least a dozen distinct beats, journalist pools and editorial priorities sit. A review from a gaming token project tells you almost nothing about how an agency would run your AI infrastructure raise.
The vintage problem
PR market conditions change fast. An agency with a 4.9 Clutch score built on reviews from 2021 to 2023 was operating in a very different media environment: more reporters on the crypto beat, higher editorial appetite, looser standards on what constituted news. Coverage that was relatively easy to earn in that window requires significantly more narrative work now. Recency of the reviews matters more than the aggregate score, and Clutch does surface review dates, so check them.
What G2 and Trustpilot add, and where they fall short
G2 is software-first and its PR agency reviews tend to be thinner and less structured than Clutch. The reviewer verification is less rigorous, and the review text is usually shorter and less specific. For evaluating a PR agency, G2 adds little beyond a second sample of self-selected positive sentiment.
Trustpilot is largely unverified by default, meaning anyone can leave a review without having to demonstrate a client relationship. In a category as relationship-heavy and reputation-sensitive as PR, Trustpilot scores are almost entirely noise. Some PR agencies have high Trustpilot scores because they actively solicit reviews from happy clients; some have low scores because a disgruntled party left a review. Neither is meaningful for your evaluation.
| Platform | Review process | Verification | Useful for | Not useful for |
|---|---|---|---|---|
| Clutch | Phone interview, structured questions | Client relationship verified | Building a longlist, checking category history | Comparing output quality, earned vs paid distinction |
| G2 | Self-written, shorter format | Partial email verification | Spotting consistent complaints across reviewers | Evaluating PR craft or media relationships |
| Trustpilot | Self-written, open access | Minimal | Very little in B2B PR context | Almost everything |
| Direct reference | Your call with a past client | You are doing the verifying | Everything that matters: output, process, honesty | Scale: takes time, hard to fake |
The signals that actually matter in a review
When you do read Clutch reviews, here is what to look for instead of the star rating.
Named outlets, not category claims. A review that says "they got us into the top crypto publications" is almost worthless. A review that says "they placed a CoinDesk exclusive on our raise and a Blockworks feature on our mainnet within the same campaign" is specific enough to evaluate. Named outlets let you check whether they match your target list. Vague category claims do not.
Mentions of the actual deliverable process. Good reviews tend to describe how the agency worked: how they developed the narrative, how they prepped the founder for interviews, how they handled a rough patch when a story fell through. That kind of process description is hard to fake and suggests the client was genuinely paying attention to quality. Reviews that only describe outcomes are less trustworthy because outcomes in PR are easy to overstate.
Negative specificity. A review that says "everything was perfect" is less useful than one that says "the earned media was strong but their social amplification was weak." Specific criticism is the sign of a real engagement. Pure positivity is often the sign of a curated, relationship-favour review.
Review recency and campaign context. Look at when the review was written and what kind of project it describes. A 2021 token-launch review tells you the agency could operate in a bull market. A 2024 or 2025 review describing a raise announcement or a protocol launch in the current environment tells you much more about how they perform under current conditions.
The reference-check questions that surface real quality
No review site replaces a direct conversation with a past client the agency did not hand-pick for you. Most agencies will provide two or three references if you ask, and those references will be their best relationships. That is still useful, because even a friendly reference will tell you things the Clutch profile will not, if you ask the right questions.
- Which specific outlets did earned coverage land in, and was any of it sponsored or paid? A reference who is honest will know the distinction and will answer directly. If they are unsure whether coverage was earned, that is itself a signal about the agency's transparency.
- What happened when a planned placement did not come through? Every PR campaign has moments when a story falls apart, an embargo breaks wrong, or a journalist passes. How the agency handled that is more revealing than any success. Did they communicate early, adapt the narrative, find an alternative, or go quiet?
- How did the agency develop the core narrative for your launch? This question reveals whether the agency does real narrative architecture or just executes a template. If the reference describes a genuine discovery process, positioning work and a narrative rationale, the agency probably does the former. If they describe a boilerplate press-release draft that went out on PRWeb, they do the latter.
- What did you wish you had known before signing? Open-ended, and the most useful question on the list. A client who is willing to be honest will tell you about the onboarding, the reporting quality, the account management turnover, or the gap between what was sold and what was delivered.
- Would you hire them again for your current stage? Not "would you recommend them" in the abstract, but specifically: given what you are working on now, is this the agency you would call? That framing forces a more honest answer.
Beyond the references the agency provides, ask whether you can speak with a client whose engagement did not renew. Most agencies will decline, but the ask signals that you are a serious buyer, and occasionally you will get a useful introduction. The comparison between how an agency treats you before the contract and how it behaves when a client leaves tells you a great deal about how the relationship will actually feel at month six.
What to do with all of this when you are shortlisting
The practical flow for evaluating Web3 PR agencies in 2026, review sites included, runs roughly like this. Start with Clutch to identify agencies that have verifiably worked in your category: DePIN, AI infrastructure, cybersecurity, token launches, whatever your specific vertical is. Look for named clients or named verticals in the reviews, not just "blockchain" as a tag. Filter for recency: reviews from the last eighteen months are more relevant than anything from 2021. Read the review text for named outlets, process description and specific critique.
Then move off the platform entirely. Ask for a case study that matches your type of launch, not their best campaign. Ask for two references outside the curated list. Run the reference-check questions above. Ask the agency to walk you through how they would build a narrative for your specific announcement, before you sign anything. An agency that can do that clearly and specifically, without retreating to generic frameworks, is demonstrating the skill you are actually buying.
The best Web3 PR agencies for 2026 guide covers what to look for by agency type, from boutique earned-media shops to full-service firms running KOL campaigns alongside press. And if you are trying to decide whether a fractional operator or a full agency structure fits your stage better, the fractional vs agency breakdown covers that tradeoff in full.
The honest version of what reviews can and cannot do
Review sites are a reasonable starting point for a longlist and a poor tool for a final decision. The structural problem is not that agencies game them, though some do. The structural problem is that PR quality is genuinely hard to capture in a standardised review format. The difference between an agency that earns a CoinDesk exclusive and one that places a sponsored article on a mid-tier site and calls it "coverage" is enormous. The difference between a retainer where the senior person you spoke to on the discovery call does the work and one where your account gets handed to a junior six weeks in is enormous. Neither of those differences shows up in a star rating.
The founders I have seen make good agency decisions in 2026 are the ones who treated review sites as a filter, not a verdict, and who invested the time in the reference calls, the case-study review and the direct narrative conversation before committing. That process takes longer than picking the agency with the most stars. It is the reason you do not end up three months into a retainer realising you bought a PR agency's template, not a PR operator's attention.
Frequently asked questions
Evaluating your options before you commit? The best Web3 PR agencies for 2026 compares agency types by structure and output. The agency checklist gives you the full decision framework. And the full playbook library covers pricing, pitch guides and narrative strategy.