Medical Device Press Release Distribution Benchmarks 2026: What Six Campaigns Actually Produced

TL;DR

Six press releases. 1,159 recorded pickups. 307 distinct websites.

VayoMed Research analyzed six PR Newswire campaigns run between May 26 and August 5, 2026 for two specialist life-sciences companies — a medical-device manufacturer and an independent ultrasound-service company 1. Then we went further than the wire report: we swept every pickup URL for liveness 2, audited the link markup on 240 of the pages 3, and cross-referenced all 307 pickup domains against roughly 2,300 tracked AI-assistant answers for three life-sciences brands 45. This is, as far as we can find, the only campaign-level public benchmark of its kind — none of the pages ranking for "press release distribution benchmarks" publishes real campaign data at all.

What the measurements show:

• The distribution network saturates after the first release. The first release reached 174 distinct domains. The next four added 4, 10, 2 and 2 new domains respectively — while each still reported roughly 175 pickups. A monthly wire program republishes to substantially the same core of about 165 sites; 85.5% of all 1,159 pickup rows sit on the 163 domains that carried every one of the six releases 1.
• 75.8% of reported "traffic" was crawler activity, not people — and PR Newswire's own guidance concedes that crawler interactions "may register as counts, skewing numbers" 6. Human-facing views totaled 8,375 of 34,543 recorded traffic actions 1.
• "Potential audience" is mostly one Yahoo! Finance row. The field is the summed monthly-visitor estimate of the host websites. In five of six campaigns, a single Yahoo! Finance placement supplied 80.6%–87.2% of the reported 49–53 million 1.
• The placements are durable — but their destination links are qualified tracking redirects. Zero of 1,159 pickup URLs had gone 404 at 4–75 days after release 2. All 720 destination links we sampled carry rel="nofollow" or rel="sponsored", and none of 450 checked links points at the client's own domain — every one routes through the wire's click-tracking redirect 3. Treat them as distribution links, not conventional followed editorial backlinks.
• The syndication tail is invisible in AI answers. Across two measurement providers and three tracked brands, not one of the 307 pickup domains appeared as an AI-answer citation source; 98.5% of tracked citations pointed at the brand's own website 45.
• More pickup did not mean more clicks. The campaign with 281 pickups and a 107.3M potential audience produced 48 tracked clicks — the cohort minimum. The campaign with the fewest unique pickups produced 145 — the maximum 1.

The conclusion the data forces is not that wire distribution is worthless. It is that the release is a distribution event, and the owned website is the asset. Everything measurable — the links, the citations, the durable value — resolves to the destination, not the placements.

1. The six-campaign benchmark

Takeaway: the middle campaign in this cohort produced 177 unique pickups, 5,253 recorded traffic actions, 84 tracked clicks and 131 journalist views. Those are reference points for diagnosing your own release — not promises, and not outcomes.

The six releases ran from May 26 through August 5, 2026. Three represented a medical-device manufacturer building an international evidence library; three represented an independent ultrasound-service company addressing distributors, service businesses and hospital stakeholders. All six were distributed through PR Newswire, and all figures below come from the platform's own Visibility Report exports, refreshed August 7, 2026, plus the full exact-match pickup detail files 1.

Cohort totals: 1,159 unique pickup-detail rows (1,312 as reported by the dashboards), 34,543 traffic actions, 514 tracked click-throughs, 771 journalist views. Medians: 177 pickups, 5,253 traffic actions, 83.5 clicks 1.

Two reconciliation lessons carried over from the raw exports. First, campaign C's dashboard reported 323 placements, but 153 of those rows were duplicates created by a correction re-publication; the detail export consolidates to 170 unique rows. Report both numbers — "323 reported placements; 153 correction duplicates consolidated to 170 unique rows" — rather than silently choosing the flattering one. Second, the traffic field reproduces exactly as the sum of four components (release views, release crawler hits, multimedia views, multimedia crawler hits), and journalist views sit outside it. A benchmark is trustworthy when another analyst can reproduce not just the final number but the reason it differs from the dashboard.

The two company sets had different shapes: the manufacturer's three releases produced 518 unique pickups, 18,434 traffic actions and 332 clicks; the service company's three produced 641 pickups, 16,109 traffic actions and 182 clicks. More pickup on one side, more action on the other. The releases differ in subject, timing, links and media, so the honest observation is narrow: pickup volume did not determine the action profile 1.

2. Six releases, 307 domains: the saturation curve

Takeaway: after the first release, a monthly wire program adds almost no new websites. What repetition buys is consistency on a stable network — not incremental reach.

This is the question a buyer actually needs answered before signing a twelve-month wire commitment: what does month six buy that month one didn't? The cohort answers it with a curve.

In chronological order, cumulative distinct domains ran 174 → 178 → 188 → 190 → 192 → 307. The first release landed on 174 distinct websites. The second added 4 new ones. The third added 10, the fourth 2, the fifth 2 — each while reporting roughly 175 pickups. Across the first five releases, 878 pickup rows resolved to just 192 distinct domains. If each release had reached a fresh set of sites, six releases would have produced roughly 1,062 distinct domains; the observed count is 307, or 29% of that 1.

The composition is starkly bimodal. 163 domains appeared in all six campaigns; 132 appeared in exactly one; only 12 sat anywhere in between. That is the signature of a fixed automated syndication network plus occasional one-offs — not of editors making monthly decisions. The 163-domain core carried 991 of the 1,159 rows (85.5%), and 150 of those 163 core domains are dedicated syndication subdomains of the form pr.<publisher>.com or smb.<publisher>.com 1.

The sixth release is the exception that explains the rule. Campaign F jumped the cumulative count from 192 to 307 — because it ran a different circuit: 212 AP-network outlets instead of 45, adding 114 broadcast-station websites the other five campaigns never touched. New domains come from changing the circuit, not from repeating it.

Read constructively, saturation is not an indictment. A stable network that republishes your approved language every month has real value: the same entity facts, product terminology and canonical links repeated across ~165 public URLs, month after month — a consistent public record that supports verification and entity recognition. What the curve forbids is the mental model in which month six delivers 177 new placements. It delivers the same shelf, restocked.

3. Traffic: views plus machines

Takeaway: the traffic tile mixes people with crawlers, three to one in this cohort. PR Newswire's own guidance says so — the problem is that the caveat lives in a resource article while the dashboard shows one big number.

Across the six campaigns, the reported traffic field decomposes exactly into four components 1:

• Release views: 3,369 (9.8%)
• Release crawler hits: 12,420 (36.0%)
• Multimedia views: 5,006 (14.5%)
• Multimedia crawler hits: 13,748 (39.8%)

Human-facing views total 8,375, or 24.2% of the 34,543 headline figure. Crawler hits total 26,168 — 75.8%. The platform's Brazilian-edition FAQ states the definition plainly: traffic includes total views and web-crawler hits 7.

Two ratios can be computed, and both need their method in their name. Tracked clicks divided by total traffic actions is 1.49% — but three quarters of that denominator is machines. Clicks divided by human-facing views is 6.14% — closer to meaningful, but still not a deduplicated human click-through rate, because views and clicks can both repeat. Neither number should be casually compared with an email click rate or an ad CTR.

Here is the important part: none of this is a gotcha. PR Newswire's own measurement guidance already concedes every element of it 6:

"Pickup confirms that a website has published your content, but it can't tell you that a human read it."

"Automated feeds often syndicate content to pages with low traffic to help with broad visibility. Plus, crawler and content management system interactions may register as counts, skewing numbers and making a press release look more widely distributed than it is."

The same page calls platform analytics "the first layer of data." The vendor is candid; the dashboard is not. A tile that says "4.5K traffic" next to a tile that says "52.6M potential audience" invites exactly the misreading the vendor's own resource library warns against. The report module's data is generated by "PR Newswire's proprietary analytics technology" 8 — definitions are the vendor's, and they can change. Our contribution here is simply to put the number and the caveat in the same place, with six campaigns of arithmetic attached.

4. "Potential audience" is one Yahoo! Finance row

Takeaway: the potential-audience figure is the summed monthly visitor count of the websites that carried the release. In five of six campaigns, more than 80% of it came from a single Yahoo! Finance placement.

Of the four headline metrics, potential audience is the largest and the least understood. Reconstructing it from the row-level data settles what it actually is: each pickup row carries a monthly-visitor estimate for the host site (sourced from SimilarWeb), and summing those row values reproduces the campaign's reported potential audience exactly in five of six campaigns 1. "Potential audience: 52.6 million" therefore means: the websites that carried this release collectively receive about 52.6 million visits a month. It does not mean 52.6 million people could plausibly have seen the release — every visitor to any host is counted whether or not they ever saw the page, and visitors to multiple hosts are counted multiple times.

And the sum is extraordinarily concentrated:

• A single Yahoo! Finance row (42.5–43.0M monthly visitors) supplied 80.8%, 87.2%, 83.1%, 80.6% and 86.7% of the reported figure in campaigns A through E, and 40.1% in F 1.
• Across the pooled cohort, the top 5 rows of 1,159 account for 50% of the summed total; the top 34 rows account for 90%.
• The 474 placements on sites drawing under 5,000 monthly visitors — 40.9% of all pickups — contribute 0.126% of the total.
• 394 rows (34.0%) carry no audience estimate at all and contribute zero.

The median host in this cohort draws 1,193 monthly visitors; a quarter draw under 500 1. So the honest description of a typical campaign's audience math is: one giant portal, a dozen mid-sized sites, and hundreds of pages that statistically nobody visits.

This is not a novel failure mode — it is a named one. The current professional standard, AMEC's Barcelona Principles V4.0 (June 2025), instructs practitioners who report an audience-scale metric to "clearly and transparently define the limitations of the metrics and how they have been calculated" 9. And independent research has repeatedly shown how weak the link is between imputed reach and actual readership: analyses by risk researcher Jim Pierpoint found correlations between potential reach and measured article readership of just 0.40–0.68, with media-monitoring tools imputing reach above 250 million on 33 separate days in one company's year — more than the adult population of the United States 10.

Use the number for what it is — a rough scale descriptor of the outlet set — and never add it across campaigns, call it reach, or put it in a board slide without its definition attached.

5. What the placements are worth: durable, with qualified links

Takeaway: every one of the 1,159 pickup URLs was still resolving weeks after release — a real, measurable form of durability. And every single destination link on the pages we audited was a nofollowed redirect through the wire's own tracking domain. Both facts are true at once, and together they define what a pickup is worth.

The durability census

On August 9, 2026 we checked all 1,159 pickup URLs — a census, not a sample — between 4 and 75 days after their releases went out 2.

Not one URL had disappeared. 98.8% still resolved; 81.3% were verified to still carry the release text in the page's own HTML. The intuitive assumption — that wire pickups rot within weeks — is wrong in this window. The blocked 10.4% are almost entirely broadcast-station sites running bot protection; two were spot-checked through a browser-equivalent service and carried the full release, so a 403 means "closed to scripts," not "gone."

Nor is there measurable decay: the 75-day-old campaign verified at 89.4%, the 11-day-old at 91.8%. Age explains none of the variance — host type explains nearly all of it. Placements on dedicated pr.*/smb.* syndication subdomains verified at 99.6% and 98.1% respectively; everything else at 14.6%, dominated by bot-blocking stations and one structurally weaker category 2:

One pickup in thirteen (7.9%) is not an article page at all. It is a rolling "latest news" feed URL with a query key, where the release is injected client-side by a PR Newswire-hosted JavaScript widget — we confirmed the widget (tools.prnewswire.com/.../landing.js) on 21 of 26 such hosts probed 2. A human with a browser sees the release; a crawler reading served HTML — which is what most search indexers and AI retrieval pipelines consume — sees a page with no release text on it. Those placements count as pickups; they cannot function as durable public evidence.

The link-equity audit

Durability would matter more if the placements passed anything to the company's own site. They do not — at all, anywhere in this cohort 3.

We parsed 240 randomly sampled live pickup pages (40 per campaign) and found 720 links to the campaigns' destinations. All 720 — 100% — carry rel="nofollow" or rel="sponsored". These are explicitly qualified links, not ordinary followed editorial backlinks. On an independent 150-page sample, we checked where the links actually point: 0 of 450 point at the client's domain. Every single one routes through edge.prnewswire.com/c/link/..., the wire's click-tracking redirect, with the client's URL buried as an encoded parameter.

That one mechanical fact explains two things at once: it is why the platform can report 514 tracked clicks — every click passes through the wire's counter — and why these links should not be valued as conventional link-equity backlinks. The tracked-click metric and the qualified-link markup are the same engineering decision seen from two sides.

This is also exactly what Google's rules require. Google's spam policy names press releases explicitly: links "with optimized anchor text in articles, guest posts, or press releases distributed on other sites" are link spam unless qualified with rel="nofollow" or rel="sponsored" 11. Google separately says links carrying these attributes will generally not be followed 11. The cohort complies perfectly. Anyone selling wire distribution as conventional followed-link building is selling a policy violation that the wire itself has already engineered away.

One more mechanical finding: Google's guidance says the "most effective solution" for syndicated copies is for partners to block indexing 12. The network does nearly the opposite — 97.9% of the sampled copies permit indexing, and 238 of 239 canonical tags point at the syndicated copy itself rather than the original release. Hundreds of near-identical pages, each declaring itself canonical. We measured markup, not rankings, so we make no claim about how these pages actually perform in search — but a reader should stop expecting either link equity (there is none) or tidy canonical consolidation (there is none).

6. Where the release actually landed

Takeaway: 90% of placements were US websites, and 2.8% sat on sites the platform itself classifies as Medical/Healthcare. Distribution breadth is not audience fit — and "global distribution" needs expectation-setting before the first release.

Of the 1,159 pickup rows, 1,044 (90.1%) are United States placements; 106 are tagged "Global"; nine rows cover everything else — Canada 4, India 2, United Kingdom 2, Singapore 1 1. Every row is classified Online; the 114 "Broadcast Media" rows are station websites, not airtime. For a life-sciences company buying international visibility, the observed footprint of a standard US national circuit is a US-domestic one. PR Newswire's boilerplate describes a network of more than 500,000 newsrooms, websites, feeds, journalists and influencers across 170+ countries 13 — the network's advertised size and one campaign's observed 307 domains are both real numbers; the gap between them is the expectation to set.

The source-type mix: Newspaper 823 (71.0%), Online News Sites & Other Influencers 178 (15.4%), Broadcast Media 114 (9.8%), Trade Publications 9 (0.8%). By industry label: Media & Information 83.3%, Financial 8.8%, Medical/Healthcare 2.8%, Business Services 2.1% 1. A medtech release achieves its breadth almost entirely on general-media syndication endpoints, not specialist clinical or procurement publications.

That is neither good nor bad — it reflects two different jobs. Broad syndication puts the approved announcement on a large stable public record; targeted editorial penetration puts the story in front of a specific clinical or commercial audience through a publication exercising judgment. The wire buys the first. The second requires a genuine news hook and direct media work, and its results appear in the journalist-views field only as a faint precursor (771 across the cohort — platform-recorded views, not articles). Measure both layers; never present one as the other.

7. Pickup, clicks and destination design

Takeaway: placement count and click count moved independently — the widest campaign produced the fewest clicks. What the clicks did show is a near-even three-way split across identity, action and evidence destinations.

Campaign F: 281 pickups, 107.3M potential audience, 212 AP outlets — and 48 tracked clicks, the cohort minimum. Campaign C: 170 unique pickups, the cohort minimum — and 145 clicks, the maximum 1. Six campaigns with differing subjects, dates and observation windows cannot establish what causes clicks (F was also only four days old at snapshot). But they conclusively disprove the monotonic rule: more pickup does not automatically produce more downstream action. If your scorecard is a placement count, you are optimizing a number this data says is disconnected from response.

Where the 514 clicks went: homepage 190 (37.0%), contact pages 170 (33.1%), content/report pages 154 (30.0%) 1. The near-even split says readers arrive with three different jobs — verify the company, act, or go deeper into evidence — and a release should serve all three with distinct, descriptive links: a canonical identity page, an evidence destination, and a governed action path. (The split is a design argument, not a proven cause of volume: the two campaigns that offered no content link were not the low performers.)

One planning note per objective: a corporate-milestone release can reasonably prioritize accurate broad pickup and durable URLs; an evidence release should be judged on content-destination clicks and what happens after them; a direct-response release on qualified contact actions; a media-development release on journalist activity and subsequent conversations. Decide the objective before the release goes out, and score against that — otherwise the biggest number on the dashboard becomes the objective retroactively.

8. What AI answers cite

Takeaway: in roughly 2,300 tracked AI-assistant answers about three life-sciences brands, the 307-domain syndication tail appeared as a citation source exactly zero times. What got cited — 98.5% of the time in the largest sweep — was the brand's own website.

PR Newswire's current positioning line for the AI era is "AI doesn't browse – it cites. If your brand isn't part of the source layer, it won't be part of the answer" 14. Our data refines that slogan with an important observation about which URLs constitute the source layer.

What we tracked: AI-answer visibility exports for the two PR clients in this cohort (collected August 5, 2026, via the Peec AI customer API — 1,604 captured answers across 57 prompts for the ultrasound-service company; 222 answers across 25 prompts for the device manufacturer) 4, plus an independent 1,440-keyword citation sweep across six generative surfaces for a third life-sciences brand (collected August 1, 2026 — 458 citation rows) 5.

What we observed:

• In the 1,440-keyword sweep, 451 of 458 citation rows (98.5%) pointed at the brand's own website. Four pointed at press-release URLs — the wire's canonical release page and its Yahoo! Finance copy — and all four occurred on queries for the company's name 5. The 255 unbranded, category-level queries produced zero release citations; the one owned-site long-form report page in the set was cited on an unbranded category query.
• For the ultrasound-service company, none of the 44 domains in the sampled citation evidence — and none of the top-20 most-cited domains out of 702 tracked — is one of the 307 pickup domains. The brand's own website was the 6th most-cited source domain overall (415 citations) 4.
• For the device manufacturer, exactly one wire-related citation appeared: the release's canonical page on prnewswire.com, surfaced on a distributor-intent prompt. None of the other 306 syndication domains appeared 4.

The pattern is coherent with the mechanics measured in Section 5: the syndication copies are duplicate pages on ~1,200-visitor hosts, one in thirteen of them not even present in served HTML, all of them self-canonicalising, none of them passing links. The canonical wire page and the brand's own site are structurally strong citation candidates; the tail is structurally weak. We present that as coherence between independent measurements — not as a causal chain.

Three limitations, stated plainly: the citation-evidence samples are partial (the per-prompt evidence blocks record about five sources per prompt, so "zero observed" is a statement about sampled evidence and top-20 rankings, not an exhaustive scan); the windows are short (one day to one week); and AI surfaces are non-deterministic and change. For sector context, Muck Rack's November 2025 analysis of over a million AI-cited links found that when AI tools cite newswire content at all, GlobeNewswire accounted for 61% of newswire citations, PR Newswire 27% and Business Wire 12% — vendor-adjacent research, but directionally useful 15.

What a reader should do is measurable without buying anything: run your priority category queries through the AI surfaces your buyers use, record which URLs get cited, and note whether they are your pages, your wire copies, or your competitors'. PR Newswire itself now advertises an LLM-visit metric in its report module 8 — ask your vendor for it.

9. What a release costs per outcome

Takeaway: the tier-1 wires do not publish prices — itself a finding — and at credible price points, a tracked click from wire distribution costs somewhere between about $2 and $36. The spread is dominated by vendor tier and by 3× campaign-to-campaign variance.

The pricing landscape divides cleanly into two worlds. The self-serve services publish rates: EIN Presswire sells a single release for $149 (with multi-release packs working out to $66–$83 per release) 16; Newswire.com tiers run $399/$599/$659/$965 per release 17; ACCESS Newswire publishes $475–$1,205 per release 18; Cision's own budget brand PRWeb publishes $120–$480 19; Send2Press lists $99–$289 20. The tier-1 wires publish nothing: PR Newswire has no public pricing page at all, Business Wire's "pricing" page contains no dollar figures 21, and GlobeNewswire's own registration page says to contact them for cost 22.

The most defensible tier-1 numbers are partner-published: eReleases, PR Newswire's national-distribution reseller, describes the PRN rate card as $715–$865 for a 500-word local/state release and over $1,500 national, plus a $305 setup fee, and prices the bundled national value at $1,145 (400 words), $1,510 (500 words) and $1,875 (600 words) 23. Third-party rate cards exist for Business Wire (a base near $760 plus per-100-word and per-image fees) but their own publisher labels them "industry estimates, not gospel" — quote that disclaimer, not the numbers, as the state of the evidence 24.

Two vendor observations worth making explicitly: Newswire.com prices its tiers on "monthly readers" of 6–200 million — the same summed-host-traffic construct Section 4 dismantles 17. And none of the six campaigns' own contract costs appear in this analysis; we apply public price points to our measured outputs.

Against the cohort medians — 177 unique pickups, 83.5 tracked clicks, 1,291 human-facing views per release — representative price points produce:

Campaign-to-campaign variance moves every one of those figures by roughly 3× (pickup ranged 170–281; clicks 48–145) 1. Anyone quoting one precise cost-per-click for wire distribution is ignoring both the price opacity and the performance spread.

The sharpest economics come from Section 2's saturation curve: measured per new domain reached, the first release cost a few dollars per domain, while releases four and five — which added two new domains each — cost hundreds per new domain at any realistic price point. Under a "reach" theory of PR, the marginal economics collapse after release one. Under a "repetition and stable public record" theory, the same spend is defensible. Decide which theory you are buying before you sign, because the invoice is identical either way.

10. What to measure instead

Takeaway: the profession's own standard already names the failure mode in this article: substitution error — treating outputs as outcomes. Measure the layers separately, and say which layer each number lives in.

The measurement standard for this industry, AMEC's Barcelona Principles V4.0 (June 2025), puts it in one sentence: "Don't make substitution error – e.g. media clip volumes and tonality of coverage do not equate to the effect on the stakeholder audience" 9. A pickup count is an output. Treating 1,159 placements as an outcome is the named error. The same document requires disclosing how any audience-scale metric is calculated — which is all Section 4 does — and its channel list now explicitly includes "search data or AI search engine outputs," so measuring AI citations is not a novelty; it is the current standard 9. AMEC has since published dedicated GEO measurement principles (May 2026) 25, and industry commentary on its platform frames the question as "are we being mentioned, cited, and trusted by AI systems?" 26.

The industry is not there yet: in Muck Rack's 2026 survey of 971 PR professionals, nearly three-quarters called generative-engine visibility important, but 39% are not measuring it at all and 29% say nobody owns it 27. If your reports confuse these layers, you are in the majority — which is exactly why a clean ladder is a competitive advantage:

Operationally: freeze metrics at matched observation windows (the platform's own documentation notes pickup data accrues for up to 60 days 7, so day-4 and day-75 snapshots are not comparable); baseline branded search, referral sessions and form volume for a month before distribution; and keep an evidence inventory so each month's release answers one real reader question instead of manufacturing news. Report contribution narratives with timestamps, not single-touch attribution.

11. Methodology and limitations

The cohort. Six PR Newswire Visibility Report snapshots, refreshed August 7, 2026, for releases distributed May 26 – August 5, 2026 by two specialist life-sciences companies (three releases each). Client identities are withheld; campaigns are labeled A–F. Pickup analysis uses the full exact-match detail exports (1,159 de-duplicated rows; one campaign's 323 reported placements consolidate to 170 unique rows after 153 correction duplicates). Traffic reproduces exactly as release views + release crawler hits + multimedia views + multimedia crawler hits; journalist views sit outside it. Potential-audience analysis sums the row-level monthly-visitor estimates (SimilarWeb-sourced on 765 rows; 394 rows carry none) and reconciles to the reported figure exactly in five of six campaigns 1.

URL sweep. All 1,159 pickup URLs — a census — checked August 9, 2026 (15:30–15:32 UTC): HTTP GET, redirects followed, 10-second timeout, concurrency 10, one retry on transport failure, desktop-browser user-agent, randomized order. Liveness scored at two levels: HTTP status, and presence of a campaign-specific brand token in returned HTML (catching soft-404s). Blocked responses (403/429) are recorded as "not machine-verifiable," never as failures; two were independently confirmed live via a JavaScript-capable extraction service. Ages at sweep: 4–75 days 2.

Markup audit. 240 pages (40 per campaign) randomly sampled from the 942 verified-live pages, parsed for robots directives, canonical tags and destination anchors; link mechanics re-verified on an independent 150-page sample with a different random seed 3.

AI-citation observation. Two providers, three brands, ~2,300 captured answers, collected August 1–5, 2026 45. The per-prompt evidence samples are partial; "zero pickup domains observed" describes the sampled evidence and top-20 source rankings, not an exhaustive scan of every citation. No before/after design exists; nothing here is causal.

Limitations. Six campaigns, two companies, one wire vendor, 90% US placements, a 4–75-day window. Platform metrics are proprietary and vendor-defined. Third-party price figures are labeled as such and disagree with each other. Potential audience is never reach; crawler hits are never people; pickup is never editorial coverage. This benchmark describes what happened in this cohort; it is not an industry census.

Frequently asked questions

How many pickups does a press release get?

In this six-campaign medtech cohort, the median release produced 177 unique exact-match placements (range 170–281), resolving to a recurring core of about 165 syndication websites. The number depends almost entirely on the purchased circuit: the one campaign on a 212-outlet AP circuit reported 281 placements, while the five on a 45-outlet circuit each landed near 175 1. Treat any pickup benchmark as circuit-specific, not as a quality score.

Is press release distribution good for SEO? Do wire links count as backlinks?

Not as conventional followed backlinks in this cohort: all 720 destination links we sampled carry rel="nofollow" or rel="sponsored", and none of 450 checked links points directly at the client's site — each routes through the wire's tracking redirect 3. That matches Google's rules for qualifying distributed press-release links; Google says links carrying those attributes will generally not be followed 11. The realistic SEO-adjacent value is indirect: a stable, indexable public record of your approved language, and whatever earned coverage the story wins separately.

What does "pickup" actually mean?

A pickup is a recorded instance of your release being republished at a tracked URL — overwhelmingly automated syndication to a fixed network, in this cohort 92% of the recurring core being dedicated pr.*/smb.* subdomains 1. It is publication evidence, not editorial coverage. PR Newswire's own guidance states that pickup "can't tell you that a human read it" 6.

Is "potential audience" real reach?

No. Reconstructed from row-level data, it is the summed monthly-visitor estimate of the host websites — in five of our six campaigns, a single Yahoo! Finance row supplied over 80% of it, and the 40.9% of placements on sub-5,000-visitor sites contributed 0.126% 1. It describes the theoretical scale of the outlet set, double-counts overlapping audiences, and says nothing about who saw your release.

Do press releases show up in AI answers?

In our tracked data, occasionally — and specifically: the wire's canonical release page or its Yahoo! Finance copy appeared as a citation on queries for the company's name, while zero of the 307 syndication-tail domains appeared anywhere. 98.5% of tracked citations pointed at the brand's own website, and owned-site pages were what surfaced on unbranded category queries 45. Descriptive observations from small samples — but consistent across two providers and three brands.

How much does PR Newswire cost?

PR Newswire publishes no prices. Its national-distribution partner publishes the rate card at $715–$865 for a 500-word state release, over $1,500 national, plus a $305 setup fee, with bundled national values of $1,145–$1,875 by word count 23. Self-serve alternatives publish rates from $149 (EIN Presswire) to $965 (Newswire.com Feature) 1617. Against our cohort medians, that puts a tracked click at roughly $2–$36 depending on tier and campaign performance.

How long do syndicated release pages stay live?

Longer than the folk wisdom says: zero of 1,159 URLs in this cohort had gone 404 at 4–75 days, and the oldest campaign verified as well as the newest 2. The durable form is the article permalink on a syndication subdomain (99.2% verified live); the fragile form is the rolling feed page whose release exists only in a JavaScript widget — 7.9% of pickups, invisible to plain-HTML crawlers 2.

Conclusion: the release is the event; the website is the asset

Four findings survive every robustness check in this cohort.

First, the network is the product — and it saturates. Six releases produced 1,159 pickups but only 307 domains, with 85.5% of all placements on a fixed 163-domain core. Monthly distribution restocks the same shelf; new reach requires a new circuit.

Second, the headline metrics measure the machine, not the audience. Three quarters of "traffic" was crawler activity; more than 80% of "potential audience" was one Yahoo! Finance row. The vendor's own documentation concedes both mechanics — the discipline is putting the caveat and the number in the same sentence.

Third, the placements are durable while their destination links remain qualified. Zero of 1,159 URLs died within the window; all 720 sampled destination links carried rel="nofollow" or rel="sponsored", and all 450 independently checked links routed through the wire's redirect rather than the companies' own sites. A wire placement is a stable public copy of your approved language, not a conventional followed editorial backlink.

Fourth, almost everything observed in the largest citation sweep was owned. In ~2,300 tracked AI answers, the syndication tail never appeared; the brand's own pages drew 98.5% of citations and won the unbranded queries. The wire's canonical page earned its citations only when someone already searched the company by name.

The operational conclusion is the one all four findings point at: distribution amplifies an asset that must exist first. The canonical product pages, the evidence library, the contact paths, the entity-consistent public record — those live on the company's own website, and they are what search engines index, AI systems cite, and buyers click through to. That is why VayoMed operates PR as one of six services inside a single Done-for-You growth engine — website, content, LinkedIn, domains, AI-visibility monitoring and one press release per month with global distribution — for one $24,000-per-year subscription 28. The PR guarantee is the production and distribution cadence, never pickup counts, coverage, rankings, AI citations, leads or revenue: this report is precisely about why those cannot be honestly guaranteed — and how to measure what actually happens instead.

Sources

1. PR Newswire Visibility Reports for six life-sciences wire campaigns, May 26 – August 5, 2026, including full exact-match pickup detail exports (1,159 de-duplicated records), refreshed August 7, 2026 — VayoMed analysis, August 2026. Method in Methodology.

2. Liveness sweep of all 1,159 syndicated pickup URLs (census; HTTP GET, browser user-agent, content verification; includes client-side rendering probe of 26 rolling-feed hosts), conducted August 9, 2026 — VayoMed analysis.

3. Link and indexing markup audit of 240 randomly sampled live pickup pages, with link mechanics re-verified on an independent 150-page sample, conducted August 9, 2026 — VayoMed analysis.

4. AI-answer visibility exports for two tracked life-sciences brands, Peec AI Customer API, collected August 5, 2026 — VayoMed analysis.

5. AI-answer citation sweep across six generative surfaces, 1,440-keyword universe, collected August 1, 2026 — VayoMed analysis.

6. PR Newswire, How to Measure Press Release Performance, accessed August 9, 2026.

7. PR Newswire, Visibility Report FAQ (Brazilian edition), accessed August 9, 2026.

8. PR Newswire, Amplify Report Module, accessed August 9, 2026.

9. AMEC, Barcelona Principles V4.0, June 2025.

10. Memo, Potential reach does not equal readership (summarizing Jim Pierpoint's reach-vs-readership research), accessed August 2026.

11. Google Search Central, Spam policies for Google web search, updated May 15, 2026; and Qualify your outbound links to Google, accessed August 9, 2026.

12. Google Search Central, Fix canonicalization issues, updated July 10, 2026.

13. PR Newswire, corporate boilerplate (distribution-network scale: 500,000+ endpoints, 170+ countries; vendor claim), accessed August 9, 2026.

14. PR Newswire, PR Newswire Sets the Record Straight on AI Visibility: "Be the Source", accessed August 2026.

15. Muck Rack, More stats from "What is AI reading?", November 17, 2025.

16. EIN Presswire, published pricing, accessed August 2026.

17. Newswire.com (an ACCESS Newswire brand), published pricing, accessed August 2026.

18. ACCESS Newswire, published per-release pricing, accessed August 2026.

19. Cision PRWeb, published pricing, accessed August 2026.

20. Send2Press (Neotrope), published price list, accessed August 2026.

21. Business Wire, Business Wire Pricing: What to Expect (no rates published; quote-only), accessed August 2026.

22. GlobeNewswire (Notified), registration page ("press release distribution is a paid service… contact us"), accessed August 2026.

23. eReleases (PR Newswire national-distribution partner), Pricing, page last modified July 6, 2026 (partner-published account of the PR Newswire rate card).

24. Prezly, Business Wire pricing, updated August 7, 2026 (third-party estimate; the publisher's own disclaimer: "take these numbers as industry estimates, not gospel").

25. AMEC, GEO Principles, May 2026.

26. Kelly Costello (Meltwater), Rethinking PR Measurement in the Age of Generative AI, AMEC, July 21, 2026 (vendor authorship; published on AMEC's platform).

27. Muck Rack, The State of PR 2026 (971 qualified responses; fieldwork May 14 – June 12, 2026), July 2026.

28. VayoMed, Done-for-You Global Growth Engine for Life Sciences, accessed August 2026.

Original analysis: six PR Newswire Visibility Reports and 1,159 exact-match pickup records retained by VayoMed (refreshed August 7, 2026); URL census sweep, markup audit and AI-citation observations conducted August 1–9, 2026. Chart data files published alongside this report preserve every displayed series.