Does journalist outreach still earn links? A platform that got shut down and bought back says yes.

A tactic that survived its own funeral

Help a Reporter Out started as Peter Shankman's mailing list and grew into a service claiming over 800,000 sources and 55,000 journalists. Vocus bought it in 2010. Vocus merged into Cision in 2014, and HARO became a Cision brand for a decade. Cision folded it into a rebrand called Connectively through 2024, then on 8 November 2024 told users that Connectively would be permanently discontinued on 9 December 2024. HARO, in any form, stopped existing.

It didn't stay gone. On 15 April 2025, Cision sold the HARO brand to Featured.com, a competitor in the same expert-quote space. Featured relaunched it the next day, with investor backing from Great North Ventures, and started sending pitch emails to the old user base again within a week.

A company doesn't buy a brand it considers worthless. Somebody paid real money for the list, the name recognition and the habit thousands of PR people and marketers had already built around it. That's a stronger signal about ongoing demand than any single case study, because it's a company betting its own capital rather than a blog post making a claim.

What the pitch-to-link numbers actually say

The most detailed public data on HARO-style conversion comes from Jolly SEO's tracked run, an agency that runs done-for-you HARO/Connectively outreach and, notably, sells that exact service. Across 31,328 pitches sent, the agency vetted 3,738 responding domains and secured 2,600 links, a headline conversion of roughly one link for every twelve pitches. Break it down further and the picture is uneven: high-authority targets converted at closer to one link per 8 to 15 pitches, while the agency's more experienced writers landed one in every 4 to 10. The average domain rating of a placement climbed from 69.8 in January to 79.9 by November of the same tracking year, which the agency reads as its own targeting improving over time.

Treat all of that as vendor-published, not independently audited. The company that ran the outreach is the same company selling you the outreach. The numbers were first published in December 2020 and have been updated since, which means the underlying media landscape they describe isn't the one HARO operates in today. Nobody outside the agency has reproduced the sample. That doesn't make the numbers wrong. But read "one link for every 8 to 15 pitches on strong targets" as a plausible range from a single interested party, not a guarantee.

Why this dodges the risk Google is actively enforcing

Guest posting and journalist outreach get talked about as the same category of work. They aren't, and the difference matters more this year than it did two years ago.

Google's current spam policy documentation is explicit about what turns a guest post into link spam: "advertorials or native advertising where payment is received for articles that include links that pass ranking credit, or links with optimized anchor text in articles, guest posts, or press releases distributed on other sites." A freelancer publishing an original, editorially attributed piece under their own byline is treated differently and explicitly called out as acceptable, even across multiple publications. The line Google draws is not "did a link appear on someone else's site." It is whether the content was paid for, duplicated at scale, and built to pass ranking credit rather than to inform a specific publication's specific readers.

Google has enforced that line hard. In November 2024 it issued manual actions against major publishers, including Forbes, the Wall Street Journal, Time and CNN, for hosting third-party content like payday-loan reviews, casino write-ups and sponsored coupon pages that existed mainly to borrow the host domain's authority. That enforcement is not uncontested: the European Commission opened a Digital Markets Act investigation on 13 November 2025 into whether Google's policy unfairly penalizes news publishers running legitimate sponsored content, with EU antitrust chief Teresa Ribera saying the rules do not treat publishers "in a fair, reasonable and non-discriminatory manner." That fight is still open, and a violation finding could cost Google up to 10% of global revenue, so this is not settled law.

Journalist outreach sits outside that entire argument, because it doesn't involve you writing or paying for the article at all. A reporter is working on a piece, you supply a quote or a data point they didn't have, and they decide whether it earns a mention on editorial merit. No purchased placement. No duplicated article. No optimized anchor text to police, because you never touch the page. Whatever risk guest posting carries under Google's current enforcement, a real journalist deciding your quote is worth including doesn't carry it.

The AI-citation case is about mentions, not links

Our own backlinks and AI search piece already covers the correlation data: two large 2025 studies disagreed on how strongly link authority predicts AI-answer visibility, and both agreed that unlinked brand mentions correlate more strongly than the links themselves. That's the same dynamic we found looking at Reddit's citation share: a platform with almost no earned backlinks still dominated citations, because engines were reading the mentions, not crawling a link graph. Journalist outreach is the mechanism that produces both a link and a mention at once, so it's worth separating what evidence exists specifically for the earned-media side of it.

Muck Rack's "What Is AI Reading?" report, tracking more than one million links cited by AI models and last updated 16 February 2026, put journalistic sources at about 25% of all citations, with the majority of cited links overall coming from unpaid, earned coverage rather than brand-owned pages. Comparing its July 2025 data to December 2025, press release citations increased fivefold, and more than half of all citations observed traced back to content published in the previous 12 months. Earned coverage isn't just a nice-to-have in that dataset. It's a moving, growing share of what AI models actually quote.

A smaller but sharper study makes the mechanism explicit. Stacker tested 8 articles across 944 prompt-platform combinations and roughly 189 unique prompts against five AI platforms, publishing 9 December 2025. Left on a brand's own site, an article got cited in AI answers 7.6% of the time. Syndicated across third-party outlets, the same underlying story reached 34% total citation coverage, a 325% increase. That's the clearest available evidence that getting a fact or a quote placed somewhere other than your own domain does something for AI-citation odds that publishing it yourself does not, independent of any link at all. The sample is 8 articles, which is small enough that the direction matters more than the exact multiplier.

None of this is a controlled experiment. Every figure here is observational. It describes what happened across a sample, not what happens if you land one more journalist quote. Treat the direction as real and the size as unproven.

Guest posting, for comparison

Guest posting has not disappeared either. Aira's State of Link Building report has it as the third most popular link-building tactic, used by 47% of surveyed SEOs. But Ahrefs' own guide to guest blogging, citing a survey of over 500 bloggers across 239 published guest articles, found the average referral traffic per article was just 50 visits. Guest posting's case rests almost entirely on the link and the association with a known site, not on any meaningful stream of readers.

That's the real difference in what you're buying. A guest post is written and placed to earn a link, full stop, and Google now polices exactly how that link was obtained. A journalist quote is earned because an editor found it useful, and it produces a link, an unlinked mention, and the citation lift documented above, all from the same conversation.

What this changes about buying outreach

If you're buying link building on the promise that any tactic converts reliably or fast, every number in this piece should worry you. A one-in-twelve pitch-to-link rate from a single vendor's old dataset, a citation-lift study built on 8 articles, and a guest-post traffic average of 50 visits aren't the pitch a confident salesperson gives you.

But weighed against the alternative, journalist outreach is the more defensible half of the link-building budget right now. It produces the earned mention that the AI-citation research keeps landing on as the stronger signal. It doesn't touch the guest-post enforcement risk Google and the EU are actively fighting over. And unlike a rented placement, nobody can retroactively decide your quote was a policy violation, because you never owned or paid for the page it appears on.

Two things worth doing before spending anything here. Confirm the target publication is one an actual audience reads, the same qualification bar we use in Link Building, because a defunct or spam-farmed outlet earns you nothing whichever tactic produced the placement. And read what we found on measuring AI visibility before judging a campaign, since a placement's citation effect will not show up in your link report at all.

What we could not verify

The Jolly SEO/PitchResponse numbers are vendor-published by a company selling this exact service, first dated December 2020, and have not been independently reproduced by anyone outside that agency. Muck Rack's 25% figure covers journalistic sources specifically inside its own observed sample, not a universal share across every AI engine and query, and it doesn't isolate journalist-outreach-earned mentions from other kinds of press coverage a brand didn't pitch for. Stacker's citation-lift number is real and dated but rests on 8 articles, a sample too small to generalize a precise multiplier from, even though the direction (syndicated beats brand-only) is consistent with everything else in this piece. No study anywhere, vendor or independent, has run a controlled test connecting journalist-quote acquisition specifically to a measured AI-citation lift. Every citation figure in this piece is correlational or descriptive, not causal, and we're saying so rather than implying otherwise.

Sources

Wikipedia's sourced history of Help a Reporter Out, including the Vocus acquisition (2010), the Cision merger (2014), the Connectively discontinuation notice (8 November 2024, effective 9 December 2024) and the Featured.com acquisition and relaunch (15 to 16 April 2025). Google Search Central's Spam Policies for Google Web Search documentation, confirmed last updated 28 August 2026, for the guest-post and advertorial link-spam definitions. Search Engine Journal's reporting on Google's site reputation abuse enforcement and the EU's Digital Markets Act investigation, opened 13 November 2025. Muck Rack's "What Is AI Reading?" analysis of over one million AI-cited links, last updated 16 February 2026. Stacker's citation-lift study on brand-only versus syndicated citation rates, published 9 December 2025. Jolly SEO's published tracking of 31,328 HARO/Connectively pitches, run by Greg Heilers, first published December 2020. Ahrefs' guide to guest blogging, last updated 12 August 2026, citing Aira's State of Link Building report and Tim Soulo's survey of guest-article referral traffic.

Common questions