Skip to content
Get started

Digital PR for AI search: earn your place in the answer

What the higher-ups care about is revenue and leads. A link doesn't really mean anything to an executive, but authority in the places AI answers read supports what you launch.

Free. Public data only, no logins. Back in 2 business days.

One story, many earned links

Your story, with data worth citing

  • National newsNew data on your category
  • Trade press
  • Industry blogs
Links to your siteMentions AI answers read
Editors link because the story is worth citing, and that coverage is what AI answers read.

Clients

Their stories ran in

Outlet coverage from our client reporting. Logos belong to their owners; no endorsement implied.

In short

Green Flag Digital's digital PR for AI search is for brands whose buyers ask ChatGPT, Gemini and Google AI Overviews before they ask a salesperson. We turn original data into stories reporters cover, so AI answers find sources that name you. You pay in points, the way you buy AI tokens, and see every point monthly.

The graph, in plain words

The web is a graph. The crudest picture of it is a word cloud, and a more advanced one is a knowledge graph of how things relate to each other and how far apart they are. Pages are the dots, links and mentions are the lines between them, and AI models learned the world from that graph. AI answers that search the web still read it each time someone asks a question. In Google, getting a variety of high-quality links from different websites seemed to move the needle the most. There were a lot of studies in the field showing that third-party references, whispers and brand associations mattered too, but Google never clearly communicated it. Brands with strong lines into a topic are the ones AI answers tend to quote: a 2025 study found AI search shows "a systematic and overwhelming bias towards Earned media" in the sources it cites (Chen et al., arXiv). Don't become a brand for SEO's sake. Become a brand because being a brand is a moat.

You can see this in the answers themselves. Joe's read: "It's been fairly obvious for a while, for those who care to look at the citations, in the footer or inline. It's constantly evolving, but you see strong patterns where the answers reflect the source listicles they used and cited." "They really like listicles, and they like data points. I haven't seen them pull a lot of quotes unless they're really asked, which is interesting." That is one person's reading of a lot of answers. The larger studies point the same way:

  • Lists get cited. In an Ahrefs study of 750 buying-research prompts, "best X" blog lists made up 43.8 percent of all the pages ChatGPT cited, across every page type (Ahrefs, December 2025). The share moves fast: Seer Interactive saw listicle citations in ChatGPT fall 30 percent from December 2025 to January 2026, so treat any single number as a snapshot (Seer Interactive).
  • Other people's pages name you. Across 21,311 brand mentions in AI answers to more than 500 commercial buying and vendor-comparison questions, 85 percent came from sites the brand does not own, and nearly 90 percent of those were lists, comparisons or reviews (AirOps, October 2025).
  • Earned media dominates. Of more than 25 million links cited by ChatGPT, Claude and Gemini, about 84 percent were earned media, and journalism alone was 27 percent (Muck Rack, 2026).
  • Mentions beat links. Across 75,000 brands, mentions on the web tracked visibility in Google's AI Overviews about three times more closely than backlinks did (Ahrefs, May 2025).
  • Data and sources lift a page. In a lab test, adding statistics, quotations and source citations to a page each raised its visibility in AI answers, by up to 40 percent (Aggarwal et al., KDD 2024).

Buyers feel the same pull. A marketplace founder told us that the more PR they have, and the more links, brand mentions and earned media links they get, the better the authority of the website, and the easier it is to rank their other pages, like their location pages. He had already tried the usual tactics: some citations, guest posting and link insertions, the regular stuff people do. But he saw digital PR as a different side. It's a different animal.

Digital PR is how a brand earns those lines fairly: a data study a reporter builds a story on, a quote in the news while it is still news, a page worth linking to. Instead of traditional PR, which is more focused on the company overall, we focus on creating interesting content and stories that are data driven, and then we pitch them to the media to get coverage and backlinks. We pick a new topic, do research on it, and then pitch it to bloggers and journalists to see if they want to cover it and reference us as the source. Whenever you pair the word data with the source, it looks more natural, and they're happier to link to something that's data. When a reporter links to you, Google is essentially outsourcing a manual editorial review of your page to the person who gave the link. One marketplace founder told us that every day he sees a new AI site with ten thousand pages in his space that just scraped his data, with no authority and no history. That is the earned graph. We build it month by month and show you which parts are working, including which links from past years are still live.

What you get each month

  • Pitching every month to hand-picked reporters on your beats, sized to each story, showing each one we did our research on what they cover and what they care about. If a reporter is a freelancer, we pitch them, because freelancers usually bring stories to the editor. If they're staff, by default we pitch the editor, because staff writers are usually handed assignments by their editors. We look for a few outlets with a few writers who love to cover a topic all the time, because that gives us a signal that it's a big focus and a lot of people are interested.
  • Each quarter, a new data study, a refresh of a past hit and a seasonal piece, planned with you. Journalists don't want to be hit up every month for the same thing, but they're fine with a quarterly. We spoon-feed reporters the key insights, make it visual, and make the data quick and easy to grab, because our primary audience is reporters who can quickly get the idea and write their own story.
  • AI answer tracking: a set of category questions agreed in month one, checked monthly across the main AI engines
  • A monthly report with every placement, its points and link, AI citations, the total against plan and the past links still live. It answers a complaint we hear about PR: a founder who had hired traditional PR companies told us it was a lot of vagueness, and a lot of "well, we don't know."
  • Every study page we build or pitch passes our 14-point study checklist, which is written into the contract

How points work

A point is a unit of earned coverage. The plan sets how many points we owe you each month, and we size it on a call. Our number one job is to deliver media placements. The content backs up into that, so we start with the links and work backwards. We developed the point system because, depending on the budget you have, we can translate it into a set amount of points. Counting placements alone, seven this month and eight the next, does the work a disservice when some of them land on really good, high-traffic sites.

Why points, and not dollars or audience numbers? Joe's answer: "What is a link? What is a placement? That's the same as asking what a car is and how much it costs. It ranges from a five-hundred-dollar car that doesn't drive to a three-million-dollar car, and everything in between, mostly in the twenty to sixty thousand range. Likewise, unique monthly visitors is a very made-up number used in PR, and the earned media value number has been debated a lot. For those who watch sports, points are a way to keep track as you go, and the points are tallied at the end to see who wins. It also feels a bit more classy than talking money all the time." The measurement trade body agrees on the money part: AMEC's Barcelona Principles say advertising value equivalents "do not measure the value of public relations" (AMEC). The usual audience figure is the outlet's whole monthly traffic times an assumed 2.5 percent who see any one story (Muck Rack help center).

  • Scored on one scale. A mention on a niche site scores less than a feature in a national outlet, and an AI answer that cites or names you scores too. We define a placement's strength as a blend of the site's domain rating and the prominence of the publication. News coverage gives you high-authority links from high-traffic sites, which really shows that you're an authority. Any mix of studies, refreshes and reactive stories adds up on the same scale.
  • Top-ups. Add a block of points for a launch or a hot story, only when you ask in writing.
  • Cheaper in bulk. Buying three or six months up front lowers the price per point. We quote it on the call.
  • Points flex. Over plan in a month carries forward. Under plan, we owe it until it is delivered, and anything still owed carries into the next term. Another buyer told us the volume of coverage from a past agency was very low for what they had agreed, and for what they were expecting in return. Owed points are how we avoid that.
  • Spend points on specific work. If you would rather have a given piece, such as a refresh of a past post or a new data post, each piece has a fixed point cost, confirmed before we start. A smaller update can be an easier pitch than a whole new piece, and reports with the new year in the URL tend to convert a little better, because journalists can see they're truly fresh. There's definitely an argument for re-pitching some older stuff and trying to work that back in.
  • Six-month terms. Plans renew in six-month terms, so studies and seasons can be planned properly. The highest long-term value comes from trying different types of campaigns over six months, finding what works and scaling up that one approach, and some of the best placements take three to six months to land. Any one campaign can do really well or flop, but across a basket of four campaigns you're going to have one overperform, one underperform, and some in the middle. We review results and the next term's plan together before each renewal.

How does this compare with retainers and pay-per-placement firms? Read our plain guide to PR pricing and packages.

Plans

Every plan runs on the same engine, the same points and the same report. The plan decides how many points a month, and so how much work.

  • Data Study PR. Data studies and monthly pitching to hand-picked reporters
  • Data Desk. Studies at a larger scope, plus a reactive news desk
  • Category Index · Scaleup. A flagship index for your category, with content and AEO
  • Category Index · Enterprise. The same at enterprise scope

No plan bets everything on one story. The main thing we tell new clients is that we can't predict which campaign will hit. It's better to try a few different things than beat one to death when it isn't working. We usually present options across a few categories so everyone can decide together: for this topic we optimize for reach, and for another campaign we optimize for keyword traffic. If we go super relevant and tightly around what the pros care about, it will get trade publication links, maybe fewer of them. If you open it up, you can get broader coverage, so it's finding the right balance between the two. Beyond the links, you're getting in front of your actual audience.

Taste, process, then the machine

Every result starts with human taste: people who know your beat choose the story, the angle and the reporter. In a boutique firm, the result is a blend of our process, our team and our know-how, with curation, working with your team and being creative. A proven process keeps the quality steady, starting with the 14-point checklist every study passes before it is pitched. No matter what, everything needs to be run through a few different sources and then manually checked at the end. Rather than taking a weird idea and trying to spin it into a visual, we back every story up with data. If it's a more fun topic, then we'll be more creative. If it's more hard data, we find that reporters in that space want to see the charts. Journalists can read our narrative and poke holes in it, so we want to provide a really cool visualization, resource or stat list that they can't provide on their own. Our platform and AI tools do the heavy lifting in between, the research and the aggregating, while the human direction setting, the quality checks and the final editing are a small share by volume but end up being most of the differentiator. More of your budget turns into coverage instead of hours.

And the aim is bigger than the citation count. Joe's advice: "Try to think bigger than just chasing citations." The goal is a study that reporters and buyers find worth repeating, not one an engine happens to cite once. That is the asset.

Case study · A national moving marketplace

89% of that marketplace's links still live six months after they landed

Measured on the 1,388 links Ahrefs could see, using a standard survival curve.

How long earned links last

A national moving marketplace, 1,388 links

100%80%60%Day 06 mo1 yr2 yr

89% live at six months

Our survival analysis of Ahrefs link data · Oct 29, 2026
Proof from client work

The work, in numbers.

placements earned for a national moving marketplace since 2021
1,960
placements earned for a national moving marketplace since 2021About 4 in 5 are estimated to still be live today, after checking live pages against Ahrefs.

A national moving marketplace runs data-study digital PR with us; the proof cards above carry the dates and sources. We do not name clients in this offer's public material without their yes.

Results and coverage from our client reporting. Results vary by company, budget and market.

JRJoe Robison

Joe Robison

Founder and principal, Green Flag Digital

Joe sets the strategy on every engagement and makes the final call on what counts as evidence. A principal reviews the findings before any client reads them.

Joe Robison on LinkedIn

What clients say

  • “Green Flag consistently delivered high quality PR pieces that got mentions and links from some of the largest media brands on the internet.”
    Ryan Carrigan, Owner, moveBuddha · via Clutch

Reviews from our Clutch profile. Results vary by company, budget and market.

See where you stand.

Tell us where to look. We come back with where you stand today, who gets named instead of you, and the work we'd take first, then walk you through it in a 20-minute teardown.

Questions buyers ask

What is digital PR for AI search?

Digital PR for AI search is earned media planned for the sources AI answers quote. Original data and expert comment earn coverage, and that coverage gives ChatGPT, Gemini and Google AI Overviews independent sources that name your brand. The piece should stand on its own, so that in theory it doesn't matter who produced it, because the content is so good that it's worth covering. We don't want to take a super strong opinion and tell the story. We want to present the data and float some interesting concepts and ideas. A growth manager at a tax software company told us they liked the idea of having high-quality content on a website and then pitching that to journalists, because it has multiple benefits.

What is a point?

A point is a unit of earned coverage, scored on one scale that is written into your contract. One buyer told us a points system can seem a bit arbitrary, to be frank, which is why placements score by how strong the outlet and the mention are, AI answers that cite or name you score too, and the monthly report lists every point with its link.

Do you guarantee coverage or a specific outlet?

No. Reporters and AI engines decide what runs and what they cite, so no outlet or answer is promised. We won't say we can get you onto a given list, because we can pitch everybody and none of them may land. For every ten things we produce, there's going to be one thing that hits. The points in your plan are what we answer for: under plan, we owe the rest until it is delivered.

Can we buy more points in a busy month?

Yes, in top-up blocks for a launch or a story that is taking off. Buying months up front costs less per point.

How long is the contract?

Six months, renewing in six-month terms. Points still owed at the end of a term carry into the next one.

How much does digital PR for AI search cost?

It depends on the plan, which sets how many points a month we owe you. We size it on a call; the plans above show what each is built from.

How do you measure AI citations?

In month one we agree a set of questions your buyers ask about your category. We check them every month across the main AI engines and report which answers cite or name you, with the points each one scores. We look at the reporting as something useful that makes the work better, not something we have to do.

Is this the same as AEO?

No. Answer engine optimization works on your own site so AI engines can find and quote it. Digital PR for AI search earns coverage on other sites that name you. Many brands need both.

Do we need our own data?

No, though it helps, because your own data is the hardest thing for a competitor to copy. When a client has it, we lead with their data, because that's the proprietary fresh data, and then census data backs it up as an official source from outside. Without it, we run a survey or combine public datasets into something new. Running a few surveys around certain topics and then tying them all together in a bigger piece tends to do well. Another route is a ranking: a proprietary ranking based on our opinion, something the LLM can't go find on the web.

How is this different from your Digital PR page?

It is the same team and the same craft, sold as a points retainer with AI answer tracking built in. If you want the full menu of campaign types, our Digital PR page has it.

What do you most want from search right now?

Three taps, then your site. About a minute. Public data only, no logins.

·