Link prospecting: tools and process for building a target list

Link prospecting is finding and vetting pages for a link or mention before outreach. In AI search, you start with the pages AI answers already cite. That changes who belongs on your target list. Competitor backlinks become a secondary source.

This guide takes an offsite angle. Offsite means content about your brand on sites you don't own. ChatGPT, Gemini, and Perplexity draw on those third-party pages for answers. Prospecting is how you earn mentions and citations there.

AirOps AI search tracking shows which third-party URLs get cited for your prompts. Below you'll find a 6-step process, a qualification scorecard, a tool comparison, and measurement guidance.

What is link prospecting in SEO?

Link prospecting in search engine optimization (SEO) is finding and vetting pages you want a link or mention from. The output is a qualified target list. Outreach starts once the list is ready.

Link building prospecting runs in three stages.

• Research finds candidate pages.
• Qualification filters candidates on evidence and fit.
• Contact finding names the person to pitch.

Four terms come up throughout this guide.

• A citation is an AI answer linking or footnoting a page as a source.
• A mention is an AI answer naming a brand, with or without a link.
• A placement is your brand earning a spot on a third-party page.
• Mention rate is the share of AI answers that name your brand.

Classic link building counted hyperlinks. AI answers also read brand names in plain text.

A named brand on a cited page still shapes the answer. Ask for the link, but accept a strong mention-only placement.

Large language models (LLMs) choose a set of sources for each answer. Those sources are your best prospects. Ali McCarty, VP and GM of Managed Services at AirOps, starts by identifying which URLs AI answers cite for the commercial prompts that matter.

That moves the starting point of your list. You begin with cited URLs. Competitor backlinks fill gaps later.

Each prospect type needs a different ask.

• Listicles and reviews need inclusion or a higher position.
• Resource pages need a relevant asset to link.
• Journalists need data or a timely story.
• Communities need genuine participation from your own people.

Relevance also keeps you within policy. Google's spam policies treat links created primarily to manipulate search rankings as link spam. A list built on topical fit stays clear of that.

The table compares the two approaches.

How to build a link prospecting list in 6 steps

Six steps turn category prompts into one qualified, de-duplicated list. Whoever runs the work acts as the link prospector. The role mixes research and analysis.

Ethan Smith, CEO of Graphite, advises monitoring which domains influence your category and targeting them strategically. Those domains set your offsite priorities.

Step 1. Pick topics and a category prompt set

Start narrow with one or two topics. Pick one topic you win and one you lose. Comparing both shows whether your work moves anything.

Write category prompts for each topic.

• Use category prompts, since branded prompts inflate mention rate.
• Cover early research questions and late comparison questions.
• Keep the same prompt set so trends stay comparable.

Step 2. Pull the domains and URLs AI cites

Run each prompt in ChatGPT, Gemini, and Perplexity. Record every cited URL, along with its domain. Repeat the runs, because cited sources shift between answers.

"I write down every source that it cites. And then this list effectively becomes your target set. And it's not a wish list of publications, it's… the model has already told you where the model is shopping in your category," says Josh Grant, founder of StackedGTM.AI.

The point is evidence over aspiration. Cited URLs show where the model already looks. Rank the domains by how often they appear. Expect your own site and top competitors near the top.

In the clip below, Grant walks through his AI search sprint, including the category-question method.

Step 3. Add classic sources

Classic sources fill gaps in citation data. Majestic's Clique Hunter, on its Pro plan, compares the backlink profiles of competing sites. Footprint searches surface roundups and resource pages.

Google documents search operators such as quotes, site:, the minus sign, and filetype:. Leave no space after the colon. Try footprints like these.

• "best [category] tools" -site:[your domain]
• site:[publisher domain] "[category] alternatives"
• "[topic]" "resources" filetype:pdf

Step 4. De-duplicate against existing partner contacts

Ask your public relations (PR) and affiliate leads which contacts they own. Remove those from your list. McCarty calls this a scalpel rather than a sledgehammer. A narrow comparison page can influence a high-intent AI answer more than a broad placement on a major publication.

Newer brands often need stronger owned content and brand credibility first.

Step 5. Tag each URL by page type and priority

Format drives priority. Tag each URL as a listicle, review, comparison, resource page, news story, or community thread.

Editorial pages such as reviews, comparisons, and listicles often appear among cited sources. Tag them as high priority.

Keep the long tail in view. Pie charts of top cited domains distort the picture. Alex Halliday, co-founder and CEO of AirOps, notes those charts show share "only as a percentage of the top 5… often, you will find a lot of opportunity in the long tail, of domains that you can build relationships with."

Give long-tail URLs a fair priority score.

Step 6. Find contacts and log them in a CRM

Find the author or editor for each URL. Log each contact in a customer relationship management (CRM) tool. Record the URL, page type, priority, and ask.

Manual list building is still common. BuzzStream's State of Digital PR 2026, a survey of 150+ digital PR professionals published February 2026, found 32.4% build media lists manually without a tool. BuzzStream sells list-building software, so weigh the figure with that in mind.

The same survey found outdated or incorrect journalist information is the top problem with list-building tools, at 62.8%. Confirm each contact still works at the publication before you pitch.

Here's the full flow for one topic.

How to qualify link prospects for AI search

Qualify prospects on citation evidence, recency, and topical fit before domain authority. This step separates AI-era link building prospecting from classic lists.

Domain Authority (DA) is Moz's score. Domain Rating (DR) is Ahrefs' score. Both are third-party estimates, not Google metrics.

Backlink counts and DA or DR scores make weak filters for AI prospects. Check whether AI answers cite the page first. Use DA or DR as a tiebreaker only.

Run each prospect through these questions.

• Do AI answers cite the page for your prompts?
• Did the citation appear in your latest prompt runs?
• Does the page topic match your category?
• Can you land in the top three of the list?
• Does the cost fit the citation value?
• Does the site mark paid placements with rel="sponsored"?

Paid placements need correct markup. Google asks publishers to mark paid links with the sponsored rel value. Nofollow remains an acceptable alternative. Expect any paid placement you buy to carry one of them.

Position matters on listicles. Aim for spots one, two, or three, not the eighth or ninth. A spot near the top keeps your brand in the part of the list readers and AI answers see first.

Apply the same test to any link building provider. Ask whether its placements get cited by AI answers today. Request examples of cited placements before you sign.

Reject placements for two reasons: cost above citation value, or poor fit.

Audit your current partners too. Some link partners appear in AI answers, and some never do. Crystal Carter, Head of AI Search & SEO Communications at Wix, suggests checking "which of the… link partners that you have… are regularly cited by LLMs."

Shift effort toward partners that show up in answers. Move the rest to a lower tier.

Score each prospect against this table.

What are the best link prospecting tools?

The best link prospecting tools each do one job well: discovery, data, enrichment, contacts, or outreach. No single link prospecting tool covers all five. Pick by the job you need done.

AirOps covers AI citation-source discovery. Its AEO tracking lists the URLs and domains AI answers cite for the prompts you track.

Prices below are list prices as of October 2026, in US dollars. Confirm on each vendor page before you buy.

Combine two or three link building tools into one stack.

• For AI-first lists, pair AirOps with Hunter or Skrapp.io and a CRM.
• For classic lists, pair Pitchbox or Majestic with URL Profiler.
• For digital PR lists, use BuzzStream for searches and outreach.

Watch compliance with search harvesters. Google's spam policies say automated access to Google Search without express permission violates its policies and Terms of Service. Keep that in mind with tools like ScrapeBox. Skip automated link posting, which Google lists as link spam.

Classic link building tools still earn a spot in the stack. You can also hand off the placement work. AirOps external content work identifies the third-party sources AI already trusts in your category and coordinates the work to build presence there.

How to measure link prospecting results

Measure mention rate, citation rate, and share of voice on the prompts you targeted.

• Citation rate is the share of answers that cite your own pages.
• Share of voice is your mentions compared with competitors' mentions.

Offsite placements mostly move mention rate. On-site content mostly moves citation rate. Track the two separately.

Set each topic's target against a named competitor. Mention rates never sit still, because competitors keep working.

Josh Grant frames mention rate as the number that replaces rank. Pick a query set and commit to it.

Per-placement effects show before topic-level movement. Watch the prompts tied to each new placement first. Judge the topic only after several placements go live.

Association is not cause, so isolate your impact. A split test does that cleanly. In the same AirOps session, Alex Halliday suggests splitting pages or topics into two groups and running activities on one.

Run placements on the test group only. Compare the change in mention rate between the groups after each cycle. Enough placements compound into AI citations across a topic.

Attribution needs two sources. Google Analytics 4 (GA4) only records the visits that click through from an AI answer. Buyers who read an answer and never click stay invisible there. Add a "where did you hear about us?" field to your forms. Read both sources together.

Avoid these common mistakes.

• Filtering by DA alone drops cited long-tail pages.
• Branded prompts inflate mention rate.
• Large, untargeted lists waste outreach on poor-fit pages.
• Judging a topic before placements go live hides real impact.

Our guide to AI visibility metrics covers tracking setup.

What to do next

Start with one topic this week.

1. Write category prompts for one topic you win and one you lose.
2. Run them in ChatGPT, Gemini, and Perplexity, and log every cited URL.
3. Remove contacts your PR and affiliate partners already own.
4. Score the rest on citation, recency, fit, and position.
5. Pitch the top URLs and track mention rate on those prompts.

AirOps for link prospecting from AI citations

AirOps shows the third-party domains and URLs cited for the prompts you track. That citation-source data becomes your first prospect list. See the AirOps platform for setup.

AirOps external content work coordinates presence on those cited pages and measures the change. You then track mention rate on the prompts you targeted. Bring one topic, and we'll map its cited sources with you.

Work with us