When you have more outreach candidates than time, you need a reasoned way to choose which sites deserve closer research. Throw out the unsuitable prospects before you rank anything. Then order what survives by fit, direction and effort.
Eliminate before you rank
Ranking a list you never filtered wastes the ranking. Make these cuts in order, because each one costs more than the one above it.
Cut on fit. Do you have something specific to say about this page's subject? Being in the same industry is not enough. A resource list on an adjacent topic will read your pitch as a stretch, because it is one.
Cut on mechanism. Does the page accept anything at all? A submissions page, a suggest-a-resource link, contributor guidelines. Without one you are asking somebody to edit a published page as a favor, and that needs a relationship you probably do not have.
Cut on requirements. Membership, fees, credentials, submission windows, exclusivity. Read what the page asks for and answer honestly whether you can meet it.
Most prospects die at the second or third cut.
Then rank what is left
1. Audience precision
The most important signal and the hardest to measure. A modest publication read by two thousand of exactly your buyers beats a large general site where your piece disappears.
Rank on how closely the readership matches the person you sell to. Not on how impressive the site looks.
2. Direction over time
A current authority figure hides the trend completely. Month-by-month link measurements show whether a site is rising, holding or sinking, and three sites with identical figures today can be three different prospects.
Two years of steady decline rarely justifies the effort. Approach a consistently rising site now, ahead of the point where everyone else notices it.
Check the January 2026 measurement change before reading any decline. From that month the bare host and its www variant are counted separately, which makes unchanged www-served sites look like they fell. Current and historical link measurements has the detail.
3. Existing relationship or mention
Any row where the site already mentions you goes to the top. The publisher already chose to name you, so you are sending a small request and not a pitch.
Next come rows where the site links to several of your direct competitors. The precedent exists and you can usually work out what created it.
4. Effort required
A directory listing takes minutes. A guest contribution is a commissioned piece of work. An award submission has a deadline and a form. A podcast is an hour plus preparation.
Balance the list. Sort purely by value and you end up with nothing but high-effort placements, and the programme stalls in week two.
A worked ordering
Say qualification leaves you twelve rows. A reasonable order:
- Two pages that already mention you without linking. Minutes each, high success rate.
- One page with a broken link to your old domain. A correction, and a favor to them.
- Three directories or resource lists with submission forms and requirements you meet.
- Two publications linking to three of your four named competitors editorially.
- Three contributor opportunities where you have a real piece to write.
- One award submission, with the deadline written down for later.
That front-loads the quick wins. Outreach programmes die from lack of visible progress far more often than from lack of prospects.
What to do with the rest
Delete it. A long list of prospects you already judged weak just gets re-read every time you open the file.
Run a new search on a different opportunity type instead. Better rows than grinding down a list you already dismissed.
What no signal can tell you
Whether the placement gets accepted, and whether the link turns out to be worth anything. Both are judgments. Link measurements describe structure, and they say nothing about editorial standards or audience quality.
Use the numbers to order a list you already believe in. Do not use them to decide whether to believe in it.
Use qualification before scheduling outreach. A smaller list with clear audience fit and a practical next step gives you work you can act on.
See how to evaluate a prospect, inspecting a site's history, or the Prospect Index.