Reviews
Part of Independent references for link building: assets, outreach, and publisher choice
Stronger link building: fixing the destination before outreach
How to improve link building in 2027 through audience purpose, original assets, small source groups, useful outreach, clear relationships, and full outcome learning.
What to take away
- Fix the asset before outreacha document, dataset, or tool a named audience can cite on its own merits.
- Set relationship, disclosure, privacy, approval, suppression, and technical-link rules before outreach or paid distribution begins.
- Judge results by qualified conversations, accurate coverage, referrals, cost, and neutral outcomes, not by live URL counts.
Learning how to improve link building starts with the asset rather than the pitch. A page that answers a real question earns references. A thin page with a good contact list earns nothing.
Improve the destination first
Start with one destination, one source audience, and one reason that audience would cite it. Publish original data, a clear method, expert guidance, a public tool, a template, or documentation that solves a source's problem.
Add author or reviewer context, dates, definitions, limitations, and licensing. Name one owner for updates. Fix broken pages and unclear claims before any outreach.
Set rules and outcomes
Define qualified referral visits, assisted actions, reputation, partnership, citation, or search outcomes. Prohibit unqualified paid ranking links, excessive exchanges, automated creation, private networks, fabricated identities, and required anchors. Approve disclosure, privacy, and opt-out rules before contact.
Build smaller source groups
Segment journalists, educators, associations, researchers, partners, customers, suppliers, local institutions, and resource curators by actual use. Review each person's beat or page, market, editorial standards, and preferred contact. Exclude irrelevant, copied, abandoned, or placement-selling properties.
Worked example: a permit-delay dataset
Asset: a public dataset of building-permit processing times for the 50 largest US cities, built from each city's open-data portal, with query text, pull dates, method notes, and a downloadable CSV.
Source group: 40 city-hall reporters at metro dailies and municipal finance analysts at research nonprofits that covered permit delays in the past year.
A sample message:
- Subject: 2024 permit times for your city
- Hi [name],
- We rebuilt the permit-delay dataset from your city's open-data portal and updated the 2024 figures.
- Pull dates, the query, and the limits are on the project page.
- Your recent story on slow permitting used the 2023 numbers, so the change may interest you.
- Use any figure you like. I can walk through the method if that helps.
- [Your name]
Benchmarks to plan around
Reply rates for cold, well-matched outreach: a typical range is 2 to 10 percent. Generic mass pitches sit closer to zero.
Time per asset: a data or tool asset usually takes two to six weeks of research, review, and design before the first pitch.
Cost per earned link: driven by research hours, design, legal review, and outreach volume. Teams commonly budget a few thousand US dollars for one substantial asset, and a few hundred for a template. Treat any figure as an estimate for your own scope.
Make the message useful
Explain the finding, intended audience, method, limitations, and available evidence without demanding coverage or a link. In the US, CAN-SPAM sets the baseline for commercial email: accurate headers, a clear opt-out, and prompt suppression.
For UK recipients, the ICO's PECR electronic marketing guidance sets subscriber consent and opt-out rules. Use honest identity and keep suppression records.
Separate earned and paid activity
Report earned coverage, partnerships, user submissions, affiliate activity, and sponsorship separately. The FTC's native advertising guide for businesses says deception is assessed from the overall context and discusses when disclosures are needed so people recognize advertising.
The FTC's Endorsement Guides apply the same standard to reviews, testimonials, and paid placements. Paid presentation and technical link qualification solve different problems; neither should be hidden inside an earned-outreach total.
Learn from full outcomes
Track qualified conversations, accurate coverage, referrals, actions, corrections, lost references, and relationship value. Preserve rejections and neutral results. Audit published samples for disclosure, link state, destinations, and reader utility.
Price research, design, review, tools, outreach, corrections, and maintenance instead of celebrating cost per live URL. Check link state in Google Search Console, Ahrefs, or Semrush before you credit a placement.
Decision table
| Stage | Working record | Link-building gate |
|---|---|---|
| Destination | Page, source need, citeable evidence, update owner | Can an editor cite this page as it stands? |
| Source group | Beat, page, market, contact preference, exclusions | Does this person already cover this topic? |
| Outreach | Identity, relevance, disclosure, opt-out, suppression | Does the message pass CAN-SPAM and the FTC test? |
| Placement | Anchor text, rel attribute, destination URL, disclosure | Is the link labeled and pointed where we claimed? |
| Outcome | Replies, coverage, referrals, corrections, cost, neutral results | Keep, fix, or stop? |
Apply US rules to the outreach
US outreach runs on the FTC Act and the CAN-SPAM Act, not on procurement policy. Keep identity honest, subject lines accurate, and opt-outs working. Label paid placements so readers know who paid.
For vendor and cost evidence, the GOV.UK technology selection guidance is one public example, not a US rule.
Teams weighing paid placements should settle those common link building questions before outreach begins.
The CISA software acquisition fact sheet covers supply-chain review for government software buys.
Common questions
What is the first link-building improvement to make?
Fix the destination. Publish the evidence, method, and limits an independent editor needs before you contact anyone.
Should earned and paid links share one KPI?
No. Classify earned, partner, user-generated, affiliate, sponsored, and other paid activity separately, with their relationships, disclosures, treatment, outcomes, and costs.
How should outreach quality be evaluated?
Review audience fit, honest identity, useful evidence, applicable rules, limited pressure, opt-out handling, qualified replies, corrections, complaints, and relationship value.
