Most backlinks strategies documents are lists of tactics with no sequencing, no target, and no way to tell whether the quarter went well. A tactic list is not a plan. A plan says which pages you are lifting, which domains you are approaching, how many links the job needs, who does the work, and what number you will look at in ninety days.
This article is the structure for that document. The tactics are the easy part; the decisions around them are what separate a link building plan that gets funded twice from one that gets quietly abandoned in week six.
Step one: pick the pages, not the site
Sitewide link building is a budget with no destination. Pick three to five specific pages that matter commercially and that are close enough to page one to move. A page sitting at position 34 with thin content is not a link problem. A page at position 12 with genuinely good content and fewer referring domains than everyone above it is exactly the right target.
Check intent alignment before you commit. If the SERP is full of comparison tables and your page is a brand story, links will not close that gap. Run a technical scan too — it is common to discover that some of your best-linked URLs are canonicalizing authority away or returning the wrong status code. Those fixes are free compared to outreach.
Step two: size the job against the weakest competitor
The single most useful number in a link building proposal is how many referring domains you need, benchmarked against a realistic target. Not the market leader. The weakest site currently holding a page-one position for your term.
SEO Rocket estimates this against two milestones: the weakest page-one competitor and the page-one median. The weakest competitor is your quarter-one goal — reachable, provable, and enough to justify the next quarter’s budget. The median is the longer campaign. Presenting both up front is what stops a perfectly healthy three-month project being declared a failure at week five because someone was silently benchmarking against the category leader.
Those estimates carry niche-based cost bands, which are directional planning figures for budgeting time and headcount. They are not quotes, and they are not a price list — buying links violates Google’s link spam policies and no cost band changes that.
Step three: build the prospect list from the gap
Prospecting is where campaigns quietly die. Scraped databases of ten thousand generic contacts produce nothing but a damaged sender reputation. The productive list is much smaller and much better qualified: sites that already link to your competitors and not to you.
SEO Rocket’s backlink gap analysis compares your profile against up to five ranking rivals and returns exactly that, ranked by how many competitors each domain links to. Domains linking to four of five rivals go at the top — they have proven repeatedly that they cite businesses in your category. It filters obvious spam and flags toxic domains, so what you review is a workable shortlist rather than several thousand raw referring domains.
Run the anchor-text gap alongside it. Seeing which descriptive phrases the SERP winners are consistently associated with — and which you have over-leaned on — tells you what to publish, not just who to email.
Step four: sequence tactics by effort and lead time
Different tactics pay out on different timelines. Sequence them so something lands in month one while the slower work matures.
- Weeks 1–3 (fast, near-certain): unlinked brand mentions, supplier and partner pages, industry associations, integration directories, event and sponsor listings. Almost every site has a dozen of these unclaimed.
- Weeks 2–8 (medium): broken link replacement against dead resources your niche still references, and outreach to the top tier of the gap list.
- Weeks 1–12 (slow, highest value): one original data asset — a customer survey, anonymized platform data, or an analysis of a public dataset nobody in your niche has bothered with. This is what earns citations for years.
- Ongoing: expert commentary. Be reachable, be fast, have an actual opinion.
Deliberately absent from the sequence: directories nobody uses, comment links, forum signatures, private blog networks, and reciprocal exchange schemes. They either do nothing or create a footprint you will pay to clean up later.
Step five: control velocity
Growth rate is a strategy decision, not an accident. A domain that is four months old gaining 400 referring domains is broadcasting a pattern. Growth that tracks your publishing output and your visibility looks normal because it is normal.
A reasonable shape for most small and mid-sized sites: five to fifteen genuinely earned referring domains per month, weighted toward relevance rather than authority scores. Slower than the pitch decks promise, and it compounds instead of collapsing.
Step five and a half: decide who does the work
Plans fail on staffing more often than on strategy. Earned link building is research, writing, and follow-up — roughly two to four hours per genuinely good placement once you include qualifying the prospect, finding the right contact, and writing something worth reading. Multiply that by your monthly target and you have an honest headcount figure.
If the number does not fit the team you have, cut the target rather than the standard. Fifteen mediocre placements bought in bulk will not do the job of five earned ones, and they create cleanup work later. This is also the calculation that exposes suspiciously cheap agency pricing: the arithmetic tells you what a very low per-link price can possibly be buying.
Assign a single owner. Link building distributed across three people who each do it when they have time reliably produces nothing, because outreach depends entirely on follow-up discipline.
Step six: report on the right numbers
Reporting is where good work loses credibility. Three rules keep it honest.
- Referring domains, not total links. One site linking forty times is one endorsement.
- Trends over 28-day windows, not daily positions. Rankings jitter two or three places as a matter of course. A day-after check measures noise.
- Search Console and GA4 as ground truth. Third-party volume, difficulty and position figures are modeled estimates from periodic crawls — excellent for comparing competitors on a level field, imprecise about your own site.
Expect a lag of eight to twelve weeks between a link landing and any visible movement. Linking pages need recrawling and signals need to settle. Watch impressions before positions; impressions usually move first.
What to write in the plan document
Keep it to one page. Target pages and current positions. The weakest-competitor link target and the median target, with the gap between where you are and each. The prospect count by tier from the gap analysis. The tactic sequence with owners and weekly volumes. The reporting cadence and the specific metrics. And a stated assumption about lag so nobody panics in week four.
That document survives contact with a skeptical stakeholder because every number in it is defensible and every target is grounded in what a real competitor is actually doing.
A final note on scope. Software is genuinely good at the mechanical parts of this: computing the gap, ranking prospects by competitor overlap, filtering spam, flagging toxic domains, running the anchor comparison, and sizing the job in links and directional cost. SEO Rocket does all of that. It does not send outreach emails, manage campaigns, or host a link marketplace. The relationships are human work — and that is precisely why links earned that way keep counting when the shortcuts stop.
