Named case studies age badly. The company that ran a brilliant campaign in 2021 was operating in a different market, and copying the specifics usually produces a worse version of something that no longer applies.
Patterns age better. What follows are six that we have watched work across multiple engagements and categories, with what each actually costs and where each fails.
Pattern 1 — The annual industry number
The shape: a company measures one thing about its industry every year and publishes it with a methodology. By the third edition, journalists cite it without being asked.
Why it compounds: each edition links to the previous one, the archive ranks for the underlying query, and being the only available source is a position competitors cannot easily contest without three years of their own data.
Effort: high in year one, moderate thereafter. Typical output: the strongest single line item in most programmes on a three-year view.
Fails when: it is not repeated. A one-off benchmark ages out in eighteen months and the position passes to whoever commits.
Pattern 2 — The free tool that outranks the product
The shape: a company builds a small standalone utility solving one irritating job in its category, ungated. It earns resource-page links steadily and converts far better than any blog post.
Why it works: linking is how people recommend tools. No pitch is required — the recommendation mechanism is built into the format.
Effort: moderate build, near-zero maintenance if kept simple. Typical output: slow accrual that continues indefinitely.
Fails when: it is gated. Requiring an email before showing a result removes most of the citation value, because nobody links to a form.
Pattern 3 — Systematically claiming the partner ecosystem
The shape: someone spends two weeks emailing every integration partner, association, conference and customer with a case-study page, asking to be listed or correctly listed.
Why it works: these links were always available. The counterparty is contractually or commercially inclined to say yes. Nobody asks because it is unglamorous.
Effort: two weeks of email. Typical output: twenty to fifty referring domains at a cost per link under $50 in staff time.
Fails when: never, really. This is the first thing we tell new clients to do, and it requires no agency.
Pattern 4 — Correcting the round-ups
The shape: audit every "best tools for X" article in the category. Roughly half list you inaccurately — outdated pricing, missing features, a description of the 2022 product — and a third do not list you at all. Contact all of them.
Why it works: publishers want accurate articles, and a polite correction email with specifics is easy to act on. Inclusion requests convert less often but still convert.
Effort: moderate and ongoing. Typical output: referral demo requests within weeks, independently of any ranking effect.
Fails when: the product genuinely loses the comparison. Being accurately listed beside four better tools accelerates a decision that goes against you.
The pattern most companies underestimate
Patterns 3 and 4 together typically produce more referring domains in the first quarter than a $5,000-a-month outreach programme, at a fraction of the cost.
They are unglamorous, they involve no creativity, and they consist almost entirely of email. That is precisely why they remain available — the work is available to everyone and boring enough that most teams skip it.
Pattern 5 — The specialist who becomes a source
The shape: one internal expert responds to journalist queries consistently for a year. At first nothing happens. By month six, writers start approaching directly. By month twelve, that person is quoted routinely across the trade press.
Why it works: journalists build a mental list of reliable, fast, quotable sources, and that list is short. Getting on it takes persistence rather than budget.
Effort: low cost, high consistency requirement — the binding constraint is four-hour response capability, not money.
Fails when: responsibility is delegated to someone junior, or every quote needs three days of legal review. Both remove the reason it works.
Pattern 6 — Publishing the thing your industry argues about
The shape: identify a claim your industry repeats without evidence — a rule of thumb everyone cites and nobody sourced — then measure it properly and publish the result.
Why it works: writers have been asserting the number for years and would rather cite a source. You have supplied one. The links arrive from people correcting their own old articles.
Effort: moderate to high, depending on data access. Typical output: unusually durable, because the claim keeps being made.
Fails when: your finding confirms the folk wisdom exactly. Interesting findings are ones that revise the number.
Three patterns that have stopped working
| Pattern | Why it stopped |
|---|---|
| Scaled guest posting on general-interest sites | The quality bar rose sharply; these placements are now a liability rather than a neutral cost |
| Infographic outreach at volume | Publishers stopped embedding them; the format's citation mechanism collapsed |
| Generic business directory submission | Links from pages with no readers and no topical relationship to anything |
How to choose between them
Not by preference. By what your situation permits.
| If you have… | Run pattern… |
|---|---|
| Integrations, partners, customers | 3 — immediately, before anything else |
| A competitive product in a mapped category | 4 — fastest measurable output |
| Usable product telemetry | 1 or 6 — the compounding options |
| A free-standing useful feature | 2 — build it out and ungate it |
| A genuine internal specialist with time | 5 — cheapest, slowest, most durable |
| Budget but none of the above | Editorial acquisition, while you build one of the above |
The most common mistake is choosing the pattern that sounds most interesting rather than the one your assets support. A company with forty integration partners and no data should run pattern 3, not commission a survey.
What all six have in common
Three things, and they are the actual lesson.
None of them starts by buying a link. Every one creates a reason for someone to link, then makes it easy.
All of them are slower than buying. The first quarter of any of these produces less than the first quarter of volume acquisition. The difference shows up at month twelve, in survival rate and in whether the domains were ones that mattered.
Most of them are available to you now. Patterns 3, 4 and 5 need no budget and no agency — they need someone to own them for a quarter.