Link Gap Analysis Read as Mechanisms, Not Domains

The useful output of a link gap is not the list of domains linking to three competitors and not to you. That list is an export, it takes eleven minutes, and roughly seventy percent of it is unobtainable for reasons the export cannot show: the link arrived with an acquisition, a funding round in 2021, a founder's former employer, a conference sponsored once. What this engagement produces instead is every link classified by the mechanism that created it, and a straight answer on which of those mechanisms you can enter this year. The mechanisms, in the order they usually pay for B2B SaaS and developer tools: Integration and partner directories, where a contract you already signed entitles you to a listing nobody ever claimed. Package and dependency graphs: for a developer tool, the README of every project that installs you, and the curated lists indexing your category. Review platforms, the two or three with real referral traffic behind the profile rather than all of them. Registers and memberships, which in Spanish fintech means the Banco de Espana and CNMV registers, AEFI membership and sandbox cohort lists, links not for sale at any price that competitors have been sitting on for years. Documentation and syllabus citations from universities and bootcamps. Category comparison pages that already list four vendors and could list five. Podcast and conference show notes. Press coverage, the slowest and least controllable of them, and the one every agency proposes first. The deliverable: One sheet, a row per acquirable link, with the mechanism, the specific action that would earn it, the person at your company who has to do that action, and an estimate in calendar weeks. A second, shorter sheet of what you cannot have and why, because knowing that a competitor's advantage came from a 2019 acquisition changes what you decide to fund. And one page of written reading: what the others are doing that you are not, stated without hedging. Sample and method: Intersect exports from Ahrefs, Semrush and Majestic, deduplicated to root domains, filtered to domains with organic traffic of their own, then read by hand. The reading is the entire value, which is why the engagement takes four to six competitors rather than twenty. Twenty competitors produce a longer sheet and a worse decision. What this is not: Not execution. The engagement ends with the document. If you want the programme run afterwards that is separate work, and I will tell you honestly whether your own team can do it, which for most of the mechanisms above they can. Not a keyword gap, not a content audit, not a technical crawl. Not an assessment of your own profile for spam, which is the audit engagement instead. Who should not buy this: Companies whose competitors are the same age and size and got their links the same way you got yours; you would pay for a document saying there is nothing to copy. Consumer brands, where the mechanisms are different and I am not the right reader. And anyone who cannot name their real competitors, because if the list is the four names in an analyst quadrant rather than the four names that appear when your buyer searches, the analysis inherits the mistake. I will push back on the list before starting rather than after. What I need to start: Four to six competitor domains, read access to Search Console, and one call with whoever owns partnerships, because half the acquirable links in a typical sheet turn out to sit inside agreements your company has already signed.