Relevance engineering and relevancy engineering sound identical and work on different layers. Relevance engineering, associated with Mike King and iPullRank, tunes content so search engines and AI systems rank it as relevant. Relevancy engineering, Petrichor's term, builds whether the market can place a company and repeat it. One optimizes the algorithm. The other optimizes the market.
- One letter apart, one layer apart. The confusion is real and worth clearing.
- Relevance engineering (Mike King, iPullRank) is about machine relevance: getting retrieved and ranked.
- Relevancy engineering (Petrichor) is about market relevance: getting placed and repeated by buyers.
- Ranking for a position the market cannot understand is expensive noise.
- Order matters. Build the meaning first, then make it retrievable.
Where the confusion starts
A founder hears relevancy engineering, searches relevance engineering, and lands on a body of excellent work about embeddings, query-document matching, and how large language models decide what to retrieve. Reasonable mistake. The terms are one letter apart.
It is the wrong body of work for the question the founder was actually asking. The founder wanted to know why the market cannot place their company. The search returned how to get a page ranked. Related, not the same.
The two definitions
Relevance engineering, a term associated with Mike King and iPullRank, is the practice of tuning content and entities so search engines and AI systems judge them relevant to a query. It lives at the level of retrieval: embeddings, structured signals, query-document matching. The goal is machine relevance.
Relevancy engineering, Petrichor's term, is the structural construction of market relevance across three legs: authentic core, category ownership, and distribution mastery. It lives at the level of meaning: whether a buyer can place the company in a category they already understand and repeat that placement. The goal is market relevance.
Same word, different layer
| Term | Optimizes for | Unit of work |
|---|---|---|
| Relevance engineering (Mike King, iPullRank) |
Machine relevance: getting retrieved and ranked by search and AI | The document and the entity, against a query |
| Relevancy engineering (Petrichor) |
Market relevance: getting placed and repeated by buyers | The company's position in its category |
Three ways they differ
The judge is different
Relevance engineering answers to an algorithm. Relevancy engineering answers to a human buyer. You can satisfy the first and fail the second: a page that ranks for a term nobody associates with you sends traffic that bounces.
The failure mode is different
Relevance engineering fails when you are not retrieved. Relevancy engineering fails when you are seen and still not chosen, because the market cannot say what you are. One is an invisibility problem. The other is a placement problem.
The order is different
Meaning comes before ranking. If the market cannot place you, ranking for the right query just delivers strangers to a page that does not land. Engineer the position first. Then make the position retrievable.
Which one your startup needs
Usually both, in that order. Relevancy engineering first, because a retrievable position nobody understands is wasted retrieval. Relevance engineering second, so the position you built is findable by search and AI. Fix the meaning, then fix the ranking.
This is the discipline behind Petrichor's Relevancy Engineering methodology. If you arrived here from the adjacent term, the brand relevance comparison is the next stop, and reputation engineering covers what accrues when the structure holds.
Frequently asked
Are relevance engineering and relevancy engineering the same thing?
No. They sound almost identical and operate on different layers. Relevance engineering is a search and AI discipline about how algorithms judge a document's relevance to a query. Relevancy engineering is a positioning discipline about whether the market can place a company and repeat it. One is about ranking. The other is about meaning.
What is relevance engineering?
Relevance engineering is a term associated with Mike King and iPullRank for the practice of optimizing content and entities so that search engines and AI systems understand and rank them as relevant. It works at the level of embeddings, query-document matching, and technical signals. Its goal is machine relevance: getting retrieved and ranked.
What is relevancy engineering?
Relevancy engineering is Petrichor's term for the structural construction of market relevance across three legs: authentic core, category ownership, and distribution mastery. Its goal is market relevance: whether buyers can place the company in a category they understand and repeat that placement without the company in the room.
Which one does my startup need?
Usually both, in order. Relevancy engineering comes first: if the market cannot place you, ranking for the right query sends traffic to a page that does not convert. Relevance engineering then makes sure the position you built is retrievable by search and AI. Fix meaning, then fix ranking. Ranking for a position nobody understands is expensive noise.