Method
In-app search analytics, as we use the phrase
Not SEO. Not a relevance-engineering bootcamp. A reading practice for the strings people type once they are already inside the product.
In-app search analytics is the disciplined reading of queries, result counts, ranking snapshots, and next actions inside a logged-in or installed product. The unit of attention is the string, dated. Dashboards are allowed only after the string has been quoted.
What belongs in the practice
Typed queries, including typos and gift language. Empty pages, bucketed. Clicks that never happen. Filter chips that fire the same event as the search box. Ranking of a frozen query list around a catalogue feed. Sampling rules that legal can read. Facet leakage: size and colour typed because the rail failed.
What we leave out on purpose
Google Search Console. Keyword tools for acquisition. BM25 lore as a personality. Vector demo theatres. Sentiment on queries. A single findability score. If those are your job, hire differently. Our courses assume you already have an engine, however imperfect.
A weekly shape
Monday: dump a fortnight, redact, sample. Wednesday: assign zero-result buckets to twenty strings. Friday: one snapshot or one merchandising sentence, not both if you are tired. That rhythm is what Query Signal Lab rehearses until it is boring, which is the goal.
If this matches the argument you are already having in your UK product team, write to us with a description of the export. If you need someone to rebuild ranking, we will say no quickly so you can spend the week elsewhere.
Describe the export