[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"glossary-integration-attach-rate::en":3,"gloss-cluster-integration-attach-rate::en":26,"gloss-next-integration-attach-rate::en":9},{"slug":4,"category":5,"name":6,"definition":7,"meta_desc":8,"faq":9,"schema_markup":9,"related":10},"integration-attach-rate","analytics","Integration Attach Rate","Integration attach rate is the share of accounts that have connected at least one integration, normally measured across active accounts over a window. It is tracked because a connected account has more of a customer's data and more of their working process inside the product than an unconnected one, which makes it one of the few product metrics that is genuinely about the surface it names rather than a proxy for something else.\n\nDefine connected before you compare anything. It can mean an authorisation was completed at some point, that a connection currently exists, or that a connection moved data recently — and those three numbers diverge sharply in any real account base, because connections get authorised and then abandoned, and because a revoked or broken connection usually still looks connected to a naive query that counts rows rather than checking state. A rate built on ever-authorised can only go up, which makes it a poor health measure and an attractive one to report.\n\nThe interpretation trap is causal, and it is worth naming because this metric invites it. Accounts with integrations retaining better than accounts without does not establish that connecting one causes retention; the customers who connect are disproportionately the ones who had already decided to build a process around the product, so the intent came first and the connection is evidence of it. The number is worth watching. A lift claimed from it needs a comparison that separates the act from the intent.\n\nIt is most useful segmented rather than aggregate. Which integrations are attached, by which kind of customer, how long a connection survives after it is made, and how many accounts have one that has silently stopped working — a company-wide percentage rising while a popular connector quietly breaks is a number that reports health while describing the opposite.","Integration attach rate is the share of accounts with at least one connection — why the definition of connected decides the number, and why it is not causal.",null,[11,14,17,20,23],{"slug":12,"name":13},"expansion-revenue","Expansion Revenue",{"slug":15,"name":16},"net-revenue-retention","Net Revenue Retention (NRR)",{"slug":18,"name":19},"onboarding","Onboarding",{"slug":21,"name":22},"product-analytics","Product Analytics",{"slug":24,"name":25},"time-to-value","Time-to-Value (TTV)",[27,31,34,37,40,43,47,50,53,56,59,62],{"slug":28,"category":5,"name":29,"updated_at":30},"ab-testing","A\u002FB Testing","2026-09-08T21:32:36+00:00",{"slug":32,"category":5,"name":33,"updated_at":30},"active-user","Active User (DAU, WAU, MAU)",{"slug":35,"category":5,"name":36,"updated_at":30},"autocapture","Autocapture",{"slug":38,"category":5,"name":39,"updated_at":30},"cost-per-resolution","Cost per Resolution",{"slug":41,"category":5,"name":42,"updated_at":30},"customer-data-platform","Customer Data Platform (CDP)",{"slug":44,"category":5,"name":45,"updated_at":46},"data-sampling","Data Sampling","2026-09-08T21:32:37+00:00",{"slug":48,"category":5,"name":49,"updated_at":30},"deflection-rate","Deflection Rate",{"slug":51,"category":5,"name":52,"updated_at":46},"double-counting","Double Counting",{"slug":54,"category":5,"name":55,"updated_at":46},"feedback-signal","Feedback Signal",{"slug":57,"category":5,"name":58,"updated_at":46},"goodharts-law","Goodhart's Law",{"slug":60,"category":5,"name":61,"updated_at":30},"guardrail-metric","Guardrail Metric",{"slug":63,"category":5,"name":64,"updated_at":30},"identity-resolution","Identity Resolution"]