The Future of Operational Analytics

Summary

Stancil argues that internal analytics fails because it’s detached from the work it’s meant to inform, unlike consumer data products (Yelp, Google Maps) that embed data seamlessly into decision contexts. The gap isn’t technical literacy — it’s product design. Core mental models:

Data Replaces Intuition When Embedded. Yelp didn’t train users to be “data-driven diners.” It embedded ratings, reviews, and photos into the restaurant selection flow. Nobody was “onboarded to Yelp’s schemas.” We just used it. Internal analytics demands the opposite: leave your workflow, open a different tool, learn a query language, interpret a chart.
Analytics Is Not Primarily Technical. The SQL-vs-everything-else debate misses the point. Technical skills don’t separate average analysts from great ones — critical thinking and communication do. Analytics engineering can either be the bridge (lowering the technical bar so analysts can focus on thinking) or the bleed (making “technical enough” the graduation standard for analysts).
Dashboards Are Data Looking for a Problem. The “what do you want on the dashboard?” question punts the hard decisions about what matters to the end user. Charts, filters, cohorts accumulate. This isn’t serving customers — it’s serving ourselves.
Three Principles for Operational Analytics. (a) Solve specific problems, not general “data access.” (b) Be ruthlessly disciplined about scope — say no to feature creep. (c) Guide decisions, don’t automate them. People want options narrowed and ranked, not to be replaced. Human in the loop.
The Illusion of Quantitative Rigor. A dashboard version of Yelp with scatter plots and dropdowns would be worse at solving the restaurant problem. Dashboards create an illusion of rigor while making data hard to find, easy to misinterpret, and intimidating to use.

Relevance

06-reference/2026-04-03-reforge-why-analytics-efforts-fail — Stancil’s diagnosis maps directly to Reforge’s framework. Analytics efforts fail when they’re project-based instead of product-based.
06-reference/2026-04-03-data-maturity-processes-tools — The maturity journey should end with embedded operational analytics, not better dashboards.
06-reference/2026-04-03-headless-bi — Headless BI is one architectural answer to Stancil’s call. Decouple the semantic layer so data can surface inside operational tools.

Open Questions

What does “Yelp for internal business decisions” actually look like in practice? Is it Slack alerts? Contextual sidebars in CRMs?
If dashboards are dead, what replaces ad-hoc exploration for the analyst who doesn’t yet know what question to ask?