Most data teams cannot explain how their work changes the business. This isn’t a communication problem—it’s a missing model of impact. And it’s precisely the chaotic middle that Cutler describes: organisations have moved beyond pure delivery thinking but haven’t yet developed shared value models to anchor their goals and investments.
Ask a software product team how their work creates value, and they can typically trace a path from feature to user behaviour to business outcome. The connections may be imperfect, but they’re articulable. Ask the same question of a data team, and the response often dissolves into abstractions about “enabling better decisions” without clear mechanisms connecting data work to business change.
The Invisibility Problem
Data value suffers from a visibility challenge other product work doesn’t share. A new application feature is tangible—users interact with it, behaviour changes observably, metrics move in response. Data work operates more like dark matter: invisible itself, yet its presence shapes everything else.
When functioning well, decisions get made faster. Opportunities get recognised earlier. Risks get priced more accurately. But because the enabling infrastructure remains invisible, attribution becomes nearly impossible. The connection between data investment and business outcome disappears into a gap that neither technical teams nor business stakeholders can bridge.
The Missing Translation Layer
What data organisations lack is not goals but a shared language connecting technical outputs to business outcomes.
Technical teams describe work in terms of artefacts: pipelines built, models deployed, dashboards created. These descriptions represent effort but say nothing about value. Business stakeholders describe needs in terms of questions to answer or decisions to make, but lack vocabulary connecting these needs to data capabilities.
Without this translation layer, adoption depends on individual initiative rather than systematic alignment. Some stakeholders extract value. Most do not.
The Measurement Displacement
Into this vacuum flows measurement—but of the wrong things. Unable to articulate impact, organisations default to counting what’s easy: data products shipped, dashboard views, query volumes, adoption rates.
Proxies aren’t inherently harmful. They become dangerous when they harden from hypotheses about impact into goals themselves. Dashboard views measure curiosity—someone looked. They reveal nothing about whether looking led to different action or better judgment.
Some argue that “better decisions” cannot be measured. This concedes too much.
Data doesn’t make decisions, but it prices the risk of decisions.
We can track decision latency, rework frequency, confidence levels, outcome variance. The measurement will be noisy and indirect, but even imperfect signals anchor teams in outcomes rather than proxies. Without such signals, teams optimise for what gets counted—and motion masquerades as progress.
Toward Shared Value Models
Escaping the trap requires developing explicit models of how data creates value in specific contexts. These models need not be complex, but they must be shared—understood across the boundaries between technical and business functions.
The work begins with shifting the fundamental question:
From “what did we produce?” to “what change did we enable?”
This reorientation forces attention toward mechanisms connecting data work to business outcomes. It surfaces assumptions and reveals where those assumptions may be unfounded.
Most importantly, shared value models create foundation for genuine dialogue about priorities and investments. When organisations share understanding of how data creates value, they can make informed decisions about where to focus and how to evaluate success. Without such models, every conversation about data value becomes translation without a common language.
Next in this series: Shared value models don't emerge from workshops or frameworks alone. They require something deeper—a human operating system that distributes trust and authority to where knowledge lives. In Where Data Strategy Actually Lives, we examine why the binding constraint on data value is rarely technical, and what genuine progress demands of leadership and culture.