Jon Moshier / Notes / Operational Transparency budding
Note · From the Notebook

Operational Transparency

Ryan Buell's finding that deliberately showing customers the hidden work done for them, and showing employees the people they serve, raises quality, trust, and satisfaction on both sides.

Standard operations doctrine says customer contact is friction: partition the work away, hide the kitchen, minimize interaction, run lean behind a wall. Ryan Buell’s research at Harvard Business School argues the wall is expensive. When customers cannot see the work done for them they undervalue it, and when employees cannot see the customers they serve they find the work less meaningful. Operational transparency is the deliberate design of windows in both directions.

The two-way window

The core move is mutual visibility. Buell frames it in his 2019 HBR article as cutting windows both into the operation (customers watch the work) and out of it (employees watch the customers). The second direction is the part most people miss. It is not a customer-perception hack bolted onto a service. It is a loop: customers who see effort feel gratitude and value the outcome more, employees who see appreciative customers find the work more meaningful and put in more effort, and that effort is visible, which closes the circle.

This builds directly on the Labor Illusion and the Effort Heuristic: visible labor reads as care, and care triggers the Norm of Reciprocity. Transparency generalizes the finding from a website’s loading screen to physical operations, back offices, and public services.

The numbers

The load-bearing study is Buell, Tami Kim, and Chia-Jung Tsay’s Creating Reciprocal Value Through Operational Transparency (Management Science, 2017), two field experiments plus lab studies in food service. At a cafeteria, making the cook and the customer visible to each other raised customer-reported quality by 22.2% and cut throughput time by 19.2%. The lab arm isolated the mechanism: customers who watched the process perceived more employee effort and were more appreciative, and the effect was strongest when both parties could see each other rather than just one.

The government version is more striking because the baseline is so low. Surfacing the Submerged State (Buell, Ethan Porter, Michael Norton) splits into two arms. In a controlled study, people who used a website visualizing Boston’s work on potholes and broken streetlights came away 14% more trusting and 12% more supportive of government. In the field arm, drawing on real data from the city’s service-request app, users who received photos of completed work went on to submit 60% more requests, across 38% more categories, over the following 13 months. The “submerged state” is the political-science observation that most government work is invisible to the people it serves, so they discount it. Transparency surfaces it. Three of these four studies come from Buell’s own research program, so the positive case, while replicated across food service and government, rests substantially on one group’s work.

The transparency paradox

The word “transparency” hides a reversal. Ethan Bernstein’s The Transparency Paradox (Administrative Science Quarterly, 2012) studied the second-largest mobile-phone factory in the world, embedding Harvard undergraduates on the line. Hanging a curtain that hid workers from managers raised production 10 to 15%. Being watched made workers worse, because they spent effort concealing productive deviations and improvised fixes rather than risk being questioned about them. Zones of privacy let the line experiment before having to explain itself.

Same word, opposite result, because the direction and the relationship differ. Buell’s transparency is lateral and mutual: customer and front-line worker see each other’s effort, which builds reciprocity. Bernstein’s paradox is hierarchical surveillance: a manager watches a subordinate, which suppresses the honest experimentation that improvement depends on. This is the same reason Blameless Postmortems work by removing the threat of being blamed. Observation aimed downward through a power gradient changes behavior for the worse; observation across a peer relationship does not. Anyone citing “transparency” should specify who is watching whom.

Where it breaks

Even the beneficial, customer-facing kind has limits. The Labor Illusion’s dating experiment showed the effect inverts when the surfaced outcome is bad: visible effort spent on a poor result reads as incompetence. Separately, more disclosure is not monotonically better. A controlled study of an algorithmic interface (Kizilcec, 2016) found a bell-shaped relationship: too little or too much explanation both lowered trust, with an optimum in the middle. Past that point extra detail causes overload and can surface complexity that erodes the trust it was meant to build. The design question is not whether to show the work but which work, to whom, and how much.

There is a sharper problem underneath. Because visible effort triggers obligation whether or not the labor is real, the same design that surfaces genuine work can stage fake work. The Labor Illusion fires on manufactured delays and performative “working…” screens, which puts operational transparency one disclosure away from manipulation and leaves open whether the effect survives once people know the status feed is a designed reciprocity device rather than a candid readout.

Try it

Instrument a real handoff (an afternoon, any service you run or use). Take a process that currently returns a silent result (an internal request queue, a support ticket, a deploy pipeline) and add a status feed that names the steps and the person or system doing each: “Priya picked this up,” “running 214 checks,” “fixed, here’s the before/after photo.” Measure a proxy for satisfaction or trust before and after (ticket reopen rate, CSAT, thank-you replies). Watch for the reciprocity signature: appreciation and perceived quality rising even when the actual completion time did not drop. Then check the failure edge from the research: on a request that resolves badly, does the added visibility make the reaction worse than silence would have? If Buell holds, good outcomes get a lift and bad ones get a penalty.

Sources

← All notes Read recent essays →