nobody goes back

Problem statement

Post-occupancy evaluation (POE) — going back to a finished building to measure whether it performs as designed and how occupants actually use it — is the only feedback loop through which architectural practice could learn systematically from its own projects. It is widely endorsed by professional institutes and required in the RIBA Plan of Work's feedback stage, and yet almost no one does it: Hay et al. report that only 3 percent of British practices regularly undertake POE on housing projects, only 9 percent of chartered practices offer it to clients, and none reported revenue from it. In the United States, the AIA's 2030 Commitment reporting shows the same shape — through 2021 only 329 projects had ever reported post-occupancy (measured) performance data into the DDx system, against 20,652 projects submitted in 2021 alone. Practitioners are designing the next building without knowing whether the last one worked, and building science, occupant health and energy research lack the field data that only practice can generate.

Why this matters

The gap between predicted and actual building performance — in energy, comfort, air quality and fit to use — is well documented, and it persists precisely because there is no routine feedback from occupied buildings back to designers; as one AIA-cited practitioner puts it, "It's not about sensors; it's not about the actual methods of data collection" but about "dedicating resources to share knowledge from our projects." Buildings account for a large share of energy use and of where people spend their lives, so a profession that cannot close its own learning loop propagates the same errors across thousands of projects, and the evidence base for design decisions that affect health, learning and productivity stays thin. The Hay et al. interviewees frame this as an industry-structure failure rather than individual negligence: "the primary time and budget drivers override everything else."

What’s been tried and why it hasn’t worked

POE methods are not the problem — decades of toolkits (occupant surveys, walkthroughs, energy and IEQ monitoring, the UK Probe and BPE studies) exist, and RIBA added a Feedback stage to its Plan of Work. What defeats them in practice is a stack of structural disincentives Hay et al. document from interviews with architects who do attempt POE: fees rarely cover it — the AIA's own reporting says POE is "seen as an extra service requiring extra fees, and owners aren't keen to commit to it if the topic comes up too late in the process," and Hay et al.'s interviewees describe it as a "luxury rather than … [a] necessity" that gets "value engineered out"; standard contracts disincentivize returning to a project after handover; architects are "rarely involved at all stages of a project," have limited contact with users, and often lack control over final quality, so a POE that surfaces problems risks drawing them into liability for outcomes they did not control — findings can be "poisonous both commercially and professionally," and delivery teams "feel it is a blame game." Market structure removes the demand signal (housing shortage means "there is no driver to build better homes"; commercial buyers value location over performance), and the retreat from regulations such as the Code for Sustainable Homes removed the policy push. Even where practices do evaluate, uncertainty over what counts as POE means much of the work is not labelled or shared, so no rigorous evidence base accumulates. Interviewees called for stronger institutional leadership and for mandatory POE on publicly procured buildings — neither of which has materialized at scale.

What would unlock progress

The unlock is a POE model that removes the three practitioner-level blockers at once: cost (near-zero-marginal-cost data capture from building systems and lightweight occupant surveys), liability (a no-fault, anonymized reporting frame in which findings feed a shared evidence base rather than a claim), and fee (a standard scope and price for a "feedback stage" that clients can procure). Adjacent fields have built exactly this: aviation's confidential, no-blame incident-reporting systems turned individual failures into industry learning; medicine's registries pool outcomes across institutions; and software's telemetry made post-release performance a routine input to the next release. Public-sector procurement mandates could supply the demand signal the market does not.

Entry points for student teams

A student team could design and pilot a low-cost, low-liability POE kit — a short occupant survey plus a few weeks of loggers or existing BMS exports — on two or three campus buildings and measure how much design-relevant insight it yields per hour of architect time, producing the cost/benefit case a firm needs. A policy/business team could draft the standard "feedback stage" scope, fee and contract language, and the no-fault data-sharing agreement, and test them with local firms and a public client. A data team could analyze why so few AIA 2030 signatories report measured energy data and prototype an automated pathway from utility data to the DDx-style report. Skills: architecture, building science, survey design, contract law, data engineering.