Target-Value Design (TVD) and the co-located "Big Room" are the two operating practices that make Integrated Project Delivery (IPD) work day to day. TVD turns cost from an outcome that is estimated after design into a constraint that drives design; the Big Room is the collaborative workspace — physical, virtual, or hybrid — where the integrated team designs to that target in real time. Together they are how an IPD team converts a shared-risk relational contract into a building that lands on budget without value-engineering the clinical program away at the end.
In conventional design-bid-build (DBB) and most design-build (DB) practice, the team designs to a program, estimates the result, and discovers overruns at milestone pricing — typically at the end of schematic design (SD), design development (DD), and construction documents (CD). Each overrun triggers a redesign-and-re-estimate loop, and the loop that lands closest to the bid is the most expensive to absorb because the design is already detailed. Value engineering (VE) at that stage tends to strip scope, downgrade systems, or compress contingency rather than rethink the article.
Target-Value Design inverts the sequence. The team establishes an Allowable Cost (what the owner can and will fund — set by the business case, pro forma, reimbursement model, or capital budget), validates an Expected Cost (what a well-run market-rate version of this project would cost, built up from estimates and benchmarks), and then sets a Target Cost that is deliberately below the Expected Cost — the gap is the improvement the team commits to engineer in. Design then proceeds toward the Target Cost as a fixed input, not a number to be discovered afterward.
The mechanism that makes this real is continuous estimating. Cost is re-validated against the target on a near-weekly cadence as design choices are made, not at three or four milestone gates. A wall assembly, an air-change strategy, a structural grid, or a headwall configuration is priced as it is conceived, so the team steers continuously rather than correcting at milestones. The constructor and key trade partners — who in IPD are at the table from validation — supply this real-time cost intelligence, which is why TVD is paired with IPD (or at least with a deeply collaborative CMAR/progressive-DB arrangement) and is impractical under arm's-length DBB.
| Term | Meaning | Set by |
|---|---|---|
| Allowable Cost | The most the owner can/will spend; the funding ceiling. | Owner's business case, pro forma, capital plan, debt capacity |
| Market / Expected Cost | What a conventional project of this scope would cost at market. | Validated estimate built from benchmarks + trade pricing |
| Target Cost | The team's committed cost goal, set below Expected Cost. | Integrated team, by agreement, at validation |
| Estimated Cost | The current running estimate as design develops. | Continuous estimating, near-weekly |
The discipline is that Target Cost ≤ Allowable Cost, and the team designs to keep Estimated Cost ≤ Target Cost throughout. If the estimate trends above the target, the team re-engineers the article early — when changes are cheap — rather than VE-ing detail late.
TVD begins in Validation (sometimes "validation study" or "phase 0/1"), an explicit early period — often a few months — in which the integrated team tests whether the owner's program, schedule, and quality aspirations can be delivered within the Allowable Cost. Validation produces a Validation Report: a go/no-go basis-of-design with a substantiated Target Cost, a risk register, and the conditions of satisfaction the team will steer by. Crucially, IPD validation gives the owner an early, legitimate off-ramp — if the program cannot be met within budget, that is surfaced in weeks, before significant design fees and commitment are sunk.
Once design is underway, the work is broken into cluster teams (also called cluster groups or, in some practices, "small groups"). Each cluster owns a coherent slice of the building — for a hospital, typical clusters include:
Each cluster is cross-disciplinary by design: it pairs the relevant designers, the constructor's lead, the trade partners who will install the work, and — for healthcare — clinical and facilities representatives. Each cluster carries its own piece of the Target Cost as a budget, estimates continuously against it, and is accountable for staying within it or trading capacity with another cluster through the governance team. This is the structural reason TVD needs co-location: a cluster cannot estimate-as-it-designs if the estimator, the designer, and the installer are in three different offices on three different cadences.
TVD teams commonly use Set-Based Design (SBD) (a Lean/Toyota-derived practice): rather than picking one solution early and refining it, the cluster carries multiple feasible alternatives in parallel, prices and tests each against the target and the clinical requirements, and converges on the surviving option only when the trade-offs are understood. SBD reduces expensive late rework because the team commits last responsibly — it keeps options open until the cost and performance consequences are known, which is exactly when a hospital's MEP, infection-control, and equipment decisions interact most.