Lowballers, Highballers, and Capacity Pricers: Instrumenting the Bias in the Estimating Room

Estimating is partly math and partly behavior — and every firm knows it, even if nobody says it in the QA meeting. Put the same takeoff in front of three estimators and you won’t get three random numbers; you’ll get three postures, each consistent, each rational, each invisible in the deliverable. Under uncertainty, people don’t produce noise. They produce defaults.

The three archetypes

They show up in every estimating room, under different names:

• The lowballer. Often the pursuit-minded optimist — or the estimator whose numbers keep getting challenged as “too high” until they pre-emptively shave. Their estimates make budgets look achievable and bid openings look like ambushes.
• The highballer. Usually someone who’s been burned — one brutal review after a low estimate, years ago, and every number since carries armor. Their estimates never trigger the bad meeting… and quietly park client money, oversize budgets, and invite “why is this so expensive?” challenges of a different kind.
• The capacity pricer. The subtlest one: their numbers drift with the room’s workload and mood — tighter when the team is hungry for the pursuit, looser when everyone’s slammed. It’s the contractor’s capacity-pricing instinct, translated into consulting, and it means the same firm produces different postures in March than in October.

Two things to say plainly. First, these are adaptations, not character flaws — each posture is a reasonable response to how the firm has historically punished and rewarded misses. Second, the problem isn’t that the postures exist; it’s that they’re unmeasured. A client comparing two of your offices’ estimates isn’t seeing two market realities. They’re seeing two personalities — and reading the difference as inconsistency, because it is.

Great firms don’t eliminate bias — they instrument it.

Evidence doesn’t remove judgment. It disciplines it.

The naive fix is to demand estimators “be objective,” which works exactly as well as demanding anyone be objective. The real fix is a reference point: put a market benchmark next to every consequential number, and something structural changes. The posture doesn’t disappear — the highballer is still cautious, the lowballer still optimistic — but the posture becomes visible, named, and sized. “I’m 8% above the market’s center on this item” is a professional position that can be discussed, defended, and recorded. “That’s my number” is a personality trait wearing a hard hat.

Once postures are measurable, three practices follow naturally:

• Make deviation a conversation, not a confession. The review question stops being “is this number right?” (unanswerable until bid day) and becomes “we’re above/below market center here — what do we know that the market doesn’t?” Sometimes the answer is real knowledge — an owner quirk, a scope subtlety. Sometimes the answer is the archetype talking. The conversation sorts them.
• Track posture over time, per team. Estimate-vs-benchmark-vs-award, scored across the portfolio, turns “Jim runs high” from hallway folklore into a measured, correctable offset — and it catches capacity pricing, which is invisible project-by-project and obvious in a trend line.
• Calibrate the incentives you’ve accidentally built. If one kind of miss triggers brutal reviews and the other passes silently, your firm is training an archetype. Symmetric scrutiny — high misses reviewed as seriously as low ones — is the cheapest bias correction available.

The other half: crisp scope

Behavioral slack feeds on ambiguity. When scope is fuzzy, every estimator fills the gaps with their default posture — the highballer pads for the unknowns, the lowballer assumes they’ll resolve favorably, and the eventual change orders prove somebody wrong. Tightening scope definition does for the inputs what market evidence does for the outputs: it shrinks the space where posture operates unexamined. Crisp scope plus market truth tightens the screws end-to-end — ambiguity out of the quantities, folklore out of the prices, and what remains is the thing you actually pay engineers for: judgment, applied where the evidence runs out, on the record.

Where PinPoint fits

You will never staff an estimating room with people who have no priors — and you wouldn’t want to; the priors are experience. The goal is a room where every prior has a number next to it. Instrument the bias, and it stops being a liability. It becomes calibration data.