Most contractors of any size have go/no-go criteria written down somewhere. Far fewer can say the process actually ran on their last ten pursuits—every reviewer heard from, every gate decided on time, every decision documented. The gap between the process on paper and the process in practice is where bad bids get through, and it rarely has anything to do with the criteria themselves. This article is about designing a go/no-go process for construction pursuits that survives contact with a busy week.
Where go/no-go processes actually fail
The failure mode is seldom analytical. Firms generally know what makes a project worth pursuing: the client's financial capability and history, the fit with current backlog, the margin potential, the competitive field, the schedule and bonding exposure. When a pursuit that should have been declined gets chased anyway, the usual cause is that the evaluation never really happened—or happened so late that the estimating spend was already committed.
The mechanics are familiar. A scorecard goes out to five reviewers and three respond. The BD lead, who wants the project, interprets silence as consent. The one dissenting reviewer is traveling, and by the time their concerns surface the proposal team has mobilized. Nobody is behaving badly; the process simply depends on a level of sustained coordination that no one owns. Processes that rely on somebody remembering to chase somebody else degrade quietly until the firm is back to deciding by instinct, with a scorecard as decoration.
So the design problem is not "what should we score?" It is "how does this process complete when everyone involved is busy?"
Fewer criteria, with a clear owner for each
A scorecard that asks forty questions gets skimmed; one that asks twelve pointed questions gets answered. Resist the urge to encode every consideration—score the factors that have actually distinguished your good projects from your bad ones, and leave room for a short narrative where a reviewer can flag what the rubric missed. Assign each section to the role genuinely qualified to answer it: operations scores team availability, risk scores contract and bond exposure, BD scores the client relationship. Reviewers respond faster when they're asked only about what they know.
Give every gate a clock
An evaluation with no deadline is a suggestion. Decide, in advance, how long each gate may take—many firms target 48 to 72 hours for an initial screen—and define what happens when the clock runs out. A reminder cadence, an escalation path, and a rule for proceeding without a chronically silent reviewer are not bureaucratic decoration; they are the difference between a process and a hope. The same goes for absences: if a reviewer is out, the process should route to their alternate rather than stall for a week.
One clean package per decision
Decision-makers disengage from go/no-go processes when a "decision" means reconstructing the state of play from six attachments across four email threads. The executive gating a pursuit should receive a single package: the composite score, the breakdown by criterion, each reviewer's narrative, and a clear question—advance, stop, or send back for rationale. When deciding takes five minutes, executives decide promptly, and the whole process inherits their speed.
It's worth saying that the package informs the decision rather than making it. As one consultant argued in a Construction Dive piece on bid strategy, leaders should stay willing to override the scorecard when a genuine strategic reason exists—a small project for a client you want a decade-long relationship with, say. A good process makes the override visible and deliberate instead of pretending it never happens.
Feed the outcomes back
The rubric is a hypothesis about which projects you should chase. Test it. Once or twice a year, line up the scores your pursuits received against how they turned out—win rate, margin at close versus margin at bid, the disputes that materialized—and adjust weights accordingly. Firms that do this end up with scorecards that encode their own history rather than a template's assumptions. Firms that don't are scoring against a guess made years ago.
The coordinator problem, solved without a coordinator
Everything above is straightforward to describe and expensive to operate by hand, which is why so few firms sustain it. The chasing, merging, and packaging typically falls to one person, and the process is exactly as reliable as their least busy week.
This is the job Bex Risk was built to do. A pursuit starts with one email—"AcmeCo wants us to bid the Oakdale renovation, ~$12M, RFP due 11/14, BD lead is Sara Lee"—and Bex takes it from there: it builds the tracking record, sends each reviewer an Excel scorecard locked to exactly their cells, scores the returns deterministically by your weights, reminds the quiet, routes around the out-of-office, escalates on timeout, and delivers the decision-maker one clean package per gate. The executive replies "let's go" or "pass," and Bex proceeds accordingly. Reviewers never learn a new tool, because the scorecard arrives in their inbox as ordinary Excel—the design philosophy we describe in meeting users where they are. Human judgment stays in charge at every step, per our humans-in-the-loop approach; Bex runs the ceremony, not the decision.
The result is a go/no-go process with the discipline of the version on paper and none of its dependence on heroic coordination. Gates close on time whether or not anyone remembered to chase.
If your firm's process exists but doesn't quite run, that's the specific problem we solve. Tell us how your gates are supposed to work—the contact link is below—and we'll show you what they look like running by themselves.