Decisions, Decisions: Why Business Choices Are Harder Than They Look

Chris Terrell
Aug 21
2 min read

We all make more decisions than we want to—most of them on autopilot. In business, that autopilot shows up as “busy” meetings, half-formed strategies, and a trail of choices nobody can evaluate later. The result: process debt. Here’s how to break the loop.

The real problem: we don’t practice decisions

Most teams treat decisions like range time with a bad golf swing—more swings, same slice. Reps alone don’t improve judgment; purposeful practice does. That means defining the decision, predicting outcomes, and coming back to compare results against the prediction.

Start with strategy (for real)

Before you grade decisions, confirm you actually have a strategy:

Problem: Whose pain are we solving?
Policy/Plan: What actions will we take and how will we resource them?
Payoff: What outcome should be visible when it’s working?

No resourcing = no strategy. No explicit pain = misallocated resources.

Make the invisible tensions visible

Great operational example: purchasing bias. Using a simple accuracy measure (like mean absolute percentage error) exposes whether buyers systematically over- or under-stock. Zoomed out, every strategic decision has a similar tension (growth vs. risk, speed vs. quality). If you don’t name the tension, you can’t judge the trade.

Decide where it matters: constraints

From Goldratt’s The Goal: systems improve when you relieve constraints. Decisions that ignore the current constraint manufacture process debt. Ask before you decide: Does this change relieve a present constraint? If not, you’re redecorating the waiting room while the operating room is jammed.

Force clarity with simple guardrails

Zero-based budgeting (lightweight): Once a year, justify major spend from zero. It won’t erase bias, but it forces intention on who gets resources and why.
Cohorts & timelines: Compare like with like. Tie decisions to expected time horizons; long-arc bets need long-arc grading.
One decision per quarter (per leader): Pick one consequential decision to forecast and retro. Quality over quantity.

Perfect practice > more practice

Treat each major decision like a micro-experiment:

Statement: What decision are we making, and what constraint does it target?
Prediction: What do we expect to happen by when (leading and lagging metrics)?
Exposure: Where the record lives (one canonical home; link it elsewhere).
Retro: On the review date, grade the call—keep, tweak, or kill.

Expect the social curve

Good calls often look political or fuzzy before they land and “obvious” after they succeed. Don’t chase applause mid-flight. Anchor on the prediction and the constraint you set at the start.

Process Debt Truth: Decisions without a forecast and a follow-up are just opinions with paperwork.