Standardize This Process or Leave Room to Adapt: Decision Tree

Why this matters

Every process decision lands somewhere between "one right way, every time" and "use your judgment." Pick wrong in either direction and it costs you. Lock down a task that needs judgment and your best techs quietly work around the rule until it means nothing. Leave a task loose that needs consistency and quality swings job to job, callbacks climb, and a new hire has nothing to learn because there is no standard to teach. This tree sorts a task into the right bucket before you write the checklist.

Start here: what breaks if two people do it differently?

Answer that one question honestly and most of the decision is made.

• If the answer is "someone could get hurt, it fails inspection, or the customer gets a worse result," lean toward standardize.
• If the answer is "nothing, the outcome is the same by a different road," lean toward leaving room to adapt.

The mistake is standardizing for your own comfort rather than for the outcome. A rule that exists so you feel in control, not because variation causes harm, is the kind of rule the crew learns to ignore.

Branch 1: is it a safety or code step?

Standardize hard, no exceptions. Lockout and verification, gas leak checks, confirming a circuit is dead before you touch it, ladder and fall setup. Here the standard is the floor, not the suggestion. "Use your judgment" is exactly the wrong instruction for a step where the failure mode is an injury. A skipped safety step is never a process-improvement conversation, it is a stop-work issue.

Branch 2: is the consistency itself the product?

Standardize. Some steps deliver value precisely because they come out the same every time: how the phone gets answered, how an invoice reads, how a job gets closed out and photographed, how the truck gets stocked. The customer and the office experience the sameness. When these drift, the shop stops feeling like one business and starts feeling like a bag of individual contractors.

Branch 3: does the situation vary more than the task?

Standardize the outcome, leave the method flexible. Field work is full of jobs where the goal is fixed but the path depends on what you find. Write the standard as a result, not a script: "the work area is cleaner than you found it" is a standard; exactly how you protect the floor depends on the site. This is the bucket most tasks belong in, and the one shops get wrong by trying to script every keystroke.

Branch 4: is it diagnostic or craft judgment?

Leave room to adapt, but standardize the guardrails. You cannot script which test a tech runs next or how they read a symptom; that is the skill you hired. What you can require is that findings get written down, that a second opinion gets pulled in past a time limit, and that nobody guesses on a part without checking. Standardize the rails, not the driving.

The comparison

When to pick which

Default to standardizing the what and the when, and adapting the how, unless safety or code says otherwise. If you find yourself writing a rule for a step where any reasonable method gives the same result, you are over-standardizing; convert it to an outcome standard. If you find two techs producing visibly different results customers would notice, you are under-standardizing; lock it down. Re-sort a task whenever it moves buckets: a craft step done a thousand times often earns a standard, and a standard the field keeps breaking may have been craft all along.

Recap

1. Ask what breaks if two people do it differently.
2. Safety or code, standardize hard.
3. Consistency is the product, standardize.
4. Situation varies more than the task, standardize the outcome, free the method.
5. Diagnosis or craft, adapt the method, standardize the guardrails.

References

• See related: The Standardize-Then-Improve Cycle
• See related: Why People Skip a Process and What It Tells You
• Trade-standard practice for field-service work instructions and safety procedures