ISTP Workplace Habits

Most technical departments have someone who quietly solves the things that were supposed to be impossible, who nobody quite manages, and whose annual review is a page of nothing much.

That is usually a Tinkerer, and this section is about the gap between what they do and what any record shows.

You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.

— Buckminster Fuller

The first ninety days

This type learns by touching the work. They will read very little documentation, ask few questions, and instead go and find out — opening the system, running the process, breaking something safely to see what happens.

They will also say almost nothing. A Tinkerer three months in has usually formed accurate views about which processes are pointless and which colleagues are competent, and has volunteered none of it. Asked directly they will answer, flatly and usefully. Not asked, they will simply route around whatever does not work.

The value shows up in month five, when something fails and it turns out one person already understands why.

Why supervision degrades the work

Worth stating precisely, because it is usually misread as an attitude problem.

This type’s method is exploratory: try, observe, adjust. It does not produce a plan in advance because the plan is what the process generates. Being required to justify each step before taking it does not slow this type down by twenty percent — it removes the method, and what you get instead is a worse version done to somebody else’s specification.

That is why a Tinkerer will accept lower pay to avoid close management, and why they are so often described as difficult by managers who were, from their own point of view, simply doing their job.

The workable arrangement is narrow and effective: agree the outcome and the deadline, and ask nothing about the method. Managers who do this get extraordinary work. Managers who ask for a plan get a plan and lose the thing they hired.

Why nobody has a record

This type does not report, does not claim credit and does not explain.

None of that is modesty or evasion. It genuinely does not occur to a Tinkerer that anyone needs to know — the thing is fixed, the fixing is the event, and describing it afterwards is redundant work about work.

The consequence is that the most capable person in a department frequently has the thinnest file. Promotion, pay and interesting assignments run on records, and this type’s record is missing almost everything they have done.

Correcting the record is not self-promotion. It is fixing a fault in a system, which is the one framing this type will actually act on. One sentence a month to whoever decides things: here is what I did, here is what would have happened otherwise.

Meetings

This type contributes almost nothing and dislikes being there, and both are worth managing around rather than correcting.

Tinkerers will not speculate, will not react to a half-formed proposal, and find a discussion with no decision attached close to unbearable. What they will do is answer a direct, specific question with an accurate, brief, useful answer — and that answer is frequently the thing that determines whether the plan is possible.

The correct use of this type in a meeting is to ask them one thing and let them leave.

Where it gets difficult

Process for its own sake. Documentation, status updates, forms, the ceremony around work rather than the work. This type does not merely dislike it; they quietly stop doing it, which becomes a genuine problem in regulated environments and is worth being honest about as a real limitation rather than a preference.

Colleagues who talk more than they deliver. Assessed quickly, written off permanently, and treated accordingly — which is visible to everyone and is one of the few ways this type generates political trouble.

Being told rather than shown. A change that arrives as an instruction gets complied with resentfully. The same change with a demonstrated reason gets adopted immediately.

Long stretches with nothing hard. This type deteriorates in a settled process, and the deterioration presents as unreliability rather than as boredom — which is what it actually is.

Why the resignation has no warning

A Tinkerer with a problem does not raise it. There is no complaint, no conversation, no visible dip. Then they leave, having decided weeks earlier, and it is not negotiable — counter-offers rarely work because the decision was made against a condition rather than a number.

Employers lose the person who understood the system and discover over the following year exactly how much of it was in one head and nowhere else.

As a manager

Better than expected, in a specific way.

Tinkerers manage by giving people real problems and getting out of the way, and they will defend a competent person absolutely. Their teams get autonomy that is genuinely rare, and the capable ones flourish under it.

What they struggle with is everyone else. This type under-communicates almost completely, does not do encouragement, and will not have the conversation with the person who is not delivering — they will simply route the work elsewhere, which is worse for everybody and takes years to resolve.

One habit closes most of the gap: a short, scheduled, low-content update to each person. Not because it is useful to you. Because the absence of it is not read as trust — it is read as absence.

What a manager has to do

Give the outcome, not the method. Ask one specific question rather than requesting a general update. Make sure their work reaches the record, because they will not put it there. And take the first mention of a problem seriously, because there will not be a second — the next thing you hear will be a notice period.

One thing to watch for this week: send one sentence about what you finished and what it prevented. To one person who decides things. That is the whole ceiling.