When everyone tells me to use AI, I want to know which work in my company is actually worth handing over, so I do not spend three months on something that saves two hours a week.

Job context

Who

An owner who personally approves every technology spend

When

After a seminar, or after a competitor announces they use AI

Trigger

An employee asks why the company is not using it yet

Situation

Everyone talks about capability; nobody talks about which task in this company

Constraints

Nobody inside is technical enough to judge, and the experiment budget is small

5 Pains

• Cannot describe the current process

  High

  The process lives in four people's habits and has never been written down.

  Root cause: Small businesses run on tacit understanding, and tacit understanding cannot be handed to a machine

• Demos look brilliant; the real work does not

  High

  Demos use clean data; the company's data lives in screenshots and chat messages.

  Root cause: Demo makers choose their inputs; companies do not

• Cannot measure what is being saved

  High

  Nobody ever measured how long the old way took, so there is nothing to compare with.

  Root cause: No baseline was taken before changing

• Staff fear replacement and do not cooperate

  High

  The person who understands the work best has the most to lose by describing it.

  Root cause: Nobody says in advance what happens to the person after the work is automated

• Wrong answers arrive looking convincing

  High

  This kind of wrong is far harder to catch than an error message, because nothing turns red.

  Root cause: Catching the error requires being better than the output

5 Desired Outcomes

• Name one specific task worth trying first

  Functional

• Know how many hours the old way costs

  Functional

• Run the test in two weeks instead of three months

  Functional

• Staff describe how they work without fear

  Social

• Catch wrong output before it reaches a customer

  Functional

5 Existing Solutions

A solution is not the same thing as a product — a customer can hire a behaviour or a workaround too.

• AI course for businesses

  Service

  Teaches what tools can do; does not answer which task here is worth handing over.

• Hire an implementation consultant

  Service

  The only party willing to sit and map the current process, billing hourly to do it.

• Try an AI assistant on daily work

  Behaviour

  How people actually learn where it fits, and nobody calls it a project.

• Industry software with AI features

  Product

  Requires no process mapping because it assumes one — someone else's.

• Ask in an owners' group

  Behaviour

  Hears what worked, always from the people for whom it worked.

2 Opportunity Gaps

• The person who understands the work best has the most to lose by describing it.

  Why existing solutions fail: Every implementation method starts with interviewing the person doing the work, and none answers the question that person is actually asking: then what happens to me.

  Potential opportunity: Stating what happens to that person, in writing, before asking them to describe the work.

• Nobody measures the old way before replacing it, so nobody can tell whether the new way is better.

  Why existing solutions fail: Measuring before changing produces nothing visible, so it is always skipped in favour of the interesting part.

  Potential opportunity: A way to measure the old process in under an hour, doable before buying anything.