Find the one broken thing
Most technology conversations start in the wrong place. The desire for a new platform, or AI, often masks the underlying cost center. The pattern is usually the same: the thing being built is rarely the thing costing the business money.
If the choke point cannot be named in one sentence, the business is not ready to rebuild.
The spreadsheet is usually the truth
When $50M companies run critical revenue processes in Excel because the system of record is untrusted, that is a diagnosis. It is not a technology preference.
The ops lead is not being stubborn. The number that actually matters is being protected: invoices that go out, jobs that get assigned, inventory that is real. If a new architecture does not move that number in weeks, it is a hobby.
Scaling does not start with a clean-slate platform. It starts by finding the constraints blocking growth and removing them in the order the business can feel. Some of the work might be ugly. It also funds the next stage.
One metric, or it is a guess
A useful audit does not produce a 40-page deck. It produces one number that can be defended in front of a sponsor.
- Hours lost per dispatcher per day, not "a better UX."
- Accrual leakage in dollars, not "the ERP is old."
- Time from quote to cash, not "should be on the cloud."
If the metric is mushy, the project will swell. Mushy metrics are for consultants. Sponsors need certainty.
Systems last, not first
Walk the floor. Sit with the person who actually closes the books. Ask what happens when the software lies. That conversation is worth more than any architecture diagram.
Once the broken thing is known, the build becomes obvious. Sometimes the answer is a thin proof on real data. Sometimes the answer is to not build at all. That is a valid result. It is cheaper than a rebuild that fails to move the number.
Find the one broken thing. Prove it matters. Then, and only then, spend real money.
