The 7 questions I ask before touching your data
You can have data and reports and still feel like you are guessing.
That is not a reporting problem. It is a confidence problem.
I do not start by querying your database.
Because starting with the data instead of the decision is how you spend money and still do not get an answer you trust.
Before I touch anything, I want one thing clear: can your data answer the question you care about, and can you trust the answer enough to make the decision?
Before I touch your data, I ask seven questions. They are the guardrails that tell us whether your data is ready to support a decision, and what is missing if it is not.
To keep it concrete, I will use one common decision as the running example: hiring another technician this quarter.
1. What decision are you trying to make?
Why it matters: If the decision is fuzzy, the work will be too.
Name the decision in plain language.
For example: “Do we hire another tech this quarter, or hold?”
2. What question would let you make that decision?
Why it matters: The question determines the data, not the other way around.
Keep it to one sentence. If it takes a paragraph, you are not ready.
For example: “Do we have enough profitable work, consistently, to keep another tech busy without breaking our margins?”
3. What does “good” look like?
Why it matters: Without a target, you cannot judge the data.
Define what success means in measurable terms.
For example: “If we hire, we need to stay above 80% utilization, maintain response times, and keep gross margin above X.”
4. What numbers would prove it?
Why it matters: You cannot make a decision with vibes.
List the few key metrics you would trust.
For example:
- Utilization by tech, by week
- Jobs completed vs. backlog
- Average time to schedule
- Gross margin by job type
- Repeat callbacks or rework rates
5. Where do those numbers come from?
Why it matters: If the source is unclear, the number is not trustworthy.
For each metric, identify the system of record.
For example:
- Scheduling tool
- Accounting system
- Field service app
- CRM
- Spreadsheet someone maintains “because the system doesn’t have it”
6. What is missing, messy, or inconsistent?
Why it matters: This is where confidence breaks.
Call out what you already know is shaky:
- Job types not coded consistently
- Time entries missing
- Revenue posted late
- Tech names spelled differently
- Work orders closed without notes
- Margin not tied to the job
If you cannot say what is messy, you will discover it the expensive way.
7. What will you do if the data is wrong?
Why it matters: If you do not have a fallback, you will freeze.
Decide what happens when the data is incomplete:
- Do we pause the decision?
- Do we use a conservative assumption?
- Do we validate with a manual sample?
- Do we fix the upstream process first?
This is the difference between “we need perfect data” and “we need dependable data.”
If your data cannot answer these questions clearly, that is not failure. It is clarity.
It means you now know what to fix, what to define, and what to stop guessing about.
If you want help turning these questions into a real, dependable single source of truth, that is exactly what I do.