Case study · Trusted reporting

Two staff were on the phones. The reports said they didn't exist.

Every call report the business ran was missing two team members. It wasn't a data entry slip. A hardcoded staff list was buried in six reporting scripts. AI found it, we fixed it at the source and checked it against live call data, all in one session.

Found in one sessionFixed at the sourceChecked against live data
Invisible in the reports
2
team members missing from every call report since they started
Reporting scripts with the same flaw6
Root cause1 hardcoded list
Time to find and fix1 session
Checked againstLive call data
Sound familiar?

Your reports look complete. Have you ever checked?

✕You make calls on staff performance from numbers you've never verified
✕New team members start working, but do they show up in the data?
✕Nobody on the team owns whether the reports are right
✕A broken report doesn't throw an error. It just quietly shows the wrong number
How we did it

What we did, step by step.

1

A person noticed what the reports didn't

The owner spotted that two team members making calls were showing zero activity.

2

AI traced it straight to the code

An AI review of the reporting scripts found the cause immediately. Six scripts shared one hardcoded staff list, and neither new starter was on it.

3

Fixed at the source, not patched over

We pulled the real staff IDs from the CRM's live API, nothing assumed, and updated all six scripts.

4

Checked every script still ran

All six were tested after the change, so the fix didn't break something else.

5

Proved it with a real call

The next outbound call from one of the new starters showed up correctly in live data. Checked, not assumed.

GoHighLevel APIPythonAI code review
Why it matters

Bad numbers don't look broken. They look normal.

That's what makes them expensive. Every decision built on them looks reasonable at the time.

Risk

The wrong people get the credit

Performance reviews, coaching and commissions all lean on call numbers. Leave someone out and the whole comparison is off.

Risk

Decisions get made on half the picture

Who to hire, where to send leads, who needs help. Every one of those calls gets made from the report.

Before & after

From "Looks Right" to "Is Right"

Before

Two people, zero calls

Two team members making calls with nothing to show for it in the reports. Performance numbers that looked complete but weren't.

After

Every person counted

All six scripts fixed and checked. Every team member's calls attributed correctly. Reports the owner can make decisions from.

Check yours in 5 minutes

Try this before you trust your next report

1

Pick your newest team member

New starters are the most likely to be missing, because systems get set up before they arrive.

2

Count their calls or jobs this week

Ask them, or check their phone or calendar.

3

Compare it to your report

If the numbers don't match, your reports have a gap. And if one person is missing, check everyone.

What happened

Start to finish

Owner flags two staff showing zero activity.
AI review traces it to a hardcoded staff list in six scripts.
Real staff IDs pulled from the live CRM API.
All six scripts updated and tested.
First new call correctly attributed in live data.

"Nobody was hiding anything. The report just never knew they existed."

Operations Intelligence
Free 20-minute reporting check

Can you trust your numbers?

We'll look at your reports with you and check whether every person, call and deal is actually being counted.

BOOK MY FREE SESSION → Takes 30 seconds to book.
Not ready to talk yet?

Keep looking around

© 2026 Operations Intelligence | operationsintelligence.com.au