An adventure in data first consulting

Your business runs on systems and data. They should be working harder for you.

That's the line on my website. Earlier this month I walked into an Australian video security operator's boardroom with five years of their production data prepared and summarised. Events, devices, alarms, customers, sites. All sitting in backend systems, generating value day in and day out, and nobody had ever asked what any of it meant beyond the operational view.

The brief I wrote myself: let the numbers do the talking.

91 things the dashboard never showed them

An hour in, we inventoried every device across the fleet by configuration state. The pull surfaced 91 cameras sitting unconfigured in the backend. Present in the system, absent from the main operational dashboard. Invisible to the team running the fleet every day.

Nobody had hidden them. The dashboard didn't have a view for devices in that state, so they never appeared to anyone. The fleet had been running for years while leaking value through a gap nobody could see.

The founder's feel for the fleet's health was wrong by about sixty devices. He couldn't have known. The data was there; the dashboard wasn't showing it.

Invisible to the market

Late morning I ran a live AI search for Melbourne construction camera providers. Deliberately specific: the words a buyer would use. Seven operators came back. This business, operating in that exact segment with a solid reputation and a respectable fleet, was not among them.

None of the seven mentioned the strongest competitive advantage this operator has. From inside the business, that capability had become so taken for granted it had stopped being articulated. Outside the business, it was invisible. Uncontested and uncommunicated. Potential customers had no way to choose on it.

The room went quiet.

The win the founder couldn't see he'd made

We pulled five years of purchase order processing for their largest project home builder customer. Two years ago, the median processing time was 25 days. The slow tail stretched past 60. In the last twelve months the median had dropped to 0.3 days. Same day, for most orders.

Nobody had computed that number. The founder had made a series of small changes over two years, manual adjustments that felt right at the time, compounding into a 99% improvement. No internal report had quantified it. No dashboard had surfaced it.

He didn't know he'd done it until the data told him.

Steering the analysis in real time

One participant asked to see the alarm data for the last twelve months only, rather than the five year view we'd been working from. The analysis ran live, carrying forward every classification decision from the earlier session.

It surfaced something nobody in the room had seen. Across five years, the genuine alarm rate sat at around 12.5 percent. Over the last twelve months it had dropped to 4.4 percent. False alarms climbing proportionally, and no one had the view that would have told them.

Over those same five years, the system had processed roughly 1.3 million events. About 1,800 resulted in a police callout. One in 700.

The founder looked at the number and said he'd never been able to put a figure on it before.

What came out of the room

One workshop. One day. Data the business had been generating for five years.

A fleet count wrong by sixty devices. A market position uncontested but invisible. A run of operational wins the founder had made without noticing. And a ratio so lopsided nobody in the room had ever computed it.

None of this required new systems, new data pipelines, or an eighteen month roadmap. Everything that surfaced in that room was already in the operator's data. All it took was a day and someone who knew what to ask.

Your business runs on systems and data. They should be working harder for you.