There’s a certain kind of store owner the software industry loves to complain about. She won’t just trust the dashboard. She exports the raw file and totals the cost column herself, in a spreadsheet, by hand. She checks the date on every report before she’ll let it near another one. She evaluates new software like she’s auditing it.
Vendors call this “resistance to technology.”
It isn’t. It’s quality control. Spend enough time with an owner who runs her store this way, and you come around to a different view: she’s not the problem. She’s the standard.
The disciplines
Watch an owner like this evaluate a new reporting tool, and you see a system — not a mood.
She totals the source file herself. Before the dashboard’s number means anything, she opens the export it came from and sums the column. If the dashboard says one thing and her total says another, the conversation stops until someone explains the difference. Not because she distrusts computers — because a number nobody has checked isn’t a fact yet. It’s a claim.
She knows what a real snapshot is. Her definition is precise: what the store physically had on hand, at that moment, full stop. A report that mixes two moments isn’t two reports — it’s zero reports. So she polices dates the way a jeweler polices carats.
She writes the parameters down. Every export she sends carries its settings in the file name — what it covers, what it excludes, the date it was pulled. Months later, there’s no archaeology required. The file says what it is.
She proposes the blind test. Her method for evaluating a new system is the honest one: don’t pre-fill the targets. Load the raw data and see whether the system arrives at the numbers she already knows are true. The test she actually runs: show me the numbers pulled from the data I’m already sending. If it matches, the tool has earned something. If it doesn’t, better to know now.
She refuses to rush. The refusal is a method, not a mood. Evaluation takes the time it takes. A vendor who needs her to hurry is a vendor who needs her not to look closely.
Why the industry gets her wrong
The software industry has a name for buyers like this: skeptics. Late adopters. The implication is always that the hesitation is emotional — fear of change, comfort with the old ways.
But look at what she actually does. She doesn’t refuse new tools. She refuses unverified tools. There’s a difference, and it’s the whole difference.
Her stance: until the data proves out, the prettiest dashboard in the world is still just theoretical. She isn’t rejecting the analytics. She’s sequencing them: data first, then trust, then use. Any engineer would recognize the posture immediately — it’s called validation, and entire professions are built on it.
The owner doing this by hand, between customers, during the busiest retail weeks of the year, is doing unpaid work her software should have done for her. That’s the real story. Her diligence is compensating for tools that don’t check themselves.
The vendor test hiding inside this
Here’s what her method reveals, and why we’d tell every owner to adopt it:
A vendor’s reaction to hand-validation tells you what they know about their own product.
A vendor who bristles — who calls the spreadsheet cross-check unnecessary, who wants the demo to move faster, who’d rather show you features than reconcile a total — is telling you where they expect that conversation to end.
A vendor with nothing to hide has an easy answer: good. Check. Leave the targets blank, load your own data, and see what we say it is. If the numbers hold, you’ve learned something about the software. If the vendor welcomes the test, you’ve learned something about the vendor.
She figured that out without any of us. That’s why she’s right.