On this page
Dashboards became the default output of hospitality software for an understandable reason. They demo well, they are genuinely useful the first week, and they make a large amount of data feel manageable.
The disappointment tends to arrive around month three, when the venue has a beautiful live view of itself and is not running any differently than before.
What a dashboard is actually good at
Two things, and both are real.
It answers a known question fast. If you regularly need to know today's takings against last Tuesday, a dashboard removes that work permanently.
It makes a state visible to several people at once. A number on a wall that the team can see is a coordination device, and that is worth something no report achieves.
Neither of those is nothing. The mistake is expecting a third thing that dashboards structurally cannot do.
The three things it cannot do
It cannot tell you what you did not think to ask
Every panel exists because somebody anticipated the question. That is fine for the general case and poor for a bar, where the useful findings are specific: comps clustering in one hour, variance in two products, a section slow every Thursday. None of those are on a standard dashboard, because a standard dashboard is designed for every venue.
It cannot assign
A dashboard is addressed to whoever opens it, which in practice is the owner and occasionally a manager. A finding without an owner produces awareness, and awareness is not an intervention.
This is the gap that matters most and it is rarely discussed, because it is not a data problem and vendors in this category are data companies.
It cannot tell you whether anything changed
A dashboard shows the current state. It has no memory of what you decided last week or whether the decision worked, so it cannot distinguish a venue that is improving from one that is being observed closely while staying the same.
Why this matters commercially
A venue that buys a dashboard and sees no operational change usually concludes that analytics does not work for bars, and stops looking. That is the real cost: not the licence fee, but the reasonable inference drawn from a tool that was only ever built to do the first half of the job.
The category has spent a decade improving measurement and comparatively little on what happens after a measurement exists. Measurement is tractable, has correct answers and demos well. Deciding what a pattern means, and whose job the response is, requires a model of how a bar actually works, and it is harder to show on a screen.
What to ask instead
Three questions, of any product including ours.
Does it surface something I did not ask for. Does it name who does what. Does it record whether the thing happened and what followed.
A product that answers no to all three is a dashboard, which is a legitimate thing to buy at a dashboard price. The problem is that most of them are sold as though the answer were yes.
CoreTAP is built around the second and third. The comparisons put the other products in the category against the same three questions, including where they beat us.