Techniek
Tools & Technology

How often should your dashboard refresh?

Live sounds better than daily, and it costs more. How to work out which refresh frequency your business actually needs.

As soon as a dashboard exists, a question follows it: how current is that?

And the honest answer is almost never as often as possible. Refreshing more costs money, in licenses and in infrastructure, and at most SMEs it produces not a single extra decision.

What it depends on

Two things, and not much else.

The first is whether you have a physical place where something can change today. A shop, a warehouse, a crew on the road. There current information has value, because there is someone who can still act on it today. You see that trade is slow and you send someone home. You see an article flying off the shelf and you call your supplier.

The second is your number of transactions per day. If ten invoices come in per day, your picture changes so little between two refreshes that daily is just as good as hourly. At hundreds of transactions a day that is different.

Put those two together and you have your answer. A physical business with many transactions benefits from a high frequency. A services company that sends twenty invoices a month does not, however much people want it to.

What people are actually asking

Pay attention to what sits behind that question, because it is rarely currency.

Usually what is being asked is: can I trust this. Someone who is not sure whether the numbers are from today or from last month is not going to use them.

You do not solve that with more refreshing. You solve it by putting on the screen when it was last refreshed, and by sending an alert when it failed. A dashboard that refreshes once a day and honestly says when is trusted more than a dashboard that is supposedly live and quietly sat stuck for three days.

That is in fact the real risk of a high frequency: nobody notices when it stops. With a daily refresh it stands out. With a refresh every half hour nobody checks any more whether it went through.

The practical trade off

Start lower than you think you need. Daily, overnight, for most businesses.

Then watch for a month whether there were moments where you would have made a decision differently with more current numbers. Not whether you would have enjoyed it, whether it would have changed a decision. Usually that is zero or one, and then you know enough.

If there are more, you scale up, and then you also know straight away which part of your report needs to be more current. Often that is one screen and not the whole thing, and that makes a considerable difference in cost.

What makes it expensive

The jump from daily to much more frequent is rarely a technical matter. It is a licensing matter, because the ability to refresh often usually sits in a more expensive tier.

So weigh that against what it delivers and not against what it costs in technology. For a business with shops and many daily transactions that jump pays for itself quickly. For a business that invoices twice a month it is a subscription to a feeling.

The claim

Live is not a property of a dashboard, it is a property of the decision you make with it. If nobody does anything differently today because of a number from an hour ago, then a daily refresh is not worse. It is just cheaper.

Gregory Moureau