Back to blog
Analytics3 min read

Four minutes a day spent setting filters

Four minutes sounds cheap. Multiply it by three people and a full year and it comes to more than a working week. The hours are the cheap part of the cost.
Four minutes sounds cheap. Multiply it by three people and a full year and it comes to more than a working week. The hours are the cheap part of the cost.

Most daily work in an analytics tool isn't analysis. It's re-setup: pick the date range, pick the platform, exclude cancelled orders, pick the channel, pick the store. Then the work starts.

The arithmetic

  • 4 minutes of re-setup each day
  • 22 working days a month = 88 minutes, about an hour and a half
  • 3 people doing the same thing = 4.5 hours a month
  • Over a year: roughly 54 hours, more than a full working week

That's the visible cost, and it's the smaller of the two.

The expensive part

Those minutes read as friction. Friction reduces how often anyone opens the tool, and that reduces how many decisions get made from numbers rather than from memory.

The tool doesn't fail loudly. It gets opened weekly instead of daily, then before meetings instead of weekly, then only when someone asks a question. By that point it's become a place you go to justify a decision rather than to make one, and what got it there is small repeated friction rather than missing data.

Before your real work starts there's a small ceremony every morning: open the dashboard, select last 7 days, filter to the right store, exclude cancelled orders, pick the channel, and now you can look at the thing you came for. Nobody counts it as work and everybody does it. A saved view makes those four minutes a one-time cost, and a shareable link means the rest of the team never pays it at all.

The screenshot problem

The common way to share an analysis inside a team is a screenshot in WhatsApp.

The screenshot has no filters, no date range, and no way to check what produced it. A month later nobody can say whether it included cancelled orders or which store it covered. It gets forwarded, someone quotes it in a meeting, and then a different number contradicts it and nobody can reconcile the two.

A saved view with a link is the same information plus its own provenance. See sharing an analysis.

When a view is worth saving

Three tests:

  1. Do you set this combination up more than once a week? If not, it's a one-off query rather than a view.
  2. Does someone else need exactly this? If yes, it needs a link rather than a description.
  3. Does it end in a decision? A view you look at without acting on it isn't a tool.

Most teams need three or four views in active use, not thirty. A library of saved views nobody opens has the same problem as the dashboard it was meant to fix, so archive what you don't use.

Some views are personal and some are shared. The daily check is personal, but the analysis you argue from should be shared, because a shared link is what makes the argument checkable.

In Flowfy, filtering by period, channel, store and campaign is live, so every view described here can be built on demand. Saving a view, naming it and sharing it by link are on the roadmap and haven't shipped. Until they do, the practical substitute is documenting the filter combination somewhere the team can find it, which is worse than a link and considerably better than a screenshot.

The practical step

Watch yourself for two days and write down the filter combination you set every morning. If the same combination comes up twice, document it somewhere the team shares, then see how often anyone opens the tool next week compared to the week before.