Why does Intelligence show nothing to act on?

Intelligence would rather say nothing than tell you something that is not true. Every possible finding has to pass a series of tests before it reaches your Actions list, and most do not. An empty list, or a short one, is normally the system working as intended.

Before you start

Make sure the data a check needs is connected and syncing. Open Intelligence, Data trust and look at Sync health: a platform that is not syncing gives the checks nothing to test, which also produces silence.

The tests a finding must pass

Each check works through the same questions. If the answer to any of them is no, the check abstains and nothing is shown.

1. Is there enough data? Each check has a minimum: enough orders, sessions, ad spend or survey answers in the period and in the period it is compared with. Below it, the check reports Not enough data to test.
2. Is the method suitable? Some comparisons only work when the figure was steady before the change being tested. If it was already moving, the check reports Method not eligible rather than blaming the wrong thing.
3. Is the change bigger than normal noise? Intelligence compares the period with the one before (or last year) and runs a statistical test to rule out chance.
4. Is it clearly different from no change? The range of likely values has to sit fully above or below zero. If the range includes zero, the check reports Interval spans zero.
5. Does it survive a placebo test? The same test is run on a period where nothing should have happened. If it finds a "change" there too, the result is not trusted: Failed the placebo window.
6. Are many tests producing lucky results? Because hundreds of checks run at once, a few would pass by luck alone. Intelligence corrects for this, so only results strong enough to stand out from the crowd are kept.
7. Is it big enough to matter? A real but tiny change is not worth anyone's time. Each check has a materiality threshold in money or share, and a result below it reports No material change.

Only findings that clear every step appear on Actions, ranked by the money at stake.

Step by step: check why the list is empty

1. Open Intelligence and look at the period control in the header. A longer period gives the checks more to work with.
2. Open Data trust.
3. Read the line at the top: Checked at a time for a period. If it says some checks did not run, a source could not be read.
4. Look at Sync health for any platform marked Not synced yet or asking you to reconnect.
5. Read What the evidence engine found to see how many findings were tested and the reasons any were held back.
6. On Actions, if My teams is on, choose Show every team in the empty state to see findings for teams you are not in.

What you will see

• "Nothing to act on" with "No revenue or profit issue crossed its threshold this period.": nothing passed.
• "Nothing for your teams to act on": there may be findings, but none for the teams you belong to. Choose Show every team.
• "Nothing for [team] to act on": you have filtered to one team. Clear the Team filter.

Troubleshooting

• The list was full yesterday and empty today. Check Data trust. If the last run was complete and the conditions have cleared, the findings closed. If a source failed, findings stay open rather than closing.
• You think a real problem is being missed. Email support@fabrik.ae with the brand, the period and what you expected to see. We will check which test held it back.