Ask in plain English. Get a chart in seconds, and see exactly how every number was built.
Revenue analytics
Production PostgresSaved results. Every tile shows what it got when it last ran.
Net revenue
Monthly, trailing 12 months
Orders
12,847
+11.2%vs the prior 90 days
Average order value
$333.30
+4.0%vs the prior 90 days
Revenue by channel
Quarter to date
Revenue by region
Quarter to date
Connects to what you already run.
01How it works
Connect a source read-only. Ask in plain English. iDash works out the right question to run, and draws a chart you can keep.
Connected. 42 tables, 318 columns, 27 foreign keys.
Point iDash at Postgres, MySQL, SQL Server, ClickHouse, Trino or a spreadsheet. It checks the login, reads what tables you have, and never writes.
Which plans grew fastest in the last six months?
Resolving plan, month, and net revenue
Growth was strongest on Scale, up 34% since February. Starter was flat.
Type the question the way you would ask a colleague. iDash picks from the measures your team already agreed on, instead of guessing.
Refreshed 4 minutes ago
Every Monday 08:00 to #revenue
Keep any chart on a dashboard, put it on a schedule, and let it land in Slack or an inbox. Each one asks its question again rather than showing an old picture.
Pulling data from two places can quietly double a total. iDash counts each record once, and stops when it cannot be sure.
Three things stop an answer cold: two ways to connect your data that disagree, a table with nothing unique to count by, and a question that would count the same thing twice. You get the reason, not a wrong total.
I cannot answer this safely. Orders and support tickets both repeat per customer, so adding them up in one question would inflate the total. Ask for one at a time.
Verified means iDash built the number from your agreed definitions and can run it again the same way. Unverified means it had to go around them, and says so.
02The product
One surface for the whole loop: ask, read the chart, pin it, then have it delivered on a schedule to the people who need it.
Sessions persist per project and bind to one data source you choose. Follow-up questions inherit the filters you already set.
How many customers ordered more than once in the last 90 days?
2,314 customers placed a second order in the window, 38% of everyone who ordered.
Split that by acquisition channel.
Sales, finance and the board all ask different questions. None of them should get a different number for the same thing.
Ask the way you would ask a colleague. The questions you come back to every week get saved, and after that the answer is waiting for you instead of sitting in someone else's queue.
All of it runs on the same agreed definitions, so a number means the same thing in a chat, on a dashboard, and inside your own product.
What your team uses
What it stands on
03Comparison
Both alternatives break in the same place. The number is right until someone copies it, joins it, or forgets which filter was on.
Getting an answer
Double counting
Freshness
Knowing it is right
Access
