It worked Illustration
Weekend promo
Started at 3 of 8 branches
After the start date, the branches with the promo rose more than similar branches without it. The gap is bigger than their normal ups and downs.
Early access · for businesses with several branches
Did that change actually work?
Log what you changed, and at which branches. OnCertain compares those branches with the ones you did not change. Then it tells you plainly: it worked, no clear change, or too early to tell.
The demo is a made-up group of eight clinics with sample data.
How it works
Three steps, using numbers you already have
-
Log the change
Write what you changed, the date it started, and the branches it touched. A title, a date and the branches are enough to start.
-
We compare it with untouched branches
OnCertain looks at your daily numbers before and after that date. It compares the branches that got the change with similar branches that did not. Things that hit every branch, like a holiday or a slow month, cancel out.
-
You get a verdict, and the team keeps the lesson
You get a plain answer and the reason behind it. When you close a change, the result goes into a searchable learnings bank. The team can see what was already tried, and new staff can too.
What you need: daily numbers for each branch, such as visits, sales or bookings, in a CSV file. One row per day, per branch. No customer names or contact details.
The method: difference-in-differences. It is a standard way to measure a change when you cannot run a lab test. OnCertain also checks how much your branches move on normal days, so it does not mistake noise for a result.
Honest verdicts
Three possible answers. One of them is "we can't know yet".
A report can always show a number, even when the number means nothing. OnCertain only says a change worked when the difference is bigger than the normal ups and downs of your branches. When it cannot tell, it says so, and it tells you why.
It worked Illustration
New booking script
Front desk, some branches
The branches with the change moved more than similar branches did. The gap is larger than normal noise.
No clear change Illustration
Longer opening hours
Weekdays, some branches
The branches with the change moved about the same as the others. If there was an effect, it was too small to see.
Can't tell yet Illustration
New price list
Some branches
There is not enough evidence either way yet. OnCertain will not guess.
Why: Too few similar branches to compare with, and the daily numbers jump around a lot.
Other reasons for "can't tell yet" include: not enough days since the change, or branches that were already moving apart before it. If a change made things worse, OnCertain says that too.
Who it's for
Teams that run more than one location
If you sometimes change things at some branches and not at others, you can use OnCertain. The branches you did not change are your comparison.
Clinics
Example:A new booking script at the front desk of some clinics.
Retail chains
Example:A new shelf layout in some stores.
Food and beverage
Example:A lunch promo at some outlets.
Service businesses
Example:A new follow-up call routine at some locations.
What it is not
A narrow tool, on purpose
Not A/B testing for apps
OnCertain is for changes in the real world: at the counter, on the shelf, in the clinic. It does not test buttons or screens in an app or website.
Not a BI dashboard
It does not try to show every number you have. It answers one question for each change: did it make a difference?
No customer-level data
It uses daily totals per branch only. If a file has names, phone numbers, emails or ID numbers, OnCertain refuses the whole file.
Early access
Request early access
We are opening OnCertain to a small group of businesses first. Tell us a little about yours, and we will reply by email.
Want to look first? The live demo shows the whole app with sample data.