See what you are selling, and how well the AI is doing
Order trends, top products, top customers. Plus an honest read on the automation itself: average confidence, how often it matched the product, and how many orders still needed a person.
Why this matters
Order data you could not see before
When orders live in inboxes and chat threads, nobody can answer "what is selling" without exporting something. Once every order is structured, that question takes a glance.
Hold the automation to account
Average confidence and product match rate tell you whether the system deserves the trust you have given it. If the match rate drops, you want to know before your customers do.
The number that matters is how many needed a person
Not how clever the AI is. How many orders your team still had to touch. That is the honest measure of whether any of this is working.
In the product
Everything listed here exists today.
Order trends
Volume over time, by month, with the detail behind it.
Product and customer analytics
Top ordered products and top customers, ranked from real order data.
AI performance
Average confidence, product match rate and how many orders needed a human.
In context
The day at a glance, and the AI held to account
How many orders came in, how many needed a person, and how the automation is performing. The last number is the honest one.
Dashboard
- Orders today
- 148
- ERP success rate
- 93%
- Avg. confidence
- 90%
- Manual review
- 11
+12% vs last week
6 need attention
Match rate 95%
of 148 orders
Where today’s orders came from
- WhatsApp62
- Email51
- File upload21
- Voice14
76% arrived as a message. None of it was typed into a portal.
What needs a person
- 4Below minimumHeld for a decision
- 2ERP errorsRejected by the ERP
- 5No product matchAwaiting review
11 of 148. The other 137 went through without anyone touching them.
Recent orders
| PO | Customer | Source | Entity | Value | Status |
|---|---|---|---|---|---|
| #12012 | Harborline Foods | Meridian | $459.46 | Need attention | |
| #12011 | Northgate Grocers | Meridian | $1,204.00 | In ERP | |
| #12010 | Palewood Catering | Voice | Cordell | $612.25 | Parsing |
| #12009 | Bayside Provisions | Cordell | $880.50 | Error | |
| #12008 | Harborline Foods | File upload | Meridian | $310.00 | Ready for Review |
Visibility
See what you sell, and whether the automation deserves your trust
Two kinds of number: what your customers are buying, and how well ManualOut is doing the job you gave it.
Order and product trends
Volume over time, top products and top customers, ranked from real order data rather than a monthly export nobody quite trusts.
AI performance in the open
Average confidence and product match rate, tracked over time. If the match rate starts to drop, you find out before your customers do.
The number that actually matters
Not how clever the AI is — how many orders your team still had to touch. That is the honest measure of whether any of this is working, and it is shown plainly.
Questions
Analytics, answered
- What can we see in analytics?
- Order volume over time, top products and top customers from real order data, plus the automation's own numbers — average confidence, product match rate, and how many orders needed a person.
- How do we know the AI is doing well?
- The performance view tracks average confidence and match rate over time. The most honest metric is how many orders still required a human touch, and that is shown plainly rather than buried.
- Where does the data come from?
- Every order ManualOut processes, structured into proper line items. Because the orders are read into data rather than left in inboxes, questions like "what is selling this month" are answered from the orders themselves.
See it run on one of your own orders
Send us a real order — the messiest one you can find. We will run it live and show you what happens when it works, and when it does not.