When manual costs more than the software
Order count decides this, not revenue. Here's the arithmetic.
› Adjust the assumptions every default below is sourced
Cost per month against order volume
Two distributors the same size, opposite answers
Everyone frames this as a revenue threshold. It isn't one. What drives the crossover is how many times a day someone types an order, so a low-value, high-volume business hits it long before a high-ticket one of the same size.
Order value cuts both ways, which is why it's in the table. Fewer orders push a high-ticket business away from the line, but each mistake costs more and pulls it back toward one. Each row runs your own assumptions at a different order value.
| Business | Revenue | Order value | Orders / mo | One mistake | Manual | On Odoo | Verdict |
|---|
Why one line bends and the other doesn't
The manual curve
Re-entry labour is a straight line. Every order costs the same few minutes, so doubling orders doubles the hours.
Mistakes are the part that bends. Volume rises, time per order falls, and the error rate itself climbs. Error cost grows faster than volume, and past a point it swamps the labour underneath.
The Odoo band
Licence is charged per person, not per order. Eight people cost the same at 200 orders or 2,000, so the biggest block is flat.
It's a band rather than a line because the plan and billing term move it: Standard on annual billing at the bottom, Custom month-to-month at the top.
The build is a one-off. A step at the start, not a slope.
Where businesses actually do this
Who buys, observed
Under $10M a year, the tier above you isn't aimed at you. That gap is where this sits.
The published walls
Hard ceilings you can check today. A better readiness signal than any revenue figure, because hitting one is a fact rather than a judgement.
The table
| Orders / mo | Re-entry hrs | Labour | Mistakes | Error cost | Manual | On Odoo | Verdict |
|---|
Sources
- Odoo licence cost
- Per user per month, list price: Standard CA$44.00 on annual billing, CA$55.00 month-to-month. Custom CA$69.00 annual, CA$86.00 month-to-month. odoo.com/pricing, checked 4 September 2026.
- Odoo discounts the first 12 months (Standard CA$35.20 annual, Custom CA$55.00). This model runs over 36 months, so it uses the renewal price throughout rather than a rate that expires a third of the way in.
- Implementation
- 120 hours at CA$85 an hour, spread over 36 months. That is Calibre's own rate for ERP implementation work, not a market average, and it sits below what Odoo charges direct. Both the hours and the amortisation are adjustable above.
- Cost of one mistake
- Mispick studies put a single incident at $22–100, most commonly $30–75 once return freight, restocking and the re-ship are counted. Voxware · DMSi
- Those figures are averages across mixed order values, so the model splits them into a fixed handling cost plus a share of the order's value. The split is fitted to reproduce the published band at typical distribution order values rather than measured directly, which makes it the least certain number on this page. Both halves are adjustable above.
- Mistake rate
- Small and mid-sized operations running manual or paper picking typically sit at 95–97% accuracy, against 99.5% and better for barcode-driven ones. Descartes Finale
- Keystroke-level transcription error in controlled clinical settings runs 0.83–2.02%, depending on whether entry is double-keyed. JAMIA, 2019
- QuickBooks Online limits
- Plus: 5 billable users, 40 classes and locations, 250 accounts, 4 custom fields. Advanced: 25 users, unlimited lists. Intuit
- Revenue bands
- The tier definitions placing NetSuite, Acumatica and SYSPRO at $10M–$250M, and the $200.5M median respondent revenue, are from the 2026 ERP Report. Panorama Consulting Group
- Distributor-specific bands. ERP Research
- Deliberately left out
- The widely repeated claim that 53% of firms between $10M and $100M run a formal ERP is attributed to Panorama across dozens of sites but does not appear in the report. Same for the "$28,500 per employee" and "nine hours a week on data entry" figures, which trace back to vendor marketing rather than a survey.
Not sure which side you're on?
We'll walk through what you're running today, where it actually hurts, and what a phased plan would look like. Thirty minutes, free, and sometimes the honest answer is stay on QuickBooks. We'll tell you that too.
Book a fit reviewThe build is costed at Calibre's implementation rate and spread over 36 months. Minutes per order, cost of a mistake and loaded wage are the three inputs worth arguing about, and the crossover is only as honest as those.
This is a decision aid, not a quote. It says which side of the line a business sits on, not what an implementation costs.