How much should a small business spend on EDI?
A retail chain has told you they "work with EDI", and you have no idea what it will cost. Naming a figure is hard, but knowing where the cost comes from makes quotes much easier to compare.
Where does the money go?
There are four items, and their order often surprises people:
| Item | Weight |
|---|---|
| Software / subscription | Usually the smallest item |
| Initial setup and mapping | One-off, but among the largest |
| Extra integration per partner | A new customer means new work |
| Maintenance and support | Ongoing, and routinely underestimated |
The real money is not in the software but in the mapping — defining how incoming EDI fields correspond to fields in your own system. That work partly repeats for every partner, because every partner's specification differs.
Three approaches
1. An EDI service provider (VAN or cloud)
You pay a monthly subscription and the provider handles translation and transmission. Pricing is usually by volume or by number of partners.
- Upside: you start quickly and carry no technical load
- Downside: a permanent cost that grows with volume, plus dependence on the
provider
2. Building your own integration
You add an EDI module to your ERP, or develop one yourself.
- Upside: no monthly fee, full control
- Downside: development time and, more importantly, maintenance. When a
partner changes its specification, you are the one who updates it.
3. Handling it manually
Open the file, read it, key the data in. It sounds primitive, but for a supplier receiving a few orders a month it can genuinely be the right answer.
- Upside: zero setup cost
- Downside: error rate and time cost grow quickly with volume
Where are the thresholds?
As a rough rule of thumb:
- Under 10 documents a month — manual handling is usually cheapest. Automation
will not earn back its setup cost.
- Between 10 and 100 — semi-automatic approaches make sense: read the file,
convert it to a table, import it.
- Over 100, or several partners — full automation pays for itself.
These thresholds shift by industry. In automotive, where one late delivery can stop a production line, automation may be mandatory even at low volume.
The common mistake: committing too early
The most expensive outcome is building costly infrastructure for a single customer and then losing that customer. When your first EDI request arrives, ask: what share of my volume is this partner, and how many more partners will ask for EDI within a year?
If the answer is "one partner, low volume", starting manually or semi-manually is a defensible decision. The door to automation stays open; the reverse is not true.
The costs you do not see
Questions worth asking when comparing quotes:
- What does adding a new trading partner cost? (It is often billed separately)
- Are mapping updates included when a specification changes?
- Is there a test environment? You do not want to experiment in production
- Is the transport channel (AS2, SFTP) charged separately?
- When the contract ends, do your mappings belong to you?
That last point matters more than it looks: with some providers, mapping logic built up over years remains the provider's property, and moving means rewriting it.
Where to start
Before deciding anything, look at the sample file you were sent. How many line items, which segments, how complex is the specification? That is the cheapest way to judge whether the quotes you receive are realistic — and it does not require buying anything yet.