ERP for Small Brands: Do You Really Need One Yet?
The question reaches us almost weekly: do we need an ERP? Usually from brands doing 300 orders a month.
My answer in most of those cases: not for a long while yet. And that has nothing to do with being frugal.
ERP stands for enterprise resource planning. Translated: a system that keeps stock, purchasing, orders, invoices and bookkeeping in one database. The idea behind it is good. The timing decides whether it helps you or paralyses you.
What an ERP actually solves
An ERP solves exactly one problem, and it solves it very well: the question of which number is correct.
When your stock level sits in Shopify, in the fulfilment partner's portal, in a spreadsheet for Amazon and in your buyer's head, you have four truths. Three of them are wrong, and you do not know which. An ERP turns that into one.
The real benefits follow from that:
- One stock level across all channels. No selling goods that left the warehouse long ago.
- Purchasing with lead time. The system knows consumption and delivery time and tells you when to order.
- Batches and traceability. Mandatory with food or cosmetics.
- Documents without double entry. Order, delivery note and invoice come out of the same record.
That is worth a lot. The question is whether you have those problems yet.
The problems small brands really have
When I talk to founders under 1,000 orders a month, the pain points sound different. Support eats the morning. Tracking numbers get copied by hand. Reporting costs two hours every Monday. Returns take too long.
None of those points gets solved by an ERP.
At MUSTAX monthly reporting used to cost 2 days. Today 2 hours. For that we needed a connection between the existing systems, no ERP. At mate order handling was 4 hours a day, today it is 15 minutes. No ERP for that either.
This is the point I have to explain to founders most often. An ERP is a data-keeping system. Most small brands have too many manual steps and no data-keeping problem. Which steps those typically are is in e-commerce process automation.
The threshold: from when it pays off
I work with a rough scale by order volume per month. The boundaries are soft, but they help you place yourself.
| Orders per month | Recommendation | Reason |
|---|---|---|
| under 500 | no ERP | Shopify plus automation is plenty |
| 500 to 2,000 | no ERP, but structure | a clean data flow between the tools |
| 2,000 to 5,000 | assess it | from here on double entry gets expensive |
| over 5,000 | usually yes | stock errors cost more than the system |
Above volume there are four triggers that weigh more than any order count. If one applies, an ERP can be right at 800 orders too:
Several storage locations. As soon as goods sit in two places, stock management becomes a problem of its own.
Several sales channels with real volume. Your own shop plus Amazon plus wholesale. How far you get without an ERP is in multichannel inventory management.
Batches, best-before dates or serial numbers. Legal requirements are not something you negotiate with spreadsheets.
Production or assembly. If you build a product out of parts, you need bills of materials. No Shopify app maps that cleanly.
The intermediate steps hardly anyone mentions
Between a spreadsheet and an ERP sit three stages, and most brands can stay on one of them for years.
Stage 1: Shopify as the centre of the system. Shopify can handle stock, orders and customers. For one channel that lasts a long time. The failure is usually undisciplined data upkeep.
Stage 2: connected tools. Shopify for orders, an inventory tool for stock, accounting software for documents. In between an automation hub like n8n that pushes the data around. That is our stage at all three of our own brands. It costs a fraction of an ERP and covers eighty percent of the benefit.
Stage 3: your own data core. A lean database for exactly the objects that matter to you. Orders, stock, purchases. Cut to fit your business, no off-the-shelf product. Why we often recommend this route is in why we build our own tools.
Stage 2 is the underrated sweet spot. It is reversible, it is built in weeks, and it forces you to understand your processes. Anyone who does introduce an ERP later then has a process map instead of a gut feeling.
What an ERP costs
Licence costs are the smaller part. Rough ranges for small to mid-sized brands in the German-speaking market:
| Item | Order of magnitude |
|---|---|
| Licence, cloud-based | 100 to 600 euros per month |
| Rollout and data migration | 5,000 to 50,000 euros one-off |
| Interfaces to shop and fulfilment partner | 3,000 to 20,000 euros one-off |
| Internal effort in the first year | 2 to 6 months of team time |
| Ongoing support | 200 to 1,500 euros per month |
The most expensive item appears in no quote: your team's time. An ERP rollout ties up the people who know your business best for months. Exactly the people who are then doing no marketing and developing no product.
For comparison: our n8n server costs around 15 euros a month. Building a single workflow takes between a day and a week. Most of our automations paid for themselves after 2 to 3 months. A full breakdown is in what automation really costs.
The counter-argument: when it is too late
I argue a lot here against early ERP rollouts. The reverse mistake is just as expensive, and I have seen it at clients.
A furniture brand came to us with 4,000 orders a month, two warehouses and fourteen connected individual tools. Each of them had been introduced for a sensible reason. Together they made a system nobody could see through any more. The migration into an ERP then took eight months instead of three, because beforehand nobody could define what a valid stock level even was.
The lesson: automation between isolated tools is a stage, and no end state. Treat it as an end state and you build yourself a web that is expensive to untangle later.
Two further arguments for an early ERP I do not want to leave out. If you want to raise outside capital or sell, an auditable system has value in itself. And if your growth is strong and predictable, migrating at 2,000 orders is cheaper than at 8,000.
My recommendation
Short version, by volume:
- Under 500 orders a month: no ERP. Automate support, order handling and reporting. That gives you more time back than any system.
- 500 to 2,000: no ERP, but tidy up the data flows. One place per piece of information, everything else pulls it from there.
- 2,000 to 5,000: sit down and check the four triggers above. If none applies, wait a bit longer.
- Over 5,000 or one trigger applies: plan the rollout. With buffer, and with a person who really knows the process.
And whatever stage you are on: before you introduce a system, describe your processes well enough that an outsider could run them. If you cannot do that, the ERP buys you somebody else's processes and you notice once it goes live.
Conclusion
An ERP is a good tool for a problem most small brands do not have yet. So the question to ask is which problem costs you the most today.
For us it was never the master data. It was the manual steps. That is why we run three brands with 2 people and without an ERP. It is a point about sequence.
If you are unsure which stage you are on: talk to us at Flowhouse. We look at your processes and will also tell you when an ERP is the right next step.