Nearly every ERP conversation opens with the same sentence: “Should we buy something ready, or have it built?”

It sounds like a budget question. It is not. It is a question about which parts of your company are ordinary, and which parts are the reason a customer picks you over the firm down the road. Answer that, and the software question mostly answers itself.

Three routes, not two

People usually frame this as a choice between two things. There are three, and the middle one is where most of the money goes missing.

Take a standard product as it comes. You change how you work to match the software. Cheapest to start, and genuinely right when your processes are ordinary. The cost arrives later, as the things your team stops being able to do.

Customise a standard product. You keep its core and pay to have the gaps filled. Sensible, and often correct — but only while the customisation stays small enough to describe in a page.

Build your own. The software matches how you work. The right answer for the few processes that are genuinely yours, and an expensive way to re-create bookkeeping.

Almost nobody needs one of these for the whole company. Most companies need two of them at once, with a clear line between.

The question that decides it

Take your processes one at a time and ask a single question about each: if we changed this to match the software, would we be a worse business?

Double-entry bookkeeping: no. The rules are not yours, they are everybody’s. Payroll: no, the rules belong to the state. Stock movements in and out of a warehouse: rarely.

How you price a job, how you decide which truck takes which load, how you match a supplier’s printed line to your own catalogue, what a customer sees when they log in: often yes. That is usually where the company’s actual advantage lives, and it is the part a standard product flattens.

Which side a process belongs on
The processHow different is yours?Where it belongs
Bookkeeping and the ledger Not at allStandard software
Payroll and tax Barely — the rules are the state’sStandard software
Stock movements RarelyStandard software
Pricing and quoting Often veryYours
Planning and assigning work Often veryYours
What the customer sees Almost alwaysYours

Where the middle route goes wrong

Customising a standard ERP is the sensible-sounding answer, and much of the time it is the correct one. It fails in one specific way, and it is worth naming because it happens quietly.

The customisation grows. Pricing does not fit, so it is customised. Dispatch does not fit, so it is customised. The customer portal does not exist, so it is built. Two years in, there is a system of real size living inside somebody else’s product.

At that point you have paid for a custom build, you are still carrying the standard platform’s constraints, and you can no longer take its updates without paying to redo your own changes.

Two ways to end up with a working system

Everything inside one product

One system · cannot move

  1. The standard modules go in
  2. Pricing does not fit, so it is customised
  3. Planning does not fit, so it is customised
  4. The portal does not exist, so it is built
  5. An update arrives and the changes must be redone
  6. Every change now needs the vendor

Standard core, your own layer

Two systems · one border

  1. The standard product keeps bookkeeping and payroll
  2. Pricing and planning live in your own system
  3. The two exchange documents automatically

The dashed steps are the ones that make the next upgrade more expensive than the last.

The spreadsheet beside the ERP

If your company already runs an ERP, there is a faster test than any evaluation.

Walk through the office and look at what is open on the screens next to it. Every spreadsheet you find beside the ERP is a process the ERP does not handle. That spreadsheet is not a bad habit. It is a specification, written by the people who do the work, for free.

Count them. If there are one or two, the system fits and the gaps are worth filling. If every desk has one, the problem is not missing features — the product does not match the company.

You probably do not have to replace your accounting

This is the fear that stalls most of these projects, and it is usually unfounded.

An accounting system can stay exactly where it is as the book of record for accounting and tax, while a second system runs sales, planning, approvals and portals. The two exchange documents and payments automatically, and nobody retypes anything.

Replacing accounting is a large, risky project with very little upside. It is almost never the first thing to do, and it is often not worth doing at all.

Deciding, in about a week

This does not need a long evaluation. It needs one page and two conversations.

  • List your processes on one page. Not modules — processes, in the words your team uses.
  • Mark each one ordinary or ours. Be honest. Most companies have three or four that are genuinely theirs, not fifteen.
  • Write one sentence for each “ours” process saying what makes it different. If you cannot write the sentence, the process is ordinary and you have just saved yourself a lot of money.
  • Ask two vendors to show those processes only — on your examples, not on their demo data.
  • Count the workarounds in what they show you. That number is your answer.

If the count is near zero, buy the standard product and stop. If the workarounds pile up on the same two or three processes, that is the line: standard for everything else, your own system for those.

Which of the three routes fits your company is usually clear after one conversation about how the work actually moves. If the answer is that you need no new system this year, we say that too.

Projects behind this article

What we build in this area

  • ERP & business management systems

    ERP for trade, distribution and service businesses — sales, stock, finance, payroll and approvals, connected to 1C and e-qaimə.

  • SaaS platform development

    Multi-tenant SaaS products engineered end to end — tenancy, billing, mobile PWAs, AI features and the console to run it all.

All articles