Business software — which tools a company needs, function by function

Business software is not a single purchase. A company does not "implement a system" once a decade — it buys, rents or builds a separate tool for each function: invoicing, sales, stock, bookings, customer service. The decision is made per function, not per company — and that one rule organises this entire section.

You can see it in our own set-up. We rent email, calendar and analytics, because they work for us exactly as they do for everyone else. We built our CRM, the call-booking module, the project brief builder and the calculators, because that is where our way of working lives. One company, two answers — each of them right for its own function.

This article is a map, not a textbook. It shows what kinds of software for businesses exist, which article in the section answers your problem, the order in which companies usually mature, and how to check whether the spending will pay back. If you are only just setting up the company and choosing your first tools, start with our guide to the technology side of starting a company — here we write for a company that is already running and looking for what to improve.

Business software — what is there?

Business software falls into four layers: accounting and documents, sales and customer relationships (CRM), operations — orders, stock and resources (ERP) — and tools of your own, built around one company's process. Most small businesses have the first layer, some have the second and third, and only a few have the fourth.

Accounting and documents. Invoicing software, bookkeeping done by an accountant or in an app, document flow. This is the layer nobody builds for themselves, because the rules are the same for everyone and change for everyone at once. It is also where e-invoicing is heading across the EU: since April 2025, member states have been able to make structured e-invoices mandatory at home under conditions the EU has set, and from July 2030 cross-border B2B sales move to digital reporting built on the e-invoice. Dates and formats differ from country to country; what they mean for automation is covered in our article on business process automation.

Sales and CRM. Contacts, the history of conversations, quotes, the stage of each deal. As long as one person looks after customers, a spreadsheet will do; when there are more, the conversation about CRM begins.

Operations and ERP. Orders, stock, production, resource planning. ERP (Enterprise Resource Planning) ties these areas together with accounting in one system, so that the order, the stock level and the invoice know about each other.

Tools of your own. A price calculator on the website, a portal for customers, an app for the crew in the field, an integration nobody sells off the shelf. This is where the shelf of ready-made products ends.

The first three layers are now almost always subscription services running in the browser — software as a service, or SaaS. Choosing one is less about features than it looks: most of what decides whether you will be happy in three years is in the contract, the pricing tiers and the way out.

What does this look like in numbers? According to Eurostat, in 2025 ERP software was used by 41.08% of small EU enterprises (10–49 people) and CRM by 24.69%. Two small companies in five have ERP; one in four has CRM. Across all enterprises with at least ten people the shares are higher — 46.45% and 28.51% — because medium-sized and large companies use both systems far more often: ERP reaches 69.93% of medium-sized and 88.71% of large enterprises.

Those numbers say something important about any purchase decision: running a small company without ERP or CRM is the norm, not a backlog. Almost three small companies in five work without ERP of any kind. So before you start thinking about a system built to order, it is worth checking whether what you are missing is simply a tool off the shelf.

Business apps — which problem, which article

This section is not meant to be read from start to finish. Each article answers one situation in which companies end up in a conversation about software. Find yours in the first column.

If more than one situation fits — and in a small company that is common — start with the one that costs the most hours a week, not with the one that sounds most serious. Scattered contacts and retyped data usually share a cause: there is no single place where everything lands. In that case CRM and automation are one project in two steps, not two separate purchases.

A few words on what to expect from each article.

Automation separates three things sold under the same word — integration, RPA and AI — and shows that the first candidate for automation is frequent, boring and based on rules. It also describes our own funnel from the contact form to the conversion report, and the steps we deliberately leave to people.

CRM starts from the point at which a spreadsheet stops being enough — four signals, none of them about the number of customers — and sets out what a small company's system has to do and how to square a customer database with the GDPR and consent for marketing.

ERP system explains how ERP differs from accounting software, from MRP and from CRM, gives the signals that show a small company needs it — stock, production, several sales channels, data retyped by hand — and says when accounting software that sends and receives e-invoices is enough. It breaks the cost into five parts, of which only one appears in the vendor's price list, and shows where implementations most often go wrong.

Online booking says when a free booking calendar is enough, what a paid one has to do, what the price lists do not show, and calculates the point from which a module of your own pays off.

Mobile apps give you the frequency test: an app makes sense when someone comes back often, needs the phone's features and has an account. If not, a well-built website will do the same job for less.

Low-code and no-code explains what these platforms are, what they are best at, and which limits hide in the price lists: the number of records, the price per person, sharing outside the company. And what you take with you when you leave — the data, yes; the logic of the app, no.

Custom software asks four questions that decide the choice, calculates the total cost over five years and describes the middle road: an off-the-shelf tool as the core, with a module of your own where the ready-made one does not reach.

Business management software — one system or several?

Anyone looking for business management software usually pictures a single window with everything in it: customers, orders, stock, invoices, the rota. The market answers this in two ways. The first is an ERP system — one product with modules, one database, one vendor. The second is a set of separate tools, each the best at its own function, connected so that data flows between them on its own.

The choice depends on how tightly your processes are woven together. In a trading or manufacturing company every order changes the stock level and ends in an invoice — there, one database saves the daily job of checking which program is right. A service company has sales, a calendar and invoicing that can be chosen separately and joined up by integration. That is how we work ourselves: our CRM connects to rented email and a rented calendar instead of replacing them.

Do the first test on paper. Draw the path of one typical order from enquiry to payment and count how many times the same data passes from one place to another. Many hand-overs that change something in stock or in the books are an argument for one system; two or three are an argument for connecting what you already have. And one caveat: no business management software organises the work by itself — it moves into the computer whatever is already there, mess included.

How ERP differs from accounting software and CRM, which signals show that a small company needs it, what it costs beyond the price list and where implementations break down, we cover in the article ERP system — what it is and when a small company needs one. If you sell online, the shop joins in and has to talk to stock and accounting — and the order in which you connect them matters more than the choice of any single product.

The order of maturity — from spreadsheet to a system of your own

Digitalisation rarely starts with a system. It starts with a spreadsheet somebody set up because the orders had to be written down somewhere. After that, companies usually climb the same four steps — and each of them is a good place to be until a signal appears that it is time to go up.

Spreadsheet. Cheap, flexible, familiar to everyone. It is enough while one or two people use it and nobody needs to remember the history of changes. The signal to move on: two people edit the same file and overwrite each other's work, or someone copies data from the spreadsheet somewhere else every day.

An off-the-shelf tool on subscription. Company software from the shelf for one function — CRM, invoicing, bookings. It works straight away, and someone else looks after updates and security. Read the contract as carefully as the feature list: pricing tiers, the price per user and the export format decide more than the demo does. The signal to move on: spreadsheets of workarounds grow next to the program, because it does not handle something you do every day.

Low-code. An internal tool assembled on a platform from ready-made blocks — a table, a form, an approval step — by someone on the team, without a developer. Good for prototyping a process and for tools used by a handful of people. The signal to move on: at least two of three — you hit the limit on records or licences, customers rather than only the team are meant to use the app, the logic starts to be the core of the business. Details and the limits in the price lists: low-code and no-code — what they are.

Custom software. A system written for your process and owned by you. It makes sense where your way of working is your advantage and the cost of working around the limits of ready-made tools exceeds the cost of building and maintaining your own. How to calculate that: custom software or off-the-shelf.

Two remarks. First, the steps do not have to be climbed in order across the whole company at once — accounting can stay on subscription for ever, while the price calculator on the website goes straight to the fourth step. Second, a lower step is not a failure. A low-code prototype or a well-kept spreadsheet is the cheapest specification of a future system, should one ever be needed.

Client portal — when a customer should have an account

A client portal is where a customer logs in to your company and finds on their own what they would otherwise have to ask for by email: documents, quotes, the status of a job, the order history. It makes sense when the customer comes back — for an invoice, for a new version of a quote, for news on where their case stands. If most customers buy once and disappear, an account will be an obstacle at checkout, not a convenience.

A good test is simple: how many emails and phone calls a week are the questions "where are we with…" and "could you send … again". Each of them is information the customer could see for themselves if there were somewhere to see it. What a customer portal will not replace is a conversation where the customer needs a decision or advice — its job is to take questions about the state of things off the team, not to take the team away from the customer.

That is how we built the customer account on this site. A customer signs in with a Google, Facebook or Discord account, or — if they prefer — with email and a password or a one-time code. Once signed in, they see in one place:

• quotes saved from the calculators — with the cost range, the monthly cost and the estimated delivery time, so that nobody has to work it all out again a week later;
• projects from the brief — a project description filled in step by step, which can be picked up and finished later;
• their own comments under articles, together with whether they have passed moderation;
• profile settings, including the preferred language.

There is also self-service account deletion: the customer confirms the decision by typing the word "DELETE" (and their password, if they set one) and does not have to write to anyone about it. We see this as part of the portal, not an extra — if opening an account takes a minute, getting rid of it should not require correspondence.

A client portal almost always starts on the website: that is where the customer comes from to log in, and a form, a calculator or an online booking system is the first place they leave their details. That is why it is worth planning it together with the website rather than as a separate project — how a company website connects with the rest of the tools is the subject of our section on websites.

Will it pay off — three questions before every decision

Whichever article on the map is yours, the calculation of whether it pays off starts with the same three questions. It is worth answering them yourselves before you ask anyone for a quote — because a contractor will give you the price of the solution, but does not know the price of your problem.

What does this problem cost today? Counted, not guessed: how many hours a week someone spends retyping, searching and chasing, how many times a month something slips — an enquiry without a reply, an invoice with a mistake, a customer nobody called back. That is the upper limit of a sensible budget. If the problem costs less per year than maintaining the solution that would remove it, the best decision is to do nothing.

Who in the company will own it? Every tool — bought, connected or built — needs a person who makes sure it is used, reports what does not work and decides on changes. Without that person the team drifts back to the spreadsheet within a few months, while the subscription or the server keeps costing money.

What happens if you have to leave? Leave a vendor who raises the price or shuts the product down, or a contractor who stops answering the phone. Can your data be taken out in a form someone else can read, and could someone else take over the code? That is settled when the contract is signed, not later — and it applies as much to a subscription at a few dozen euros a month as to a system built over months.

Once you have an answer to the first question, set it against the total cost of the solution — not the price of implementation, but the sum of implementation, subscriptions and maintenance over several years. How to run that calculation over five years for both options, and where the break-even point lies, we show in the article on custom software.

Technical questions have their own section

This section answers "whether and why". Once the decision is made, questions come from the other side: what it costs to build an application and why quotes differ, how the project runs and what the client's role in it is, what a web application is and how building a phone app differs. They are all collected in the section on web applications.

The order matters. A conversation about technology and price before anyone knows which problem is meant to disappear usually ends with a quote for a solution that looks good in the proposal and fits the company poorly.

The split is deliberate. The same text cannot at once convince an owner that a problem is worth solving and explain how to build the solution. If you are talking to a contractor, read both sections: this one before the first meeting, that one before you sign.

Where to start

Start with one function, not with the whole company. Pick the one that costs the most manual work today and find it on the map above. Before you choose a tool, write down how that function works today: who does what, in what order, where data is retyped and which exceptions only one person knows about. That sheet of paper is the cheapest specification there is — it will help you choose an off-the-shelf program, build a low-code prototype and talk to a contractor.

If you are wondering whether a ready-made subscription service will do, or whether you need something written for you, take the quiz: off-the-shelf SaaS or custom software — the result comes with its reasoning and may point to a solution cheaper than the one you had in mind. If the answer is to build, the web application cost calculator gives you an indicative budget before you speak to a contractor.

And once the decision is made and you are looking for a contractor, the scope of our custom software development is described on the service page.