A customer order at a mid-sized manufacturer passes through half a dozen systems before it is paid: the CRM where the deal was won, the ERP that books the order and the invoice, the supply chain system that plans the parts, the HR system that knows who is on shift, and the dashboard where finance watches the margin. That set of systems is what people mean by enterprise software. Most of it is now bought as a subscription from a handful of very large vendors, and the parts worth building or configuring yourself are the ones where your business differs from everyone else’s: the industry core, and the integrations that hold the rest together.
What is enterprise software?
Enterprise software, also called enterprise application software (EAS), is software an organization runs to manage its own operations, used by many people in different roles over shared data. Payroll, the general ledger, the customer database, the warehouse stock count and the service desk queue are all enterprise software. A spreadsheet that one analyst keeps on a laptop is not, even when the company depends on it.
Four traits set it apart from personal or consumer software:
- It is a system of record: the place where a fact about the business (an invoice, an employment contract, an insurance policy) officially lives, so other systems copy from it instead of keeping their own version.
- Many people work in it at once, with roles and permissions that decide who can see and change what.
- It keeps an audit trail, because auditors, regulators and tax authorities ask who changed a record and when.
- It connects to other systems through APIs, files or message queues. An ERP that cannot receive orders from the CRM becomes a second place to type them in.
Spending on it keeps growing. Gartner’s July 2026 forecast puts worldwide software spending at $1.47 trillion in 2026, up 15.5% on 2025, out of total IT spending of $6.37 trillion (Gartner). It is a forecast, and it counts all software, not only the business applications in this guide.
Why companies run enterprise application software
Few companies adopt a new enterprise system because of a strategy document. They adopt it after a specific failure: a month-end close that takes three weeks because figures are retyped between files, a customer who gets two different answers from sales and support, an audit question about who approved a payment that nobody can answer.
Enterprise application software fixes those problems in the same way. It keeps one record per fact, so finance, sales and support read the same customer and the same invoice. It puts the process (an approval chain, a stock reorder rule, a leave policy) into the system instead of into people’s heads, and it records who did what.
The costs are just as concrete. A packaged system brings its own data model and its own idea of how a process should run, licences are charged per user or per module every year, and each new system adds integrations that someone has to keep working. Those trade-offs decide which systems to buy and which to build, and we come back to them after the twelve types.
12 types of enterprise software systems
The twelve types below cover most of what a mid-sized or large company runs. The boundaries blur in practice: Microsoft sells ERP and CRM applications under one brand, Dynamics 365, and most ERP suites include some HR and supply chain functions. The matrix gives the short version, grouped by the job each type does; the sections after it follow the same order.
ERP: enterprise resource planning
An ERP system is the financial and operational backbone of a company: general ledger, payables and receivables, sales orders, inventory, purchasing and, in manufacturing, production planning. Every transaction that touches money ends up in it, which is why the ERP is usually the system of record the others reconcile against.
The best-known suites are SAP S/4HANA, Oracle Fusion Cloud ERP (with NetSuite for mid-sized companies) and Microsoft Dynamics 365 Finance and Business Central. Replacing an ERP is often the largest software project a company runs in a decade, which is why the end of mainstream maintenance for SAP’s older ERP, covered in the deadlines section below, affects so many companies at once.
HCM: HR and payroll software
Human capital management (HCM) software follows an employee from hiring to leaving: contracts, the organization chart, leave, performance reviews, training and, in most suites, payroll. Payroll used to be a category of its own, and many companies still run it separately or hand it to a payroll provider, because the rules change with each country’s tax and labour law.
Workday HCM, SAP SuccessFactors and Oracle Fusion Cloud HCM are typical suites for large employers. An HCM system holds some of the most sensitive personal data a company keeps, so access rules and the audit trail matter as much as the feature list.
SCM: supply chain management
Supply chain software plans and tracks the flow of goods: demand forecasts, supplier orders, warehouse stock, transport and delivery. The ERP records the transactions; dedicated SCM tools such as Blue Yonder, Kinaxis or Oracle Fusion Cloud SCM do the planning, deciding how much to order, from whom and where to hold it. Recalculating that plan across thousands of products, sites and suppliers is a heavy computation, and how often it can run sets how quickly the business can react to a late shipment.
EAM: enterprise asset management
EAM software looks after physical assets over their whole life: machines, vehicles, pipelines, buildings. It holds the asset register, schedules inspections and preventive maintenance, sends work orders to technicians and records what each repair cost, so a company can tell when an asset has become cheaper to replace than to keep fixing.
Utilities, transport, manufacturing and the public sector are the heaviest users. IBM Maximo and IFS are typical products, and the large ERP suites have asset modules of their own.
CRM: customer relationship management
A CRM holds the commercial side of the customer: leads, opportunities, accounts, contacts, quotes and, in its service modules, support cases. Sales managers forecast from it, and a service agent who opens a case sees the customer’s history on one screen. Salesforce Sales Cloud and Service Cloud, Microsoft Dynamics 365 Sales and HubSpot are common choices.
Most of the trouble sits at the boundary between CRM and ERP: which of the two owns the customer’s billing address, and what happens to an order when they disagree.
CMS: content management systems
A content management system lets editors publish and update web pages, documents and media without a developer for every change. Enterprise products such as Adobe Experience Manager and Drupal add approval workflows, several languages and brands on one installation, and personalization.
A headless CMS stores the content and serves it over an API, so the same product description can appear on the website, in the mobile app and in a partner’s shop without being copied by hand.
Marketing automation and email
Marketing automation tools send email campaigns and track how people respond: segments, journeys triggered by what a contact does, landing pages, and lead scores that pass the most interested prospects to sales. Salesforce Marketing Cloud and HubSpot are typical. The part that needs the most care is consent. Under the GDPR in the EU, the system has to know, for each contact, what the person agreed to and when, and stop sending the moment that changes.
Collaboration and messaging software
Team messaging has replaced much of internal email: a channel per team or project, direct messages, calls, shared files, and alerts from other systems posted where people already talk. Slack and Microsoft Teams are the two names most companies choose between.
Older lists of enterprise software companies still name Slack on its own, but Slack has been part of Salesforce since July 2021 (Salesforce). What a large company needs from messaging on top of chat is retention rules, legal hold, single sign-on, and integrations that post a failed deployment or an approval request in the right channel.
Project and work management
Project management software breaks work into tasks with owners and dates and shows progress against a plan. Software teams track issues and releases in Jira and write specifications in Confluence; business teams often use lighter tools such as Asana. The value is one shared view of who is doing what, which replaces a good share of status meetings and status spreadsheets.
BI and analytics
Business intelligence tools turn data from all the other systems into reports and dashboards. A BI tool rarely owns data of its own: it reads from a data warehouse fed by the ERP, the CRM and the rest, so the quality of those feeds limits what any dashboard can show. Microsoft Power BI, Tableau (part of Salesforce since 2019), Google’s Looker and Qlik are the usual candidates.
Integration and application platforms
Not every need fits a packaged product, and not every gap justifies a software project. Two kinds of platform fill the space between.
Integration platforms connect the other systems. MuleSoft, part of Salesforce since 2018, and similar products move data between applications through APIs and ready-made connectors; in Java systems, Apache Camel or Spring Integration do the same job in code. We spent several years as a Salesforce and MuleSoft development partner, including a RAML plugin built for MuleSoft’s own Anypoint Studio, and for a global beverage company we designed an integration architecture that linked regional on-premise systems to cloud applications without rewriting systems already in maintenance (Salesforce and MuleSoft case study, integration architecture case study). The rule that work taught us: decide which system owns each fact before connecting anything.
Low-code application platforms such as Microsoft Power Apps and OutSystems let a company assemble internal applications (an approval form, an inspection app for field staff, a small case tracker) from visual components and a data model, with code only where the components stop. They suit departmental tools well. Systems with heavy business logic, high volumes or strict audit requirements tend to reach the limits of such a platform, and those usually end up as configured vertical products or custom code.
Industry-specific (vertical) software
Everything above is horizontal software: it works the same way in a bank, a retailer or a factory. Vertical software is built around one industry’s central record: the policy in insurance, the account in banking, the patient record in healthcare, the booking in travel. Guidewire and Duck Creek in property and casualty insurance, Temenos in banking and Epic in healthcare are well-known examples. For many companies the vertical core is the most important system they run, because it is where their product actually lives.
The insurance example we know best is Openkoda, an insurance policy administration system (PAS) that insurers configure with AI, from quote to claim: rating, underwriting, policy administration, claims, billing, workflow automation, document generation and bordereaux reporting for delegated authority. It is built and licensed by our sister company of the same name, as a subscription, and runs fully managed or on premise.
Changes to a product are made through configuration rather than code. With the AI Product Builder, a product manager describes a new product or a change in plain English; the AI drafts a reviewable change-set of coverages, pricing, rules and workflow from a fixed catalogue of operations, a person approves it, and each change is versioned and recorded in the audit trail. Policies already in force stay on the product version they were bound under.
Stratoflow is Openkoda’s implementation partner: we configure products, integrate the platform with the insurer’s existing core and finance systems, migrate data, and build custom modules in Java and Spring. Core Travel Insurance went from contract to a live platform in eight weeks in 2026, and SkyGuard runs life and aviation policy administration on the same platform (case study). Openkoda is an insurance product, not a general business suite, which is why it appears here and not next to Microsoft or SAP below.
Top enterprise software companies in 2026
Six companies cover most of the horizontal categories above, and each started from a different place: Oracle from the database, SAP from ERP, Salesforce from CRM, Workday from HR, Atlassian from tools for software teams, and Microsoft from the desktop. Revenue gives a sense of their size, with two caveats. The figures are total company revenue, not enterprise applications alone (Microsoft’s includes Windows, Xbox and the Azure cloud), and the fiscal years end in different months.
Microsoft
Microsoft’s business software runs from Microsoft 365 (Word, Excel, Outlook and Teams) through Dynamics 365 for ERP and CRM and the Power Platform (Power BI and Power Apps) to Azure, the cloud many of these systems run on. Revenue for the fiscal year that ended on 30 June 2026 was $331.8 billion (Microsoft). Dynamics 365 now ships with prebuilt AI agents for sales, service and finance work, such as an account reconciliation agent for accounting teams (Dynamics 365).
Oracle
Oracle started with its database and now sells Fusion Cloud Applications (ERP, HCM, supply chain and customer experience), NetSuite for mid-sized companies, and Oracle Cloud Infrastructure. Total revenue for fiscal 2026, which ended on 31 May 2026, was $67.4 billion; cloud revenue made up $34.0 billion of it, up 39%, and cloud applications $15.9 billion (Oracle).
SAP
SAP, the German ERP vendor, sells S/4HANA in public and private cloud editions, along with SuccessFactors for HR, Ariba for procurement and Concur for travel and expenses. Total revenue for 2025 was €36.8 billion, of which cloud revenue was €21.0 billion (SAP). Many SAP customers still run the older SAP ERP 6.0, usually called ECC, and its maintenance dates are among the deadlines below.
Salesforce
Salesforce built its business on CRM delivered as a cloud service, and much of its range came by acquisition: MuleSoft for integration in 2018, Tableau for analytics in 2019, Slack for messaging in 2021 (Salesforce acquisitions) and Informatica for data management in November 2025 (Salesforce). Its current push is Agentforce, AI agents that work inside the CRM. Revenue for fiscal 2026, which ended on 31 January 2026, was $41.5 billion, including $399 million from Informatica (Salesforce).
Workday
Workday sells HR and finance software on one cloud platform: HCM, financial management, planning, spend management and contract management, among others (Workday). Large employers often pick it when they want HR and finance in one suite rather than an ERP with an HR module. Revenue for fiscal 2026, which ended on 31 January 2026, was $9.552 billion, up 13.1% (Workday).
Atlassian
Atlassian makes tools that many software and IT teams plan and document their work in: Jira, Confluence, Jira Service Management, Bitbucket, Trello and Loom, with Rovo as the AI layer across them (Atlassian). Revenue for the fiscal year that ended on 30 June 2026 was $6.572 billion, up 26% (Atlassian). Atlassian is moving to the cloud only, and its self-hosted Data Center products are being retired (dates below).
Buy, configure or build: matching each system to the job
Where every company does the same thing (general ledger, payroll, messaging, email campaigns), buying the product is almost always the right call. The vendor spreads the cost of tax rule updates, security fixes and new features across thousands of customers, and a company that builds its own payroll pays for all of that alone. When a client asks us to build something a product already does well, we say so.
Where the company competes, the answer changes. An insurer’s products, a lender’s ledger and a travel company’s search and pricing are the business itself, and a horizontal suite either cannot model them or forces them into its own shape. There the choice is between an industry platform configured for the company and a custom system. Legerity took the second path for FastPost, the rule-driven accounting platform we developed as its main engineering team, which processes one billion financial transactions in under an hour (FastPost case study).
| Situation | Usual answer | Example |
|---|---|---|
| The process is the same in every company: ledger, payroll, messaging | Buy a product, usually as SaaS | Workday for HR, Slack for messaging |
| The process is specific to your industry, and a vendor models it | Configure an industry platform | Openkoda for insurance policy administration |
| The process is how you compete, or the volumes are beyond what products handle | Build it, next to the packaged systems | FastPost, a billion transactions in under an hour |
| Two systems need the same data | Integrate, with one owner for each fact | MuleSoft, Apache Camel or Spring Integration |
Whatever the mix, the systems have to exchange data, and a large share of the effort in any enterprise software project goes there. Four questions settle most integration designs:
- Which system is the record for each fact: customer, product, price, invoice?
- Does a change have to arrive within seconds, or is a nightly file enough?
- What happens when the other system is down: retry, queue the message, or stop and alert?
- Who is told when a message fails, and who corrects the data?
Enterprise software support deadlines to plan for
Enterprise software is rarely switched off by choice. It is replaced or upgraded when support ends, and several support dates in the next few years land on many companies at once.
| System | Date | What happens |
|---|---|---|
| Spring Boot 3.x | 30 June 2026 (passed) | Open-source support ended for every 3.x branch (Java and Spring support calendar) |
| SAP Business Suite 7, including SAP ERP 6.0 (ECC) | End of 2027 | Mainstream maintenance ends; optional extended maintenance runs to the end of 2030 at a 2% premium (SAP) |
| Atlassian Data Center: Jira, Jira Service Management, Confluence, Bamboo, Crowd | 30 March 2028, then 28 March 2029 | Last date for existing customers to buy, then end of life; Bitbucket Data Center is the exception (Atlassian) |
| Oracle Java SE 8 | December 2030 | Oracle’s extended support ends (Java support calendar) |
For the packaged products, the work is a migration with the vendor’s tools and partners. For the custom Java systems around them, which carry the integrations and the logic specific to the business, it is upgrade work: current Java and Spring versions, done step by step and without a risky rewrite. That is the work we do. Our guide to legacy application modernization explains how we order it across a portfolio, and the strangler fig pattern shows how a system is replaced one slice at a time when an upgrade is not enough.
If you run custom Java systems alongside packaged enterprise software and want to know what bringing them up to date would take, a scoping call with us ends with a written ballpark and the assumptions behind it.