Custom software, defined.

It matters when your workflow is the part that sets you apart and no packaged product fits it without workarounds. The page on custom software development explains how a build runs.

Enterprise work involves approvals and several people on each account. See custom software development for what such a project covers.

It starts quickly and fits when your process is close to the standard one. The guide on when to buy SaaS lists the cases.

You get hosting and updates, but accept the vendor's features and limits. The guide on buying SaaS or building sets out the trade-off.

It suits simple forms and approvals, but complex logic often meets the platform's limits. The low-code section of the build or buy guide shows where it stops.

Sales, service and account teams stop keeping their own lists, so a customer's history is the same for everyone who talks to them. See sales and customer operations for how these systems are built.

It fits when your sales process differs from the standard pipeline or must connect closely to other systems. The page on sales and customer operations covers what a custom CRM can include.

When these areas share one database, a change in one shows in the others without retyping. Read about back-office operations for how modules are chosen and built.

Small tools often replace a shared spreadsheet or an email chain. The page on back-office operations shows the common kinds.

Back-office systems decide how fast orders move and how reliable the numbers are. See back-office operations for what moves off spreadsheets first.

It helps when decisions depend on numbers that are otherwise spread across exports and spreadsheets. The page on back-office operations explains how dashboards sit on top of your systems.

It cuts routine status questions and gives customers an answer at any hour. See sales and customer operations for how portals connect to the CRM.

Each partner sees only their own records, which keeps your data separate while sparing both sides emails. The page on sales and customer operations covers access rules.

It replaces phone calls and order emails with one record both sides can trust. See sales and customer operations for the ordering flows it supports.

It shortens the back-and-forth in purchasing and leaves a record of who confirmed what. The page on sales and customer operations shows how external logins are handled.

It removes waiting and retyping and leaves people the exceptions. See AI and automation for what can be automated.

It is wider than one task, so it needs a clear picture of who does what. The page on AI and automation walks through the approach.

It can bridge systems that have no other connection, but it breaks when a screen changes. An integration through an API is usually steadier; AI and automation covers connecting systems through their APIs.

It keeps a customer, order or invoice the same everywhere and removes a common source of errors. See AI and automation for how integrations are built.

Most integrations run on APIs, so whether your tools offer one decides how easily they connect. The section on which systems can connect explains what to check.

It is the most common style, so most business tools and developers already work with it. See which systems can connect for how this applies to your tools.

It suits screens that combine data from many places, such as a mobile app or a dashboard. The page on which systems can connect covers API styles in practice.

It keeps connections in one place, so replacing one system does not mean rewiring all the others. The insight on connecting accounting, messaging and your CRM shows an example.

Poor migration leaves new systems with wrong or missing data, so it needs its own plan. See AI and automation for how records move between systems.

It can take over routine multi-step work, as long as its permissions and limits are set deliberately. The page on AI and automation explains where it fits.

Typical uses are sorting incoming requests, qualifying leads and drafting replies for a person to approve. See AI and automation for the tasks that suit an agent.

It handles common questions but does not act in your systems unless connected to them. The section on AI agents and chatbots compares the two.

It keeps judgment calls with people and limits the cost of a wrong answer. The guide on human review of AI agents describes where to place it.

It is a reason to give agents narrow permissions and to check what they do with outside content. See the limits of AI agents for the safeguards.

It reaches phones and computers from one codebase and skips app store review. The section on web, PWA or native helps you choose.

It tests the main idea before the full build and keeps early spending focused on what matters. See MVP or full build for how to decide the first release.

It is where cost, risk and priorities become clear before building starts. The page on the discovery call describes how it begins.

It suits projects whose requirements are clear and stable. The guide on fixed price and time and materials compares it with the alternative.

It suits projects whose scope will change as you learn, and it needs regular reviews to keep spending in view. See fixed price and time and materials for the comparison.

It fits products that keep changing after launch, because the team already knows the code and the business. The section on when a dedicated team fits lists the signs.

A clear one gets comparable answers, because every vendor responds to the same scope. The software project brief template shows what to include.

Software that handles personal data of people covered by the law needs a lawful basis, limits on use and ways to delete or export a person's data. The data protection laws section covers the rules in practice.

It asks for consent, a stated purpose and safeguards, which affects how a system collects and stores customer details. See data protection laws for the practical side.

Buyers ask for it as evidence of how a provider runs its own operations; it is a report about a provider, not a property of a piece of software. The certifications section covers what a provider can and cannot claim.

Software that stores or sends such health information has to follow the HIPAA rules on access, logging and agreements with its providers. See data protection laws for how this is handled.

A DMS holds the dealership's operating records, while a CRM focuses on the customer relationship, and many dealers need both to work together. The section on a dealer CRM or a DMS explains the difference.