Customer Relationship Management (CRM)

GLOSSARY

Definition

A CRM is the system of record for every interaction your company has with a prospect or customer. It stores contact details, conversation history, deal stages, and activity logs. Sales teams use it to manage pipelines. Marketing teams use it to segment audiences. Support teams use it to track tickets.

Why it matters

The CRM category dates back to the 1990s with Siebel Systems, which Oracle later acquired. Salesforce moved it to the cloud in 1999 and made it a subscription. Today the line between CRM and marketing automation is blurry, HubSpot started as a CRM and added marketing, Salesforce started as a CRM and added everything else. The question for buyers is less 'do I need a CRM' and more 'how much of my stack does the CRM eat.'

How it works

A CRM sits between your sales process and your customer data. When a lead enters the funnel, the system records the source, the owner, every call, email, and meeting, and the current stage in the pipeline. Reports then roll that activity up per rep, per team, and per product. The value is not the storage. It is that everyone sees the same record, which stops the classic failure where sales and support each hold their own partial version of the truth.

Practical uses

Mid-market teams typically use a CRM for pipeline forecasting, territory assignment, and renewal tracking. Marketing teams plug the same system into campaign data so a lead score, a form fill, or an abandoned checkout updates the record before the rep ever picks up the phone. Two figures frame the ROI question: a well-run sales process converts a weighted pipeline, and the CRM is what produces the weights. Without it, forecasting is opinion.

How to choose

Compare CRMs on data model, integration depth, and per-seat cost, not on AI feature count. The market splits into three shapes. Salesforce and HubSpot sell platform CRM with ecosystems around them. Freshworks, Pipedrive, and Zoho compete on price and ease of use. Open-source options like Twenty and EspoCRM give you the schema and the server. If your team lives in email and spreadsheets today, a lighter tool beats a platform you will not configure.

The numbers

Self-hosted CRM economics: the software license is free but plan on $20-80 per month for hosting plus migration time measured in weeks. Per-seat SaaS at $14-25 per user per month looks cheap until headcount doubles; a 30-seat team crosses the 5-year point where self-hosting pays for itself somewhere between seats 15 and 25, assuming anyone on staff can run Docker and a backup schedule.

Common mistakes

The most common failure is buying a CRM before defining the sales stages it should track. Teams then record deals as either open or closed, the pipeline report loses all meaning, and the tool gets blamed. The second failure is duplicating records between the CRM and a spreadsheet, which guarantees drift. The third is granting every user full rights, so field hygiene dies within a quarter.

What changed with AI

AI agents in CRM now draft follow-up emails, score leads, and summarize call transcripts. The capability is real, but it depends on the data underneath. Agents inherit every duplicate, every stale stage, and every missing owner in the records they read. Cleaning the schema first is what makes the AI output usable. That is the same discipline that made the CRM useful to humans in the first place.