"We need a CRM" is a sentence that can mean five different things depending on who's saying it. A sales team means a pipeline. A service business means a scheduling and job-tracking system with a contact list attached. A clinic means something closer to patient records with a CRM layered on top. Before scoping a build, it's worth being specific about what a CRM actually needs to do for your business — here's the checklist I use.
The Baseline: What Every CRM Needs
Regardless of industry, these are the features that separate a real CRM from a glorified address book:
- A real lead/contact record — not just a name and email, but a full history: every interaction, note, and status change in one place.
- A pipeline or stage system — so "where does this lead actually stand" is a glance, not a question someone has to answer from memory.
- Follow-up and reminder tracking — the single feature most responsible for "we lost that lead because nobody followed up," which a spreadsheet structurally can't solve.
- Activity logging — a real audit trail of who did what, which matters the moment more than one person touches a record.
- Search and filtering that actually works at the volume you'll realistically have in a year, not just today.
The Layer Most Off-the-Shelf Tools Get Wrong
Generic CRMs model a generic sales process. The features that make a custom CRM worth building are almost always the ones specific to how your business actually operates:
Role-Based Access That Matches Your Real Org Chart
Not every feature should be visible to every user. A sales rep usually shouldn't see company-wide revenue; a support team usually shouldn't be able to delete records. Real CRM software enforces this server-side, not just by hiding a button in the UI — which matters a great deal more than it sounds like it should, the first time someone tries the API directly.
Commission and Revenue Tracking, If That Applies
For any team working on commission, the question "what have I actually earned this month" needs to be computed live from real data, not reconciled by hand against a separate spreadsheet that can silently drift out of sync with what the CRM itself shows.
Industry-Specific Workflow
This is the feature category that varies the most, and the one generic tools handle the worst. A medspa needs treatment history and scheduling. A logistics business needs dispatch and multi-drop delivery tracking. A retail operation needs inventory tied to every sale. The right question isn't "does this CRM have a feature for that" — it's "was this CRM actually built around how this specific kind of business runs."
Automation That Removes Real Manual Work
The useful kind of automation isn't flashy — it's a status change that triggers a notification, a follow-up that gets auto-scheduled, a report that generates itself instead of getting assembled by hand every month. Each one is small; together they're the difference between a system people actually use and one they route around.
AI, Scoped to One Real Problem
AI in a CRM is worth building when it solves one specific, named problem — scoring which leads are worth calling first, drafting a follow-up email from a conversation history, summarizing a long client record before a call. It's not worth building as a vague "smart" label with no concrete feature behind it.
Two Real Examples, Two Very Different Feature Sets
ADEPT MedSpa OS and MT Logistics Group CRM both started from the same baseline — leads, contacts, a pipeline — and then diverged completely based on what each business actually does. ADEPT's CRM layer feeds clinical charting, an AI-assisted injection mapper, and a branded patient app. MT Logistics' CRM layer feeds a multi-drop dispatch board, a driver app, and live customer tracking. Same baseline, genuinely different system once you get past the contact list — which is exactly why "what features does a CRM need" doesn't have one universal answer.
Frequently Asked Questions
- What's the minimum a CRM needs to actually be useful?
- A real contact/lead record, a pipeline or stage system so you can see where everyone stands, and some form of reminders or follow-up tracking. Everything past that — reporting, role-based access, automation — is valuable but not the baseline that makes it better than a spreadsheet.
- Should a CRM include AI features by default?
- Only if there's a specific, concrete use for it — lead scoring, drafting follow-up emails, summarizing a long client history. AI bolted on as a marketing checkbox with no real feature behind it isn't worth the added complexity; AI solving one named problem usually is.
- Do I need role-based permissions if it's just my team of three?
- Probably not urgently at three people, but it's worth building in from the start if you're planning to grow, since retrofitting permissions onto a system that assumed everyone sees everything is a real rebuild, not a quick toggle.
- What's the difference between a CRM and a full operations platform?
- A CRM tracks leads, customers, and the relationship. An operations platform adds whatever actually runs the business on top — dispatch, scheduling, inventory, compliance, depending on the industry. Most businesses start by needing a CRM and discover the operations layer afterward, once the CRM part is solid.
The Bottom Line
The baseline features are the same for everyone. The features that actually decide whether the CRM feels built for your business, or like your business is being forced into someone else's template, are the ones specific to how you actually work. That's the part worth scoping carefully — and the part a generic tool almost never gets right. If you're still deciding whether you need one at all, I wrote a more basic checklist on the signs your business has outgrown a spreadsheet, and a full breakdown of what custom CRM development actually costs.