The honest case for Notion as a CRM
Start by conceding what the purists will not. Notion as a CRM works, for real, for a meaningful number of businesses, and telling those founders they need Salesforce is how consultants earn a reputation for selling complexity. A database with contacts, a board view with stages, a relation to companies, a few properties, and a template — that is a CRM. It took an hour. It is free of a migration. And critically, it is a CRM the founder will actually open, which is more than you can say for the sophisticated one they bought and abandoned last year.
It has genuine advantages that get dismissed too fast. The record is a page, which means it holds the actual texture of a relationship — the notes, the proposal, the odd context, the thing they said about their co-founder — in a form a human reads rather than in a Notes field with a scrollbar. Real CRMs are strangely bad at this. The schema is yours in thirty seconds; adding a field to a real CRM is a change management event, and in Notion it is a click. And it is where your other knowledge already lives, so the deal sits next to the SOP that will run when it closes.
So who should stay? A business with a small number of high-value relationships, a sales process that is genuinely conversational, no volume of inbound, and where the founder is the entire go-to-market. Consultancies. Agencies with four clients. Anyone whose deal flow arrives from their own network. If you are moving eight to twelve deals a year at high value and every one is a relationship you would recognize on the street, a real CRM is overhead in a costume, and you have more expensive problems than this.
The mistake is not using Notion as a CRM. The mistake is not noticing the day it stopped working — because that day does not announce itself.
The four things it structurally cannot do
No timers that fire when nobody is looking. This is the fatal one and everything else is negotiable next to it. A CRM's most valuable object is the no-touch timer: every deal has a last-meaningful-touch timestamp, every stage has a threshold, and when it trips something interrupts a named human on the day. Notion can display that a deal is old. It cannot make anything happen about it. The distinction is total. Momentum fades the same way it does in relationships — not through rejection, but through neglect — and a system that can only show you neglect after you go looking is a system that lets you neglect.
No identity or dedupe. A CRM has an opinion about what a contact is. It knows that this email address is this human, refuses to create a second one, and merges when you did. Notion will happily hold the same person three times because someone typed their name differently, and now your board has three deals that are one deal, and the day you find out is the day two people from your company email them.
No pipeline state machine. In Notion a stage is a select property — a label a human drags. There is no entry criterion, no exit event enforced by anything, no transition that fires anything, no history of when the transition happened unless someone thought to log it. Which means your stages become descriptions of your feelings, and a stage with no exit event is where deals go to die politely while inflating the forecast you are using to decide whether to hire.
No deliverability. Sending is not a feature you add to Notion. Which means your emails live in your inbox, your texts live in your phone, and the conversation — which in a service business *is* the deal — exists nowhere in the record. You are the reconciliation layer, which is the exact problem the CRM was supposed to solve.
The failure that actually gets people, and why they miss it
Here is how it ends, and it is never dramatic. Nothing breaks. No error, no outage, no lost data. What happens is that your business gets slightly busier — one big project, one bad month, one new hire — and the board stops getting updated, and because updating it is the only mechanism it has, the board silently becomes fiction. Forty deals, half untouched since spring, and a founder who cannot tell which of them are alive.
The reason this specific failure is invisible is worth sitting with, because it explains why smart people stay too long. A deal you neglected produces no wreckage. There is no complaint, no refund, no bad review, no angry email. It just quietly does not become anything, and next quarter you conclude that marketing is not working and spend more money at the top of a funnel that is leaking at the second inch. Everything else in your business tells you when it is broken. A neglected pipeline does not, and Notion has no capacity to be the exception.
The second miss: founders benchmark on volume. They think 'I only have twenty deals, I do not need a real CRM,' and volume is the wrong axis entirely. Twenty deals where every action is you deciding to act during a workday is fine forever. Three deals where one of them needed a follow-up on a Thursday you were in delivery all day is not fine, and it is not fine at any volume. The question is never how many. It is whether the system needs to act without you.
And the third: the board looks great right up until it doesn't. A Notion pipeline is accurate immediately after someone updates it and degrades continuously from that moment. It is a system that lies with a confident interface, and it lies most convincingly during the exact weeks you are too busy to check — which are the weeks the lie costs the most.
What to do about it, in both directions
If you are staying, stay deliberately and buy the one thing you are missing. Since Notion cannot fire a timer, the timer has to live somewhere that can: a scheduled job in Make or n8n that reads your Notion database on a cadence, finds records past the threshold, and posts into Slack tagging a named human with the deal and a drafted next message. That is an afternoon of work and it converts Notion from a board into something that will interrupt you. It does not fix dedupe, state or deliverability — but those are limitations you can see, and this was the one you could not.
Also, if you are staying: give the stages exit events anyway, written in the page. Notion cannot enforce them, but you can, and a stage whose exit event is written down is one a person can argue with. And put a last-touch date property on every record and update it as part of the touch, not as a separate act of bookkeeping — bookkeeping loses to client work every time.
If you are going, go for the right reason. Not because a comparison post said so. Because something in your process now needs to happen without you: a form that arrives at 4:40pm, a follow-up that should fire on Thursday, a sequence that must stop when someone books, a closed-won that must start onboarding. Those are the reasons, and they are decisive when they are true and irrelevant when they are not.
And when you go, do not move everything. That is the mistake we watch founders make: the whole Notion workspace gets ported into a CRM and now the SOPs are in a CRM, which is worse than where they started. The line is clean — Notion holds the definitions, the CRM holds the state. The contact, the opportunity, the timer, the conversation go to GoHighLevel. The stage definitions, the qualification criteria, the onboarding procedure and the reason each workflow exists stay in Notion, because configuration that lives only in a tool's settings cannot be argued with, improved, or taught to your next hire.
The question underneath the question
Nobody asks 'can I use Notion as a CRM' because they are curious about databases. They ask it because they suspect their pipeline is not working and they are hoping the answer is a tool. It usually is not.
Run the real test before you migrate anything: for each stage in your process, name the observable event that moves a deal out of it — an event with a date that someone other than you could verify from the record. Not a feeling. Not 'they seem warm.' If you cannot do that for every stage, then your problem is not Notion, and moving to a real CRM will simply encode an undefined process in more expensive software. That is why so many founders are on their third CRM. The tool was never the problem; the stages were labels, and labels do not become definitions by being typed into better software.
In 17+ years of building operations and 1,400+ founder interviews on The Wantrepreneur to Entrepreneur Podcast, the tell is consistent: the founders whose pipelines work can answer that question in a sentence per stage, and they are running on all kinds of tools. The ones whose pipelines do not work have excellent software and cannot answer it at all. Your problem isn't how many people you meet. It's how many people you move.
So: if every action is you, acting on purpose, during a workday — keep Notion, add the scheduled timer, write the exit events, and go do something more valuable than a migration. The day something needs to happen while nobody is looking, move the state and keep the definitions. If you are not sure which day that is, that is precisely what an OPERATE Report ($1,997) exists to tell you.
Notion is a legitimate CRM right up until something in your sales process needs to happen while nobody is looking — it has no timer that fires into the world, and a neglected deal leaves no wreckage to warn you. If you stay, buy the missing piece: a scheduled job in Make or n8n that reads the database and interrupts a named human in Slack. If you go, move the state and leave the definitions where they are.