Every CRM is built on a 1990s assumption: that a relationship is a record. After 100+ interviews, I think the model itself is the problem. Here's the case for something different.
Hey, I'm a founder who's spent the past two years deep in the relationship intelligence space. I've talked to 100+ VCs, BD directors, accelerator operators, angel investors, and ecosystem builders about how they manage professional relationships.
I came into this respecting CRMs. I still do, for what they were designed to do. But I've come to believe the CRM model has a foundational problem that no amount of AI features, integrations, or workflow automations will fix. And I want to make the case for why, and hear where this community thinks I'm wrong.
The 1990s assumption
Every CRM, HubSpot, Salesforce, Pipedrive, Close, all of them, is built on the same underlying model: a relationship is a record you create, a row you maintain, a contact you update. The architecture assumes that managing relationships means writing things down and organising them into pipelines.
That made sense in the 1990s when the alternative was a Rolodex and a notebook.
But here's what I keep seeing: records decay. You enter a contact, tag them, maybe log a note. Six months later, the record is stale and hard to maintain. The context from your last three conversations lives in your email, your calendar, your meeting notes, and your head, not in the CRM. The record tells you someone exists. It doesn't tell you who they are now, what you last discussed, who introduced you, what changed in their world, or whether there's a reason to reach out today.
Every professional I interviewed had tried using a CRM for personal relationship management. Every single one stopped within weeks. Not because the CRM was bad software, but because the model doesn't fit the problem.
Records decay. Memory compounds.
The difference matters. A record is static. It captures a snapshot and immediately starts going stale. Memory is dynamic; it accumulates context across every interaction, across every channel, and becomes more useful over time.
When a VC partner told us, "the bane of my existence is having to do things the computer should do for me," he wasn't asking for a better CRM. He was describing the absence of memory. He wanted a system that remembered what he discussed with someone six months ago, who introduced them, what they're working on now, and whether the timing is right to reconnect, without him manually logging any of it.
An angel investor spends 15–20 hours a week on relationship admin across 200+ active relationships. He's tried three CRMs. None reduced that number because the work isn't data entry, it's context maintenance. And CRMs don't maintain context. They store records and wait for you to update them.
An open innovation lead described having hundreds of conversations a month, with "a lot of learnings which somehow get a bit lost." He had a CRM. The CRM didn't help because the context lived across LinkedIn, email, calendar, meeting transcripts, WhatsApp, and his own head. The CRM captured maybe 10% of it, and only what he manually entered.
Where I think the model breaks
CRMs were designed for sales funnels. The core data model is: contact → opportunity → pipeline stage → close. That's a transaction engine. It works brilliantly for tracking deals.
But professional relationships aren't transactions. They don't have stages. They don't close. They compound, go dormant, resurface in unexpected ways, and evolve over years. The person who wasn't relevant to your work last year might be the perfect warm introduction this quarter. The context from a coffee chat in 2023 might be the reason a deal closes in 2026.
No CRM captures that. Not because they're badly built, because they're built for a different problem.
The case for NRM: Network Relationship Memory
I've started thinking about this as a fundamentally different category from CRM. Not an add-on, not an AI layer on top of a CRM, but a different architecture entirely, one built for human networks instead of sales funnels.
What I mean by Network Relationship Memory:
- Memory, not records. Context accumulates automatically across every touchpoint — email, calendar, LinkedIn, meeting transcripts, notes — without manual input. The system remembers so you don't have to.
- Networks, not pipelines. Relationships exist in webs, not funnels. Your network has structure — clusters, warm paths, mutual connections, shared context — and the system should reason about that structure, not flatten it into rows.
- Signals, not updates. Instead of you updating records, the system monitors what's changing in your network — job moves, company news, engagement patterns — and surfaces what's relevant to you right now.
- Compounding, not decaying. Every interaction makes the system smarter. The more you use it, the deeper the memory, the better the recommendations, the warmer the paths. The opposite of a CRM where records go stale the moment you stop manually updating them.
- Network sharing. Every new person that joins your team brings with them a network and set of professional relationships your business doesn't know how to mobilise. Providing mechanisms to share this securely across trusted partners allows for a company or business to see where the trust lies, who can open doors to new markets and what their collective business network looks like.
Where I might be wrong
I want to be honest about this: I'm building in this space, so I have an obvious bias. It's possible that the CRM model isn't the problem; maybe it's just a configuration and discipline problem. Maybe the right CRM with the right integrations and the right workflow automations does everything I'm describing.
But after 100+ conversations with people whose careers depend on relationships, I haven't found a single person who's made that work. Not one.
What I want to hear from this community:
- Am I wrong about the model? Has anyone here successfully used a CRM for ongoing relationship management (not deal tracking)? What did the setup look like?
- Does the NRM framing resonate? Or does it sound like a CRM with extra steps?
- For people managing large professional networks — what's your actual system? And does any part of it feel like "memory" vs. "records"?
- Do you think AI will fix the CRM model or is the underlying architecture the constraint?
I'm not here to sell anything. I'm here because this community understands relationship management tools better than anyone, and I want to stress-test this thinking. Genuinely curious where the pushback is.