Which CRM Should You Use in 2026/2027? Follow the Agents

What CRM should you use today? One of the classics, Salesforce or HubSpot? If so, which? Or one of the next generation, from Lightfield to Monaco to Aurasell?

I was just asked this question by the CRO of a leading AI company that is planning to add 250 sellers this year. And is thinking of leaving one leader for another.

My Suggestion Today: Follow the Agents.

Let’s step back. There has always been innovation at the bottom in CRM. The first investment I ever made was Pipedrive, which back in the day was the breakout leader for simple CRM. It even arguably served as the template HubSpot used for its initial free CRM.

But my advice in general for startups from 2013-2021 or so was the same: Use Salesforce.

Why? Because if for no other reason than it’s what your VP of Sales will want, once they start. So you might as well start there. Plus, almost all of the CRM ecosystem of that generation (Gong, Outreach, Salesloft, etc.) was focused on Salesforce.

And it began to include a lot of the top third-party functionality in it. And it had native integration with HubSpot marketing, which really helped align sales and marketing functions. And for a while, it was cheap. It’s not now.

Starting in 2020, I’d ask the SaaStr community who was using what, Salesforce or HubSpot. By 2022, Salesforce and HubSpot were neck-and-beck. And HubSpot was already pulling ahead, and kept pulling ahead through 2024/2025.

But then everything changed again, at least for some of us.

What Changed at SaaStr: Salesforce Went from Shelfware to Our AI Agent Operating System

I’ll be specific about what I mean, because we lived through this transition in real time.

A year ago, Salesforce was shelfware for us. Still paying for it. Still technically “using” it. But barely. We had a few also sort of using HubSpot. The reality is our dated marketing automation app, Marketo, was really our “CRM”. We though about switching to HubSpot twice. Both times we burned a ton of time and mindshare on the evaluation, and both times the migration was too painful to justify.

Then we went all-in on AI agents. And Salesforce became the most important piece of software we run.

Today we have 20+ AI agents plugged into Salesforce, and it’s the center of everything we do in GTM. Not because we planned it that way. Because our agents all needed somewhere to live, a central hub where data flows in, context gets shared, and 20 autonomous systems don’t step on each other.

Here’s what the stack actually looks like:

Artisan runs three separate AI SDR outbound campaigns: ticket sales to past attendees, sponsorship outreach, VIP reactivation. 15,000+ messages sent, 5-7% response rates. All of it pulling prospect history from Salesforce, all of it pushing engagement signals back in. Monaco handles a fourth outbound campaign and was booking meetings with some of the biggest names in AI from Day 1 after go-live. All syncing to Salesforce. Qualified powers our inbound AI agent on saastr.com. 700K+ sessions. $1M+ in closed sponsorship revenue. In one month, 71% of our closed-won sponsorship deals came from AI-qualified leads. Our historic inbound average? 29-34%. Qualified has full context on every visitor from Salesforce: past attendance, sponsorship history, company data, engagement patterns. That context is what makes it work. Agentforce handles our warm lead win-backs. After SaaStr Annual, I audited Salesforce and found roughly 1,000 people who filled out our “I’m interested in sponsoring” form and got zero human follow-up. Ever. We put Agentforce on it. Because it has full CRM context on every contact (past events, prior sponsorship levels, every interaction), the outreach doesn’t feel like “recovered lead” email. It feels like a relationship continuing. The result: 72% open rate. 10%+ response rate on contacts considered dead. Already closing deals from leads ghosted 6+ months ago. For comparison: cold email averages 2-4% open rates. Our Artisan outbound gets 5-7% response rates. Agentforce on warm CRM contacts at 72% open is a different category. Momentum auto-transcribes every sales call and pushes structured data into Salesforce. Next steps, objections, competitor mentions, decision-maker signals. No rep touches CRM. Attention layers on more call intelligence and auto-populates Salesforce fields. Our reps literally never manually update CRM anymore.

Plus 4-5 specialized agents for email campaigns, de-anonymization, event coordination, and sponsor management. All connected to Salesforce as the data layer.

Salesforce Even Bought 2 Of Our Core AI Agents

Notice something about that list? Salesforce acquired Qualified. Salesforce acquired Momentum. They built Agentforce with 2,000 people. They’re not building every agent from scratch. They’re buying the best agents on the market and pulling them natively into the platform.

The thesis is clear: the CRM that becomes the hub for AI agents wins. The one that doesn’t becomes a database you’re overpaying for.

We now pay more for our AI agents than we pay for Salesforce itself. The agents are where the value gets extracted. Salesforce provides the infrastructure that holds it all together.

Then There’s The Next Generation

There are now real, well-funded AI-native CRM platforms that didn’t exist two years ago. And for the right company, they might be the better answer.

Lightfield started over with a CRM built around “complete customer memory.” You never enter data. Connect your inbox, and five minutes later you have a populated pipeline. 2,500 companies in three months. 100+ YC startups. $81M raised at a $300M valuation. Hundreds already migrating from HubSpot. If you poll current YC startups, the vast majority are using neither Salesforce nor HubSpot. Lightfield is eating that cohort. Best for founder-led sales through ~50 employees at $36/user/month. Monaco is Sam Blond’s play (ex-CRO Brex, ex-Founders Fund). $35M raised from top angels and Founders Fund. The differentiation: AI-native CRM + prospect database + AI agents, all supervised by experienced human salespeople Monaco employs. You’re not just getting software. You’re getting a sales team. And unlikely other AI GTM vendors, it literally gets you appointments set in your calendar with opt-in prospects. We use Monaco ourselves and it was booking meetings with top AI companies from Day 1. Best for outbound-first seed and Series A startups. Still in public beta. Aurasell goes after mid-market tool sprawl. Ex-Harness CRO and CTO, $30M seed from Menlo and Unusual. Replaces 15+ GTM tools with one AI-native platform. The smart move: they also offer a GTM OS layer that runs on top of your existing Salesforce or HubSpot, so you don’t have to rip and replace on day one. Reevo has raised $80M from Khosla Ventures and Kleiner Perkins. Founded by David Zhu (ex-Head of Engineering at DoorDash, scaled from $700M to $75B market cap) and three other DoorDash/Square/Stripe operators. Similar to Aurasell in ambition (replace the whole fragmented GTM stack) but broader in scope, spanning marketing, sales, and customer success in one platform. 90 employees already, engineering-heavy team. The DoorDash platform DNA shows in the architecture: they generate their own first-party activity data instead of relying on integrations. Attio has become the CRM of choice for AI-native companies (Lovable, Granola, Modal, Replicate). $141M raised, $52M Series B from Google Ventures, 5,000 customers, 4x ARR growth. Flexible, programmable data model for technical teams that want to shape CRM to exactly match their GTM motion.

The common thread: none of them were built on the assumption that humans enter data. They all assume AI agents do the work. Salesforce and HubSpot were designed in an era where the rep typed notes after a call. These platforms were designed for an era where the AI listens to the call, updates the CRM, drafts the follow-up, and schedules the next step.