Will We Need Big-Footprint CRM in 5 Years or Will Agents Follow Us Around Wherever We Work?

I spent a few good minutes recently with Jeff Rosenthal, head of CRM product strategy and go-to-market for North America at monday.com. Rosenthal is a 25-year industry veteran who spent more than a decade at Salesforce heading up strategy on the Service Cloud and Sales Cloud platforms.

He shared how monday built a sales-focused CRM application on top of its work management platform/project management foundation and in two years attracted more than 27,000 CRM customers focused on what happens after the deal closes. Additionally, we talked about the future of enterprise applications and whether traditional enterprise CRMs will remain relevant as AI tools and microservices evolve to work with unstructured data lakes, enabling users to retrieve information directly through natural language queries without relying on centralized systems.

Below is a clip and edited transcript from our conversation. Check out the full conversation at https://youtube.com/live/EMzJX6Mud8k?feature=share.

Edited transcript

Brent Leary: I had a recent conversation with Don Scheurman over at Pega. We talked about the importance of really being able to focus more on workflow orchestration vs. prompt engineering. And it feels like what you're saying kind of matches up to that because you're focusing more on processes and workflows.

And, and I think that's where your customers want to focus , particularly when it comes to now that the sale is over, what happens next? You don't really want your customers doing the prompt engineering stuff. You really want them to be able to focus more on, you know, workflow automation, process workflow orchestration, and how some of these technologies help to improve and automate that.

Jeff Rosenthal: And we're doing that in the business layer, and we're doing that in the end-user layer. And I think that's a little bit about, you know, why we call it a work management OS. We really put a lot more of the empowerment to build workflow designs and recipes for automations down to the end-user level to do that on their own as opposed to I need an IT project to build out a flow.

And I have to go through more testing on that and more data integration. You know, I think we really tried to bring that down to the end-user level to, to automate things that they just wanted to, to make their life better and happier. You know, it, it's at a different level than big IT-driven projects and governance of all the workflow and automation design and and building. It's more at the user level.

And I think for a lot of like collective teams, they just want to do these things on their own as much as possible, too.

Brent Leary: This is probably a very unfair question to ask, but this has been such a really interesting conversation. What do you think will have more impact on business, in enterprise, like what's happened in the last year and a half with genAI or what's currently going on and moving forward with agentic AI?

Jeff Rosenthal: Yeah, I think the million-dollar question, no, it's probably the billion-dollar question is, you know, what is the future of these big applications that have been built for CRM when everything can just, you know, go to a data lake and be extracted by queries and questions and agents is, you know, do we need to, do we need to have big-footprint enterprise CRMs in five years, or does it all just go to some future state of unstructured data?

And then the AI tools, the user tools, just follow wherever the seller or the user or whoever is doing their job. If they're in LinkedIn, if they're in some other tool and they can just sort of prompt, you know, do I have an opportunity here or is there a deal here, or are they in my database.

Like, why do I have to log into a destination that is the keeper of all of this? And I think that's what makes this an interesting time. In large enterprise, there's the emergence of a giant AI that can search and parse and retrieve and categorize and answer questions. And, you know, in natural language, like there's that often the need for, you know, this big, enterprise system with the Oracle database on the back end that's highly structured.

I don't know the answer to that. I think reality is somewhere in between; that there's still a mission around, you know, the big data within CRM. But you're going to see an emergence of more of these AI microservices for a specific job I want to do while I'm doing my job.

And I don't have to go into CRM to do that. But if I had a crystal ball, you know, I would be able to tell you a lot more that I don't know.

Brent Leary: You know, it's always coming back to that, like over the last five or six months we've had this conversation about, well, the, the rise of LMS and machine learning and all these different kinds of AI and the ability for them to deal with unstructured data.

Will that topple the relational database foundation of enterprise applications, and some think it's a mixed bag. Some are saying, yeah, eventually it will.

And then I had this really good conversation, over at Zoho Day with Sridhar Vembu, who was the, CEO. Now he's the chief scientist. And he said something I hadn't really thought of before, which is, LLMs are great, but who deletes information from one of them? Who has the ability to do that?

With a relational database, organizations can insert, update, edit, delete, whatever. But when it goes to an LLM, how does it get out of an LLM if somebody wants it to? And that's a really great question to ask.

And until that happens, how can how can enterprises really transition away from that current architecture?

Jeff Rosenthal: Well, I'll be following your voice in the industry and your podcasts and, I'm always going to be in the audience when you do these things. So I'm going to be looking for the same answers you are, Brent.