HubSpot Admin HUG's 100th Meetup!
This is a YouTube video titled "HubSpot Admin HUG's 100th Meetup!" recorded in August 2024. It features a conversation primarily between Dharmesh Shah, co-founder and CTO of HubSpot, and Kyle Jepson, a community host. The discussion covers topics relevant to HubSpot administrators, including the evolution of HubSpot, AI agents, automation, company culture, and career advice. The event celebrates the 100th meetup of the HubSpot Admin community and includes interactive Q&A with attendees.
Foreign.
Welcome everybody to another event of the HubSpot Admins Hug. Super excited to have you all here. If you are not familiar with this page, please check it out, bookmark it. We are here every Tuesday at this time and this page is your source of truth for all of our upcoming events. Oh dear, what did I do? All right, you didn't. You saw nothing. Here is where we are. You are here. Where? This is our hundredth meetup. Hundred times we've come together. Thank you all for being here. Many of you have been here for the majority of those really appreciate that you've dedicated yourselves to building this community, connecting with each other. It's really been fantastic and I'm delighted to have Dharmesh here with us. We will get to him as soon as possible. But you know the drill. I got to tell you what's coming up. So next week we're going to have our friends from Tech Form. The week after we're going to have Portent coming and then it was bound to happen eventually. Here is a session about super coming up in September and then we are at Inbound already somehow we'll be having an in person meetup. There will be lots of in person meetups at Inbound. If you are coming, please connect with Deanna connect with me. Going to be lots of admin, sorts of things going on. Would love to meet up with as many of you as possible who are going to be there and then afterwards we're going to have a couple of weeks dedicated to just recapping what happened whether you were there or not. Every year at Inbound there's a ton to talk about, ton to discuss, ton of implications for admins in particular and would love to just have you there as we, as we work through all that and then we have more events farther out that you can look at if you are interested. So that's a look to the future, looking to the past. All of our sessions are recorded and end up on our YouTube channel. We are currently at 99. This one will make 100. We made it. So excited for that. If you ever want to review a session, if you miss one, if you want to share it with a friend, if you they're all here. It's YouTube. You know how that works. It's a great resource. We've got lots of topics on lots of things. Again, a hundred of them. So very excited about that. We also have some space on community.HubSpot.com, it's a great place to connect and talk about events and things. We have a group on LinkedIn if you want to join that share things on LinkedIn, that's great too. But I have to say my favorite place to connect with fellow admins is a slack community that sometimes loads. We are on the cusp of having 2000 HubSpot admins here in the sprocketeer space. This is not in any way officially endorsed or organized by HubSpot. It's an independent community. It's very active. Lots of people in there. Whether you're at the beginning of your admin journey or you've been doing it a long time, this is the best place to brainstorm, collaborate, meet network. All the things you might want to do as an administrator. Definitely check it out. HubSpot Academy runs boot camps. These are free six week launches, cohort based. You come to live trainings on live on Zoom and we still September dates have not been announced but keep an eye on it, they fill up fast. There will be a September session of the Super Admin Boot Camp. I definitely recommend you getting in there if you can. This platform we're using is called Bevy. We are already at 200 people in the room and that number is slowly ticking up. We are a little unclear on exactly what the limit of Bevy's abilities are, but thankfully the troubleshooting is usually pretty simple. If anything seems strange, if someone is frozen, if things are out of sync, if you can't hear what's going on, refresh the page, leave and come back, try a different browser. Those frontline basic troubleshooting techniques are usually all you need to get it to behave itself. Then again, just to reiterate, we do this every week. We've done it 100 weeks and we're going to do it at least 100 more. If you have any suggestions or ideas or requests for speakers or topics you would like to see addressed here at the Admin Hub, we want to make this as valuable and relevant to you as possible. But enough of the icebreakers, let's bring Dharmesh in here. Super excited to have Dharmesh. I assume most of you know, but just in case you don't, Dharmesh is one of the co founders of HubSpot, kind of the original HubSpot admin, our CTO and he's just here to hang out, answer some questions. We're just going to have a conversation. So Dharmesh, thanks for being here.
Thanks for having me. It's an honor.
So before we pepper Dharmesh with questions and you all are welcome to drop Questions into the Q and A if you would like. I would love for him to get to know you a little bit. So let's just play a game where I ask you questions and you just drop answers into the chat and we pretend like we have polling capabilities inside of Bevy. So first up, I would love to know how. How many years have you all been using HubSpot, either as an admin or as a user, Whether continuously or. Or broken up. Where are we at? I see a 12. I see a 2016. That's a lot of years. Lots of 7s, lots of 4. I feel like 4 is probably the most common number here. Nines and tens, four months. Welcome. Hurry up. Glad to have you here.
All right.
Similarly, I would love to know how you're here at the HubSpot admin tug. How did you come to identify yourself as a HubSpot admin? That's not really like an official trajectory people have. Was it accidental? One thing led to another? Is it you love HubSpot and you just want to invest more? Do you have. Do you have. Yeah. Okay, we'll get these questions in. And then I think the question we have to start with, Dharmesh, is can you just tell us some stuff about the things in your background? Because that has been a lot of questions.
Oh, my background in the. In the shot. So that painting, I think, which some folks were curious about, is a painting of John Lennon. And it's an actual painting by an artist here in the. In the Boston area. And what's cool about it is that if you kind of get up really close to it, all it is, there's no, like, lines. All it is is a bunch of tiny little swirls pulled together. So if you just kind of zoomed in on just a segment of it, it'd be impossible to tell that it was an actual painting or a portrait. And as you zoom out, as you kind of step away from it is when you can kind of see that it's John Lennon. I'm a John Lennon fan, Love Imagine the song. But anyway, a funny thing. I have a 13 year old and it kind of crushed me in this. Back when he was, I think, nine, he had some friends over and one of them asked, like, why do you have a picture of Harry Potter on your wall? That's so cool. I'm like, oh, man.
Yeah, it is, it is. So I've got a 10 year old who is a big fan of the Beatles. This is him, actually. His name is Paul. But yeah, it's important to pass this cultural knowledge along or else people lose it. So, Darth Mash, you're kind of the. The OG HubSpot admin. I'd love to know how many. Are you still an admin in any HubSpot portals?
Yeah, I was the creator of the first portal on HubSpot, which portal id 0 back in the day. That one was for my graduate class and teammates, so it was more of a community. So it's like a portal portal. That's where HubSpot was born. Then I have my personal portal, Portal ID 10, my wife's Portal 15 on startups.com, which is kind of my blog and what I use for like my angel investments and things like that. So I probably have. I'm probably admin of like, I'll say maybe eight active portals. The largest one I have a portal for. I built this game called. It's kind of like wordle with friends. It's like basically it's a version of wordle but you can actually have tournaments and play other people. That one has 250,000 contacts or people users in it. The game's been played 150 million times now, so that's probably the largest. So I find myself in and around HubSpot doing both adminty and non admin things every day. Yeah.
So I'm curious not to get too far into the weeds here, but Portal 0 was that app.HubSpot.com 0 or it was.
I think that domain naming thing came a little bit later. So at the time. So the domain attached to that portal was. But yes, you could get to it that way. It was mitsloanfellows.org is what the website was. It had exactly like 85 users and that was all it was ever going to be. So no longer active. That was a long time ago. Everyone's since discovered Facebook and other things, but.
Yeah, that's right. Yeah. Well, since you mentioned side projects like wordle with friends and things, I'd love to talk a little bit about AP Agent AI. This is a thing I've heard you mention to the extent you're able to share and comfortable with a group of, you know, just 268 of our closest friends and family. What's going on there?
All right, so we'll take a step back. I'm, you know, so last year, for those that have been, you know, coming to Inbound or following me, it was the last year was the year of kind of chat spot and just chat UX and co pilots and all those kinds of things which everyone was rightfully excited about. This year, I think it's going to be the year of agents. Quite simply, agents are just pieces of software that use AI to accomplish goals that involve multiple steps. That's the kind of simplest definition. Opinions vary in terms of do they need to be autonomous or not to be, really be an agent. I'm sort of a minimum viable agent kind of guy that says, okay, well, as long as it does something with AI, requires multiple steps and accomplishes some goal, however big or small, it's an agent. So my thinking here is that what agent AI is that over time that agents are going to become very, very good at discrete tasks and that they're going to be at some level part of what I think of as hybrid teams of humans, plus these AI agents kind of working together to accomplish something. Similar to how we might today have people that are on the team full time, we might have some freelancers, might be working with an agency and we all sort of come together that says, here's the thing that we're doing. And I think, I'm not saying they're going to take over jobs or take over entire roles, but they can take over tasks and they can be part of the mix of what's getting done. Now, it's unclear whether that takes a year, two years, or 10 years, but I think that's the way the world is ultimately going to head. What agent AI is, is a professional network for agents. Because if you imagine that world in which agents are kind of participating as part of teams and things like that, well, they should have profiles. So agents can have profiles on agent AI. They have just like we would have on LinkedIn. They have like, oh, here's the things that I do, here's my experience. I've been executed or been running for two years, whatever. The thing happens to be just like a human counterpart might have. And at some point you should be able to, similar to like how Fiverr works, is be able to hire an agent to say, hey, I want an agent that does X, like create dad jokes or build a meme or do whatever it is. And you should be able to hire these agents. So agent AI is the first exploration. So right now I'm using it as a platform to deliver some of the things that I've built, software I've built over the last 30 years that historically that I just built for myself because there was no convenient way to really launch it. Relaunching some of those reimagined as agents on Agent AI, they're proving popular and useful. It's 30,000 users already, even though it hasn't officially launched yet. So, yeah, a lot of fun.
It's really interesting. Sort of turning the gig economy into a place for bots to find work and satisfaction in life.
I have to say this, and I'll say this at the talk again, in terms of the kind of replacement of humans it is, I could see why it would be a little bit scary and anxiety inducing. But the way I think of it is that I think a human is much more than some collection of the tasks that they do.
Certainly. Yeah.
So like, if you look at me, it's like, oh, well, there's AI software now that could write code, there's AI that can respond to emails. It's like, okay, so let's say I had an agent to do all those things. Does that make me redundant or unnecessary? I would hope not. Right. All it really does is to say, hey, if I can find agents to help do the things that I may or may not enjoy, may or may not be good at, it kind of raises kind of my value to the world, I think, because I can focus on kind of higher value things. And the other thing agents, and I'm experiencing this already, that AI is capable of doing, is that it sort of opens up doors, allows you to do things you weren't even able to do before as a human. Right. So there may be a skill set you just didn't have. I can't design things, I'm not a designer. But now with some of the image generation tools or whatever, I can sort of piece together things and have some degree of independence to be able to accomplish more than I want people to. You know, AI before aided stuff.
Yeah, I love that vision and I will say a point of optimism on the, on the AI front and I would love for all of you listening in to weigh in on this. But just a couple weeks ago here at the Admin Hug, we had a session. A couple of people from HubSpot's product team just did kind of a listening session. Hey, admin, how do you feel about the concept of agents? And if HubSpot were to build an agent, what would you want it to do? What would you not want it to do? Where are these guardrails? And, and I found it really impressive and inspiring that the group who showed up, which certainly a sampling bias there.
Right.
The people who come to a session on agents are maybe more excited about, but they were all really enthusiastic. There were some poll questions they did in slido that were like, I'm excited, I'm nervous. And everybody kind of tended toward the excited thing. Which to me is really interesting because a year ago at Inbound, HubSpot announced AI, right, we're going, we're doing all these AI things and when I did my session things everyhelps admin needs to know, I just sort of in passing showed like. And the settings page right, right now is just a big toggle switch where you can turn all the AI assistance off. And people applauded. They were excited. Yes, we can turn it off. Because they needed time, you know, to get used to this concept and understand what it was doing and how it worked. But now, less than a year later, I feel like a lot of those same people are like, no, I get it now and I'm excited for this. That's what a time to be alive.
Indeed.
Yeah. So we did collect some questions ahead of time and I see we have 25 questions in the Q and A now. So we'll figure out how to sort through this while I read off some of these questions that some of you were kind enough to submit through our form ahead of time. If everyone who's listening in wants to go over to the Q and A and start up voting questions you're interested in, we will go most popular to least popular over there. But we have a question about AI agents and maybe you've already answered this. Daniel, if you're here, asked, do you see business applications building based on AI agents, that is a swarm of AI agents doing various jobs and delivering results as a service together becoming available to non technical users in the near future?
The answer is yes. Once again amongst us friends, here's how I see this playing out with agent AI. My first pass at it was a platform for myself, so I can easily develop these agents and launch them publicly so I can make them available to the world. The big secret thing that I've been working on that I'm hoping to get completed sometime this year is around allowing other people to build agents on the platform that do multi step things or whatever, invoke any manner of large language models or image generation models and things like that and be able. So the way to think about it, since you folks are admins, you'll totally get this, is that so imagine a world before marketing automation existed, right? There was no HubSpot Workflows app. This is back 20, 30 years ago. @ that time you had to sort of be. I mean automation still happened in some sense, but you had to sort of be a developer in order to be able to really do these kind of programmatic follow up things or whatever. And that's exactly what larger companies did. What HubSpot and others provided was a kind of what many would now call like a low code, no code way to develop automation scripts. That's what the workflows app in HubSpot is. Now if you pull that forward into 2024, you can imagine something similar to that that says okay, well that was great. App gives me access to HubSpot data, I can update things, I can send emails, I can do the things that we do in either marketing automation or sales automation. I think we're going to have there's a similar model that could exist that's AI native that was born to be able to build these AI agents that are scripts. It could look a lot like a workflow app that just happens to be AI powered. That's the thing that's noodling in my head is can I make it possible for non developers to actually solve because they have the domain expertise, they know what they need to do either for the end users or for their own company is to see if I could. It's an experiment. We'll see how far I could get down that road. But yeah, that's the fun part.
Yeah, yeah. And to me it feels, I don't know, 2024 to me feels like in a lot of ways 2006 all over again. Like the Internet is just being reinvented right today. I don't know, I don't know how many of you noticed, but HubSpot just launched AI Searchgrader which is like website Grader was for SEO in 2006 but is now for ChatGPT. Answers to questions and things to say how you're doing. And I just imagine a lot of these patterns will repeat, but I assume kind of go faster. Right. Because we know no code. Low code solutions are what we all need and want for a variety of things. And so for AI to get to that point, if I can drag and drop and build an agent to do something that would just be, that's a future I'm excited to get to.
And one thing that I think that would be interesting, and this is what I've kind of experienced building agents, even just for my own personal use, is that it's really a great way to drive your own individual productivity. So I'll bet you buddy Kyle, that you have a set of things that you do every week as either related to this particular show or other things that's like, oh, I go here, then I download out of Bevy and put it in there. Then I do this and put it in Google sheets. Then I generate a chart or create a slide, whatever it is. There's these kind of series of things that you do, some that involve HubSpot data, some that may not, but it'd be nice to say, oh, well, I want to try and automate some steps of some parts of those things instead of all being manual. Now with the power. You couldn't do this three years ago, but now with the power of AI, because it can understand natural language, because it can go through a blog post or get a transcription from an audio or a video track and summarize that there's a bunch of things it can do now as part of its tooling that just wasn't possible before. That's what.
Yeah. And I feel like there's a lot of things we've all been trying to do, and there have been kind of blunt object solutions to macros that, like, where it makes your mouse click actual buttons, or like, you know, things like Zapier or make, or things where it's taking this piece of data to just be able to tell a piece of AI, like, here's the job I'm doing. Go do it.
I think one of the mistakes the industry made. So last year there were a slew of projects that launched, which is sort of when agents first started. There was baby AGI and auto GPT and things like that, where you basically typed in whatever random goal that you wanted and it would use large language models to break that up into tasks and recursively go through it, which was a fascinating idea, but just too many degrees of freedom because it was almost invariably failed. It was just a matter of time. By the time it got to step five, you got the compounding error rate that even if it fails only one out of 10 times, by the time you get to six, seven, eight steps, the failure rate is near 100%. And I think that's. And I'm going the exact opposite way. Instead of having this do everything, it's like, okay, can we say, like, I know what steps need to be executed, I know what's possible with the technology. Just let me glue it all together. That's what I really want. I don't want to just take over and figure all of it out. It's like, okay, I just. I want both of these kind of discrete steps. That's okay.
Yeah, yeah. So I don't know why this makes me think of this And I apologize to everyone if I just ask questions that are interesting to me. But something that's been really interesting to me to watch is the development of Chat Spot. Because somebody here in the audience is bound to remember 2017 when you did Growthbot, which was like trying to accomplish Chat Spot before the technology existed. Right. And I'd love to hear from you, like, how has it been different this time around with Chat Spot? Because you mentioned, like you just said, like, trying to make things simpler and fewer steps and reducing errors. I remember Growthbot being sometimes a little hard to communicate with, and Chat Spot seems much friendlier, you know, how has that been for you as the creator of these things?
Yeah, it's been great. So the, you know, the analogy I like to use, I'm big on analogies, is that AI is just not another kind of coat of paint over a house that says, oh, now it's painted with AI and look, it's bright and shiny. It's actually a new material that we're building with. Right. It's a new thing that says, okay, well, some of the physical constraints we had with houses before now no longer apply. We can build anyway. So we can build stronger, we can build faster, we can do things that just weren't possible before. And that's what makes AI exciting. I've been developing software for three plus decades now, and this is like the most excited I've been because it actually expands the kinds of things we can do with software. We saw that with the Internet, once every computer was connected effectively to every other computer. A whole slew of apps, whole slew of use cases were now solvable that weren't solvable before. We're going to see the same thing. There's a bunch of things now that computers can almost think for some definition of think. There's just a whole lot of things that we're capable of doing now that we just couldn't do before.
Yeah, I'm excited to see how we even approach building these things differently. If we think of even pre Internet, like the advent of different building materials, there are some really cool old wooden bridges. Right. But once we got lightweight, very strong metals, we didn't just build wooden bridges, same design, but now out of metal. Right. Like, we could build them different shapes, we could make them span different lengths, and I expect AI will do very similar things with, with technology. Back to questions from the actual group. 1. One question we got here that I think is interesting is what's your opinion on how companies should balance software versus human interaction in B2B interactions.
Yeah. My advice here is to start slow. You start off with a human using AI. The next step of evolution is you still have a human in the loop to prove something before it goes out to a customer or a public blog post or something like that. My sense right now is that I just don't think the software or the technology is ready yet for just handing the keys over completely autonomously. It's like, okay, I've kind of written my bot, I've done the things I want to do, now go have at it, right? That's the way to kind of get comfortable. Like once you see it, do it correctly 50, 100, 500 times and you have some sense of how, what percent of the time it's accurate and when it's wrong, how wrong is it, in what kinds of ways is it wrong? Sometimes it's okay to be wrong as long as it doesn't say something like completely offensive, right? That might be all right. Like, okay, humans make mistakes too, but it's not okay to insult the user or say something just shockingly racist or something like that. Those are things you want to make sure you protect against. My advice is use AI. Start weaving it into your business, into your daily flows, or even as your own personal use. Maybe you're not automating things, you're just using the HubSpot tools or other tools out there to do things. Then as your level of comfort increases and you know the scope of risk, right, there's always risk with all new technology, then you can sort of expand the use. It's like, okay, now we sort of know what's going on. We know that it's good at writing social media posts, but maybe not as good at writing 900 page white papers on a topic in this very kind of niche industry yet. So just sort of know the limitations and get a feel for that first.
Yeah, it's interesting to hear you say that because when I think about just traditional automation, things like workflows and stuff, so often what we emphasize is you have to have a good process, right? This is going to accelerate your process. And if you have a bad process, you're going to. AI is a. It's those same principles apply, but it's also, it's a lot more, it's a lot less like setting up an automation tool, a lot more like training an intern or something, right? Like if I wanted to hire my 10 year old son to be my social media manager, I would not tell him to like go look at all the comments on your favorite YouTube videos and just do that on LinkedIn.
Right.
Like that would, that would be a recipe for disaster. And if you think about how you're going to train it AI and at what point you're going to, to send it off to work on its own, like there are lots of intermediate steps you need to take.
Yeah. One thing I will say, because this comes up with customers and partners all the time, is that I understand the anxiety and apprehension around AI, but my advice would be that you shouldn't wait another six months, 12 months, until the technology gets to a point where you're 100% comfortable with it, even if you don't use it in production. Start getting familiar with what the possibilities are. So that way you can kind of sort of follow along in terms of how the technology is evolving and it's going to vary from business to business, industry to industry. But what you don't want to do is kind of sit on the sidelines waiting for the technology to hit critical mass in your industry because you sort of miss this opportunity. Not necessarily to get the advantages of AI, but to learn the things you need to learn. Because then when you do start, then you're starting from a cold start and you're trying to figure it all out and try to figure out how to roll it out. It's much easier to build on something that already exists. And we've learned this not to keep bringing it back to writing code or workflows or automations is like always don't try to sit down and write a 50 step automation. No one does that in their right mind. They're just like, oh, I'm going to write a three step automation that does exactly these three things, looks at this property, updates that property and send a very simple email. And then you sort of build on something that works incrementally over time. Right. So the same way, don't try to put AI in waiting nine months. It's like, oh, in the future I'm going to do it all, do it incrementally. It's okay to take it in baby steps. It's new technology, it's understandably anxiety inducing sometimes. But yeah, start slowly to whatever degree you're comfortable.
Right.
It just should be zero. That's my, that's my argument.
Yeah, yeah, yeah. Well, and I think in, in any field or any discipline, the, the mark of a true expert is not knowing the answers, but being able to formulate intelligent questions, understanding what, what questions a tool can be used to answer. Right. Like if you're Learning to code for the first. Right.
If you're.
You're writing your very first Python scripts, understanding the sorts of things you might use a Python script to do is a huge first step in being able to do anything else. And I think getting in and playing around with AI tools and things like agents as they come around and just understanding like, oh, this is the sort of thing I can think I can offload, and this is the sort of thing that I want to keep my hands on, is you got to build those intuitions.
This is right up. I was going to make one more. So one of the things, I think this is going to sound immodest, because it is, but I think one of the things that HubSpot has been good at has led to our kind of modest success that we've had so far is that we like teaching things, we like demystifying things, right? So even back 18 years ago when we started, it was like, oh, there's this thing called SEO and there's these social media sites and things like that. We don't want to make things a black box. We want customers to understand what these new things new to them are, and we're very good at it. You are a manifestation of this, Kyle, which is you love to teach, right? Like, I watch your videos all the time and including product updates and things like that. And I think, you know, HubSpot's approach to this is going to be very, very similar to the approach we've always had, which is, yes, it's new technology. We believe our customers can and will benefit from this, but we're going to hold their hand and help them through the education process because we think that's the right noble thing to do.
So, yeah, I always tell people the thing I love about HubSpot, and the reason I think I'll be here forever is it's a philosophy first company. We are an opinionated platform. We believe there's a right way to do business and to use technology, and we're trying to help people do that. HubSpot was a blog before it was software, right? And, and that's another reason I feel like 2004 is 2006 all over again, especially with this AI search. Greater right. We're going to help people understand how they're doing in AI answers, and then you can bet we will have opinions and, and methodologies on, on how to succeed in this new world.
We'll.
We'll get there. Okay, I'm gonna look at the live Q A now. There are 45 questions? Let's, let's see. I'm gonna sort by most upvoted. Oh, this is having trouble loading. Okay, 2018. Oh, it doesn't know how to count. All right, we'll resort this again. Give me a second. All right, I like this one from Kyra. What has been your favorite product launch HubSpot has done in the history of HubSpot?
I guess as a maybe not even a parent now grandparent of all these features, I love them all equally. Of course. The thing I get most excited about, I'll say this, is when we open the platform up. Things that new APIs as we're constantly enhancing new UI extensions that we're doing now to open up the UI of HubSpot so third parties, including our partners, can build applications. The reason I say this is that deep down inside my core, I know that HubSpot is not going to be able to build all the things that all the people need. That's just simply physically impossible to do. And the degree to which we can kind of make HubSpot more open via APIs and UI customizations, and we saw this with custom objects back in the day, like all those things, I think it sort of unleashes folks that are on this call to say, oh, if only HubSpot had a way for me to get into this part of it, then I could do these things to help my company out. Those are the things that excite me. It's like every time I'm watching the product updates or whatever, I'm looking for those, it's like, oh, yeah, that's something that, that's going to be so cool. It'll be interesting to see. I will say this is that this year I've seen more Progress on those two fronts like APIs and UI extensions, because it's a big priority for us that I've seen in probably over a decade.
Awesome. Love that. I assume this audience loves that also. Richmond asked the question. This is a great question, Richmond. This is hard pivot. Big change of topic here. But if I remember correctly, I learned that you've remained an individual contributor and have no direct reports. Do you think more people should follow that path? Does it make sense for a HubSpot admin? How do you move up the career ladder that way?
Great question. So for those that don't know, when Brian Halligan and I co founded a HubSpot, we had our kind of founders meeting and we decided two things amongst many others, but one which is I was not going to be CEO. He was Going to be CEO, which was awesome because I'd been CEO before and I suck at it. But then the follow up, which I had not planned going into it, which is, oh, by the way, Brian, I'm also not going to have any direct reports. And that obviously is unconventional. And I explained, I'm like, okay. I'm a reasonably intelligent person. I think with some coaching and some training I could become passively okay at management and managing people. I don't want to spend 5, 10, 15 years of my life becoming passively okay at something. So let's have an agreement right now, before we even start the company that I will never have any direct reports. And we've held and like, look me in the eye, it's like, even though our VP of engineering quits, you know, five years from now or whatever, we're not even going to put me in a temporary, transient, interim, whatever role because. And like looking back on it now, that was one of the kind of better decisions both for HubSpot and for me personally because I get to do the things that I enjoy, bring value to HubSpot in the ways that I can bring value and I don't hold the company back. Now, it's not the right thing for everyone. But I will say this as kind of macro level advice since the question was asked. I'm a big believer in, and I think this is broadly applicable, that a lot of people spend too much time trying to shore up their weaknesses and not enough time doubling down on their strengths. So if there's something you are really, really good at, now get me wrong, if you have one of your weaknesses is that you're a serial killer. Stop doing that. Let's get rid of that weakness. But the world will let you get away with a lot of weaknesses if there are a handful of things you are world class at in the fullness of time. I think that shows up because in the fullness of time, whatever org you are in org you go to, the value you will get from the organization will always be a function of the value that you're adding. And in lots of organizations, because of the way org structures evolved over time, value has historically been measured by how many direct reports you have and what the downline is. I can't believe I used the word downline. I've never used that before anyway. But so my advice would be to kind of stay open to remaining an individual contributor even if you have managerial responsibilities that sometimes those are thrust upon you. That's the way life works sometimes, is to keep your Hands in the thing that you love and enjoy and keep mastering that craft. So even if you're spending 80% of your time on managing, because now you're a manager and that's your job, try to find a path that you continue to do the things that you actually enjoy, that you're really, really good at. And over time, I think that pays dividends. That's why I still write code, like every day, right? Like, not that I. Because I have to. Because I think all my. I love it and I'm pretty good at it, right? Like, that's. It's a great way for me to add value to the world into HubSpot.
I love that and I feel like I've been a beneficiary of that. HubSpot invests a lot in having good managers. And every manager I've had is someone who's a manager because they want to be a manager, not because they were thrust into this position because that's how the career ladder takes them. And I really appreciate all the growth I've been able to. I get to do this for a living, right? I have no direct reports. I am all in on being an individual contributor until the day I die. And there's been space for that here, which I think a lot of organizations maybe wouldn't have made. So thanks. What is your biggest advice for an admin within HubSpot?
This is a really good question. Not that you aren't already, but wherever you are right now, in terms of your understanding of the actual business that your business is in, your company is in. Get closer and closer to that. Get closer and closer to the actual sales, closer and closer to the actual marketing and the stuff that's actually happening in the business. Spend time on sales calls. Spend time with the marketing team as they launch a campaign or whatever it is they're doing. Because what that helps you do, it's a little bit. I'm a product guy through and through. The best way to get good at product. And that's what admins are producing too. We're all producing product in some shape or form. The closer you can get to the customer, the better the things that you will produce. The more understanding you'll have of what's actually necessary, the better you can push us on things. It's like, hey, if you come to us and it's like, I've talked to 20 of my 30 sales reps and they're all pissed off by X. If we hear that from enough people, it's like, okay, this is like A real issue versus. Okay, well that's just Dharmesh in his kind of admin world thinking weird thoughts or whatever. But if you could kind of bring it back to the customer, bring it back to the actual kind of end users, so to speak, I think that helps. So get close to business, get closer to your internal customers.
Love that. Amanda asks, and I will preface this with no spoilers. This can just be a binary yes or no. Have you written your dad jokes for Inbound yet?
No, but I have attempted to write a dad joke generating agent. Oh, it is not going well. So I've done. This is a weird. We're gonna have to take this little digression because the question was asked because it's a fascinating experience. So I also tried to build a meme generator that takes popular memes and current so old classic memes that have gone viral in the past and then new trends or new things or new things in the news. However, Nichi and combine those two things into a new meme that cleverly weaves those two things together that's going really, really well. So it's free to try go to Agent AI, sign up for a free account to get 100 credits and that does a pretty. I give it like a seven to an eight, give it two or three tries and it gets it right. For some reason, dad jokes are harder. I haven't quite dissected why yet, but yeah, anyway, so they're not, they're not written yet. I might have to sit down and find another path. We'll see.
All right, John asks, tell us about how the HubSpot culture code came to be and what you learned that might be able to help us influence our own company cultures.
Yeah. So the HubSpot culture code started as a 16 slide deck internally to kind of help us assess what kinds of people would fit in at HubSpot and do well at HubSpot. And it's since kind of grown. It's 128 slides now. It's been 120 slides for a decade. Plus a couple of kind of key lessons to bring out from that, which is every company actually has a culture. Whether you like it or not, it's going to have a culture whether you do it deliberately or not. But we think of HubSpot of culture as a product. Shocking, right? It's like on the product guys like, oh, well, well, culture is just a product that you build for your internal employees versus the product that you build for customers. And so in that vein, the way to really kind of make because culture can be a very mystifying, very polarizing topic. Even at HubSpot, when we first started talking about culture, people were like, ah, you can't really build culture. That has to come organically. It's like there's all really good arguments, but if you think of it as a product, then lots of pieces fall into place which is, hey, you would never build a product without asking customers whether it's working for them or not or what features would they like, what bugs have they found? So we do this at HubSpot across all 8,000 plus people. And Kyle, you know this. We ask, it's like, hey, how happy are you with the culture product working at HubSpot on a scale of 0 to 10 and why did you give that answer? And we look very, very closely at that feedback that we're getting and we iterate on culture. So kind of step one, if you're not already doing it or if your company's not already doing it, collect just those two simple questions just to get a sense for whether people there are happy or not and what's making them happy or not. And then just treat it like a product and say, you know what, we're not going to turn the ship overnight. But of the 14 things we heard of the bugs that are in our current culture, whatever happens to me, it's like, oh, it's too bureaucratic to do this or it takes too hard to do that or we have too many jerks joining the company. At least identify the problem and say, this is the thing we're going to start working towards improving and you'll see that show up in the numbers to say, oh, fewer people mentioned that as a bug. It looks like that's getting better. But you've got to start somewhere. Once again, I'm a big believer in the kind of dream big but iterate small. It's like, okay, here's what I would love my company's culture to be. What can you do this week, this month, this quarter? Right? And just start small and take baby steps. That's my advice.
Yeah. And I have to say the culture code is the reason I applied to work at HubSpot, which doesn't make me special. I think a lot of people it is, but I applied with a cynical part of my mind saying this is probably just marketing hogwash, trying to get starry eyed dreamers like me to, to apply to this company. But I, on the off chance this is real, I want to, I want to try. I've been here nine years now. And I, I'll tell you, all 278 of you, it's, it's real. A lot of it's aspirational. Like, we're not perfect, but we're working towards it. And for me, the biggest evidence that proves out the hypothesis that culture is something you can build is not what's happening in HubSpot, but it's happening in the ecosystem at events like this. Every HubSpot event I've ever been to, anytime I go to Inbound, everyone I meet is. Is this type of optimist and dreamer and trying to not just build their business but, but, but do things in a good, sustainable way. And that, that to me really speaks to what's happening internally at HubSpot, that there's this, this sort of aura around HubSpot that is in our solutions partner, it's in our app ecosystem, it's in our customer base. Somehow we just foster this sense of belonging and passion and growth that is really unlike anything else I've seen anywhere.
Yeah, that's been one of the most heartwarming things about HubSpot's 18 years old now is that we have a certain set of values in our culture which we've now documented, made public. That culture shows up in the HubSpot product itself. You will feel it. There's things that the product won't do because we think it's contrary to what we believe is good over the long term for customers and humanity. But the other thing is that just like the community, like this one. Right. And the other hugs, and there's all sorts of this kind of mini communities in this kind of broader ecosystem. And it's not that we actually like created that. What happened is that we drew in people that thought about things a certain way and then we sort of at some level gave them permission that it's okay. There are other people like you look over here, and that's what's really been heartwarming, is to see this kind of a community come together. And just like you said, it was a blog before it was a product. It was also community before it was a product. Because we started having those kind of early kind of meetups before the product was anywhere near complete. Right. And that was because it was a movement, it was about inbound marketing kind of back in the day. But we're still kind of driving that same kind of ethos to say, you know, small businesses are worth serving, we can help them grow better. And it's through education, it's through product, it's through community and connection.
Yeah, yeah. I love that idea of giving people permission. It makes me think of inbound five or six years ago when we came up with Grow Better, Right. And we had that Megan Keeney Anderson voiced over video that made the whole auditorium cry. Right. Because we were talking about how this idea of growth at any cost and without any regard for downstream effects, it isn't how you have to live your life or build your business or navigate your career. And that's really freeing to be able to buy into. All right, another big change of topic, but this one I love. Chris Evan after my own heart what does the everything is an object world look like in platform? And how soon might we get there?
Okay, so this is one of those 18 plus years in the making. So let's take half a step back for those unfamiliar. You all know the kind of standard objects in HubSpot contacts and companies and things like that. But what makes an object an object has evolved over time. There's this underlying framework called the HubSpot Data Framework that all these kind of objects kind of share software capabilities from. So when you look at a list of contacts or you look at a list of companies, you will notice that you can do the same kinds of things regardless of what the object type happens to be. Right? You can create a list of them, you can build workflows off of them, you can add custom properties, all these things. That is a very intentional, very deliberate, decade long project that's still ongoing. Because what happens when you have that framework underneath is that every time we add a new object, we're not starting the software from scratch. It's like, oh, by virtue of being an object, you are now endowed with all these capabilities, even as a piddly little new object that just started life yesterday, was born yesterday. Now you have reporting and workflows and triggers and custom. It's like all this stuff that you get as an object. I'm trying humanizing objects just for a second here, but it's like it's got superpowers from day one, right? Versus if someone were just building an app that's going to track whatever the thing is, they would have to build all that surround sound functionality. That's what's interesting is that HubSpot grows along two dimensions. One dimension is the framework gets more and more powerful in terms of what the framework is itself supports. And then we add more and more object types within the core platform or through custom objects and things like that. And those are what math people would call orthogonal things. Orthogonal dimensions that can grow independently. So we could just say, oh, the data framework is going to stay exactly what it is right now and just keep adding objects and the product would get better, the platform would get more usable. Or we can say we're not going to add any more new objects and the data framework and just get better and better and those objects now are capable of doing more things. All right, so what I'm most excited about is two things. One is things that are in the early parts of the product, before the data framework was a thing we have things like in the marketing Hub product right now are not full fledged objects. And this is what you know like and you folks have experienced as you go into parts of the product. It's like this feels that we're like, well I can do all of this over here, why can't I do the same thing over there? Why is it different? Right? It's sort of like if you have a dog, your dog would bark at it, right? Something's not right. And so I'm excited about two things. One is, I'm excited about and this is happening, right? All those what I'll call some of the kind of legacy entities that are within the HubSpot software are being promoted and migrated over to being full fledged objects. And you've seen that happen over the course of the last several years. We're going to continue to push on that and we will get there and all new things obviously kind of start life as objects because the framework does exist. But the other thing that's really interesting to me is new framework level features that we can add and thereby add capabilities to every single object that's out there. Let me give you an example that does not exist. I'm not promising it. It's been in my head for a long time. Imagine you could have any object, let's say you're looking at a company or something like that and you want to subscribe to that object. The subscription can be like, I want to be notified if this object changes or someone deletes it or something happens or changes stages a campaign goes from not having been published to being published, whatever happens to be. And here's how I want to be notified. I want an email and it shows up. That's not just, oh well, people are interested in when a landing page gets updated or something like that. We didn't go build a feature inside the landing pages thing. We're just like, no. If a landing page is an object and objects are subscribable now every Single thing. It's like we don't have to pre anticipate when subscriptions might be useful, who might want to subscribe to what kinds of objects. We make the unilateral universal decision that subscriptions are a good thing, let the user decide what things they want. That subscription is just one example. We could do 1000 different things. I've got a 20 year roadmap in my head of things we could do in the data framework that I think would be super cool and useful.
Yeah. The nine years I've been here that I've loved watching this march toward being framework supported, one of my favorite demos and HubSpot history. So for all you, HubSpot has this internal monthly meeting called Science Fair where the product team just shares demos of things they have been building. And there was one long time ago when we first decided we're going to do this thing called Service Hub. We hadn't even decided it was called Service Hub. Sophie Higgs did this demo of the tickets object and it looked exactly like the Deals object. It had drag and drop, it had all the stuff. And it was simultaneously the most boring and most inspiring demo I've ever seen at HubSpot because it was like, we have this new object and we were all expecting something new, something, something otherworldly that we'd never seen before. And it looks exactly like deals. And it's like, I'm not disappointed. I know how to use this today on day one. And as HubSpot has moved more and more onto that framework, we've seen all sorts of roadblocks fall down. Right. Many of you have been watching as this year, workflows have been very close to object parity. Things you could only do on contacts you can now do on others, things you couldn't do on contacts you now can. And that is the fruits of these labors. Alba wants you to spoil some inbound releases. I assume that's not a thing we're gonna do.
Nothing I can do. Honestly, I'm not sure exactly where the cards are gonna fall in terms of things that we're gonna announce or kind of talk about. And like, there are two things that I. This has kind of kept me out of trouble, that I never talk about things that involve numbers because then the lawyers are not happy, and things that involve dates or features because then the product team is not happy. So I do not commit the product team.
Yeah. One thing I will say as a person who is making a living by announcing the product team's work is it's really interesting to pay attention to what's happening in the month before Inbound. Some years we have a ton of public betas hitting and when Andy Petrie gets on stage and says we have all these new features, you here in this crowd are already opted in and using them. Sometimes things hit it inbound that just like, will feel like they came out of nowhere and like, whoa, I had no idea this is coming. And I don't really know which kind of Inbound this one will be, but I would say watch the product updates tab in the next few weeks because there's either going to be a lot of stuff coming or an eerie silence. And that will be.
I will say this. This I can say is that I can use a scale, but I'm going to say this in terms of the product features, launches, announcements. This is probably one of the most prolific years the product team has in terms of things that you'll see at Inbound. So it's a very product heavy. This is just like Apple. Apple sometimes has what some people call the S years where they kind of tweak the phone. They don't really do anything dramatically new. This is not one of those years at HubSpot. This is a full fledged product here. That's the star of the show this year.
All right, let's move on before we get ourselves in trouble. What were your initial goals for HubSpot and did you have any idea that the company would be where it is today?
Did not know. And one really can't know for sure what something's going to turn out to be. But we did. I'll say this, when Brian, I started, by the way, I was not supposed to start company. I was. So HubSpot actually wasn't supposed to happen. I promised my wife, after having been an entrepreneur and a startup guy for as long as she'd known me, she'd met me before I actually started my first company. I wasn't going to do it. And I promised that I'm going to go to grad school. That's why I went to grad school and maybe get a PhD and then maybe go teach. So that was kind of my original plan A. But then I met Brian Halligan in grad school. And so we made a choice and I said, okay, well. And both of us had had, you know, some success in our lives. And this was what. I'm not a sports guy, but this was our last step up to bat, right? We're like, okay, if we're gonna do this, you know, because I had you know, I had some singles and doubles before, I've had some successes. But no, like, you know, flat out, whatever, the slam, whatever. Yeah, that what you just said. And it's like, okay, so Brian and I were both in perfect alignment on this is that we're going to swing for the fences, right? So every time there's a decision to be made, we are going to take the decision that maximizes the probability that we have that kind of spectacular outcome. So either it's going to be a spectacular outcome or we're going to go down and crashing burning flames, right? There's no, we're not looking for the middle road, we're not looking for the modest outcome. And then so that's I think contributed and it's made a lot of things easier, a lot of decisions easier. So whether it comes to raising capital or how we're growing, it's like because we felt like a company like HubSpot needed to exist in the industry that HubSpot exists in. Back in the marketing days it's like okay, well inbound marketing happened, right? It's like okay, that was a new thing, not enough people were doing it. It's like small businesses should be benefiting from the Internet. Like this is a noble calling, this is something we can fix big enough market, all those things. Then as we've evolved into kind of CRM, it's like okay, well a built from the ground up CRM platform needs to exist. That should be a thing. Not with no disparagement at all intended towards the enterprise. CRMs that are out there, they're really, really great products for a segment of the market. If you are a Fortune 1000 company or you're running a small country, those are the products you should be using, right? But for a vast majority of small to mid sized businesses, that's kind of overkill. It's like driving a tank to the Trader Joe's or something like that, right? It's just not, it's not the right fit anyway. So that's so we, we've had a relatively kind of smoothie curve. No breakthrough like we woke up on a Monday and you know everything. Tedx overnighter like it was been a very steady. Keep listening to customers, keep making the platform better, keep iterating on the culture as we're evolving and just try your best to be around amazingly great people and the rest seems to work itself out.
Have a related question from Mike at Resonate. What is the future of HubSpot that we should be prepared for? I see that some BI features are being implemented and there are lots of exciting developments for developers. Will we see new hubs like a business Intelligence Hub, project Hub, Accounting Hub, or a fully featured development platform? Or will the focus remain solely on marketing, sales and service?
I literally can't answer that because I don't know the answer. This is something we talk about every year, right? In terms of what does the product footprint look like? Is it going deeper on some of the hubs we have? Is it adding a new hub? Is it making some kind of platform level enhancements? And we sort of, kind of based on the state of the market, we try to balance those things so that we have, you know, kind of our strategic off site every year. So right now it feels, I'll say this like, feels relatively complete. Like when service Hub was missing, it felt like there was a leg of the stool missing. I'll say that, right? It's like, okay, well this, these things sort of really go hand in hand. You sort of need that in order to be a full kind of customer platform, which is what we at the time had aspired to be. So I don't think we have like the burning, oh, this has to exist in order for us to be a customer platform. I think we're kind of beyond that point now. But we have lots of ideas for hubs and for features and for platform stuff. It's just a matter of sequencing. We'll get around to it or most of it probably over time. So yeah, sorry for the bit evasive of an answer, but it's.
Well, and the thing is, HubSpot adapts a lot. We sometimes set one plan. We thought 2023 was going to be the year of commerce and then it turned out to be the year of AI. And we still delivered commerce stuff, but it took a little longer because we had to pivot and that's how things work over here. But I would encourage all of you as admins to keep speculating, keep requesting, keep dreaming. A lot of people lately have been telling me once users are an object, I would love for HubSpot to replace my like employee management software, right? Like keep telling us that. I don't know if we'll build it or not. Maybe, maybe, maybe not.
This is a biased opinion. Goes back to the kind of earlier question, like, what would I do if I were you, if I were a HubSpot? Admin is really like learn and understand. Not that you don't already, but even deeper than you know right now. Get to know the platform, right? This is like the early days of anything, whether it was the iPhone or the web back in the day. Because there's going to be massive opportunity both within your company and outside your company. You know, you're going to live a long and full life that will pay dividends because the platform is going to be more and more open, we're going to be solving more and more use cases and the world is going to need people just like you. That's what I will say that the even right now as we sit, there's an asymmetry between the market need out there in terms of people that really get HubSpot, both the product and the company and the vision and all of it and love staying up with all of this stuff and the number of people that like the. There's an asymmetry. So you get to kind of meet that, meet that moment in the market right now. I think the next five to 10 years are going to be really, really good for people. I'm biased obviously, but the platform is getting richer, bigger and creating more opportunity for all of us.
Karen asked what has been your favorite stage of growing HubSpot and why? And then I guess taking sort of a metaphor from you, her choices she gives you are conception, infant, adolescent, young adult.
I'll give you the sense, the philosophical answer and the true answer happen to be the same, which is I love the stage that I'm in right now and this is a truthful answer, which is I'm having more fun right now than I've had in the history of HubSpot and the reason is that I know more things, we can make bigger bets now, we can plan out and I love the startup years, don't get me wrong. Just like I love, you know, my son when he was an infant, but I love him now too. Now he's got a brain, I can have a conversation with him. It's a different type of relationship, right? And so as HubSpot has evolved, I get more gratification out of it. Not just because the stakes are higher, but we can make longer term choices and we see the ecosystem thrive, we see the opportunities ahead of us. When you're a startup, you're sort of just trying to survive, right? That's the thing. It's like, how do we make this a business that is still a business a week, a month, a quarter from now? We don't have to worry about that, obviously. We still have to worry about lots of things and staying relevant. But it's nice to be able to take a longer range view and be able to have conversations like these where like, oh, stay tuned, there's going to be more objects and more APIs and more custom UI extensions. There's all this stuff coming and it is. I know for a fact.
Love that. Well, we are just about a time. Dharmesh. Thanks so much for being here. The last thing I want to ask you is if any of the 246 people here want to follow you, be in touch with you, connect with you out on the Internet, where are the best place or places to find you?
I'm practically unavoidable on the Internet. I'm armesh everywhere on LinkedIn. On Twitter x. I will never just completely call it X. Oh, by the way, we should call it X. Twitter. I heard that. I didn't make that up. It's like the formerly known as Twitter.
Anyway, yeah, no, that's really good.
Clever. But anyway, yeah, LinkedIn is probably the best way to kind of get in touch with me. I'm posting there all the time, so just leave a comment.
Amazing. Well, thanks so much for being here. Thanks to all of you for participating in the chat and asking questions. This was the 100th meetup of the HubSpot admin hug.
Thank you.
Hope to see you all around for another hundred. Enjoy your Tuesday.
Sam.
