Building For The End User With HubSpot’s Christopher “Millsy” Miller
April 25, 2023
I can’t tell you the number of times founders ask me how their companies can become product-led. I never know where to start because the truth is: there are so many ways. Enough ways that I’ve launched a Mastermind Community for select founders to explore this very question (and all other things PLG). To get the party started, I’m interviewing experts once a month about OpenView’s 11 Principles for the New Era of PLG.
We recently covered Principal #1: Build for the end user. Christopher “Millsy” Miller, VP of Product at HubSpot, knows a thing or two about PLG—touting over a decade of experience in the product space. Millsy led monetization at Runkeeper through the Asics acquisition, and had a brief stint in e-commerce before landing at HubSpot. He is currently overseeing their product-led growth strategy and self-service business.
I chatted with Millsy about growing products and businesses by solving for the end user. I’ve done my best to summarize our Q&A below, highlighting the parts that most landed with me.
1. What does “Build for the End User” mean to HubSpot?
Building for the end user is core to Millsy’s philosophy at HubSpot. Of all the activities that can fill the top of the funnel, solving for the end user is paramount. It’s pretty simple: if you fix problems for your user in unique and clever ways, you’ll kickstart growth. Millsy puts it best:
“Growth is a downstream effect. It’s like a lagging indicator. It’s a residual effect of doing things right upstream, which is just solving problems for the people who are using your product. If you do that, well, chances are they’ll pay for it. They’ll not only pay for it, but they will evangelize your product through word-of-mouth growth, virality.”
2. How does HubSpot think about end users versus buyers?
Ideally, your end user is your buyer but we all know that’s not always the case. At Millsy’s first product job, a B2B2C company, his end users and buyers didn’t even work for the same business.
The company sold products to issuing banks and merchant service providers (the customers) who white-labeled and resold them to small businesses and merchants (the end users).
“You could probably imagine what this led to. It led to us building to spec from a person who was never going to use that product a day in their life. They didn’t need that product in order to do their job. The people who used the product at the end of the day were very ‘meh’ about it,” he said.
Fast forward to Millsy’s current work at HubSpot, where he says “the number one mantra internally is ‘solve for the customer.’” Since it’s SMB and mid-market (not enterprise) focused, there aren’t as many stakeholders involved. As you might expect, HubSpot’s buyers (sales managers) are usually its end users.
“Our mantra is if we can make the end user fall in love with our product, even if it’s a top-down buying decision, it’s a bottoms-up adoption motion, and that’s the key to success for us. That’s the lens through which we view our product strategy and our growth strategy.”
3. To build for the end user, do you recommend hiring a generalist growth team or one with domain expertise?
It turns out generalists are definitely the way to go. Millsy didn’t know anything about CRMs when he joined HubSpot but was able to quickly learn through osmosis. He values communication, organization, tempo, and grit over subject matter expertise. His ideal initial team would be made up of a few “swiss army knives,” such as:
- A full-stack developer
- A product manager to tackle wireframing, write SQL, and manage the analytics dashboard
- A marketer to align product motions with go-to-market strategy
“Oftentimes, former marketers make great growth PMs. I think out of probably 20 PMs in my org now, 25% of them were former marketers at HubSpot who I plucked out of marketing because they were pretty technical,” says Millsy. “They knew the product really well and they were very fluent in data. I’m like, ‘You would be an amazing growth PM. Come on over here, come on to the dark side.’”
4. How do you build for end users without a big team?
Millsy said having a small team can actually benefit you because it keeps you experimenting. When your team increases, your install base and data set also increase. This buildup can lead to conservative, data-driven decisions that stall growth.
Your competitive advantage is not the size of your team but the closeness to your customer, according to Millsy.
“The more you know about the customer through face-to-face interactions, hours and hours of conversation, that’s the thing that’s going to actually fuel a perspective, a vision,” he says.
For him, that customer knowledge lets you start building. “Use your product intuition and use your gut feeling,” he says.”You have to have a strong product vision, and that’s something you can have independent of the size of your team, independent of the size of your install base, independent of the size of your data set.”
Here’s a simple visual Millsy likes to use:
- Risk: the likelihood an assumption will put your company out of business or get you fired
- Unknown: the knowledge we lack about a particular problem from previous experience
Low risk, low unknown: Don’t bother testing, just go for it! Even if you’re wrong, your assumption probably won’t harm your broader strategy.
High risk, high unknown: Don’t abandon the assumption; just spend a lot of time validating it to avoid endangering your broader strategy.
5. What are the biggest mistakes early-stage companies make when building for end users?
Tons of things! Here are the most common issues Millsy sees:
- Having too high expectations for a head of growth: “It’s hiring a head of growth, but having no resources and expecting this person to pull a rabbit out of their hat. That’s not going to happen.”
- Growth hacking: “That’s like a dirty word on my team, I don’t let anybody use that term because I think it’s a little silly, but I also think it kind of incentivizes bad behaviors. We’re not really trying to hack anything. We’re not trying to introduce dark patterns and pull the wool over our end users’ eyes. We’re just trying to solve problems. There’s nothing hacky about it; it’s good old-fashioned product practitioner work.”
- Operating without a strong conviction: “Not really having a rhyme or reason or strong conviction about why [something] would work [and] just throwing a bunch of guesses against the wall and seeing if you can just scrape a point to drop off out of your funnel. I think that’s a big pitfall.”
- Copying another company’s playbook: A lot of folks try to follow HubSpot’s acquisition strategy around content, for example. “I’m like, ‘It’s really, really hard to do that if you don’t have a huge war chest. You might have to think about it differently depending on the stage of your company.’”
- Sprinting towards experimentation before finishing data hygiene: “Things not being instrumented, having no operating system for how to document learnings, no paper trail and just throwing things against the wall. Cool, it gives you the illusion of moving really fast. It’s locomotion, but it’s not necessarily productive all the time. Start with fundamentals. I think people often try to run before they can crawl or walk.”
6. How can a company become product-led if it doesn’t have enough visibility into data and can’t track usage?
In the absence of quantitative data at scale, just lean really heavily into qualitative feedback. Talk to your customers, 10 if you can, but even by five or six, you’ll be able to hear some signals that can lead you to “just spot trends,” says Millsy.
“Obviously, that’s not going to be super hygienic in terms of being able to prove causation with changes. It’s not going to be super scientific, but it’s going to probably [give] pretty strong signals that you can hang your hat on and use your gut.”
In a previous podcast episode with OV, Millsy shared with us the 3Ds Framework they use at HubSpot to classify and solve user problems:
|A user can’t find an action because it’s not visible enough.
“Can users see it with their physical eyeballs?”
|E.g., Users didn’t see an action item tucked under a menu.
|A user doesn’t want to take an action because they’re not ready or interested.
|E.g., Users don’t take an action because they don’t yet have authorization from their managers.
|A user can’t complete an action because “it’s just too damn hard.”
|E.g., Users can’t get their payment information to load properly.
Discoverability: “You can uncover that through interviews and just talking to people. That’s an easy fix. You just pull it out or you work with your designer to figure out how to make discoverability better for that sort of core action.”
Desirability: “You don’t necessarily need a gazillion analytics events firing for you to be able to answer that question…If the problem’s desirability, you’ll probably get a sense of the solution through those conversations too.”
Doability: “In those instances, you maybe use technology to try to solve it. Maybe you use more clever design… Doability issues are kind of the meat and potatoes of: ‘We need to build this thing differently to make it easier for people to do it.’”
7. How do you know a sales-led company should transition to PLG?
“Oftentimes I think folks can go down the path of assuming that PLG is all or nothing. Either your entire go-to-market is product-led or it sucks or it’s not going to work, and I think that’s a pretty myopic way to look at it,” says Millsy. “The way that we kind of look at it is that you can take a PLG approach across a bunch of parts of your funnel. It doesn’t need to be all or nothing. It can be piecemeal.”
Since PLG can exist on a spectrum, there’s never a perfect time to instate it. Instead, determine where you want to gain leverage in your funnel and invest in those motions.
The areas where you might need PLG-driven leverage are:
- Acquisition: e.g., you’re not seeing organic traffic through traditional lead gen but don’t want to pay for advertising
- Activation: e.g., it’s expensive to get users started and you don’t want to pay for customer onboarding specialists when your company’s cranking
- Monetization: e.g., you don’t want to hire sales reps and pay commissions (even if you’ll have lower average selling prices)
“I would say just be really tactical, be specific about the reasons you’re doing it and be able to tie it to outcomes that are not just more broadly, ‘we want the business to grow.’ Where do you want the business to grow and how? Then, invest your energy in those motions.”
8. When should you include the sales team in product-led “land and expand” customer acquisition?
Data-driven segmentation should dictate whether to rely on PLG or traditional sales. Involve sales if it will create sizable LTV versus CAC; rely on PLG if not. Millsy points to HubSpot’s evolution from sales-led to “being retrofitted with PLG scaffolding” as an example.
“I think a big tool for us was having a pretty robust data set that was able to allow us to point to areas where it didn’t make sense to have humans involved,” he said.
The data showed poor CAC when it came to sales-led interactions with five-person companies. So Millsy said, “We’re going to let the product do its job.”
Retention improved as this segment used HubSpot’s starter product and showed interest in pro, though. At that point, the unit economics of sending the sales team to “bang the phones” made much more sense.
“We really just let the data dictate the go-to-market motions…and didn’t try to do anything based on philosophy.”
- Growth moves downstream. If you nail fundamentals upstream, users will not just pay for your product but evangelize it.
- Courting end users is key to growth. Even if it’s a top-down buying decision, it’s a bottoms-up adoption motion
- Your competitive advantage is the closeness to your customer, not the size of your team.
- Contrary to popular belief, PLG is not all-or-nothing. Instead, it can work in a piecemeal fashion, occurring across just some of your funnel.
Next up: Principal #2
If you want to hear another PLG expert share hard-won wisdom, sign up for future sessions. Ben Williams, former VP of product at Snyk, will discuss how to “build to be discovered” next.