Brian Balfour on creating meaningful growth

As VP of Growth at HubSpot, Brian oversees the growth teams for new products such as Sidekick and HubSpot CRM. You may also recognize his byline for articles about growth strategy, user acquisition, process, and teams from his popular blog, Coelevate.

I caught up with Brian recently to talk about the differences between growth, marketing, and sales; the dangers of chasing vanity metrics; and how to create a foundation for meaningful, long-term growth.

If you like what you hear, check out more episodes. Or subscribe to the Inside Intercom Podcast on iTunes, Stitcher, or over at Soundcloud.

If you’d prefer to read Brian’s insights, what follows is a lightly edited transcript of our conversation.

What does a growth team actually do?

Des Traynor: Thank you for joining us. For the sake of our listeners, could you define how you see a growth team within a company? Specifically, how it’s distinct from the concerns of the product team or the marketing team?

Brian Balfour: It’s interesting you asked this as the first question. I’ll be the first to admit that a well-defined answer is still developing in our industry because if you look around everyone has a slightly different definition. Different growth teams work on different things, they’re organized differently, and they have different skill sets, so there isn’t really a one-size-fits-all answer. Nor should there be one.

Before I give my opinion, let me take a step back. We need to think about what has changed in software and technology and why “growth” has become a thing. I think there are three key things:

The first is within software. The lines between where marketing ends and product begins have really started to blur, especially with the rise of platforms like Facebook, Twitter, and Pinterest, and the amount those channels help you with acquisition as well as integrate into your product.

The second is that “marketing” has become way more technical and data-driven as software in the internet ecosystem has developed. I think to continue to succeed, we’re going to have to keep pushing down that technical and quantitative path for a lot of marketing efforts.

The third is that the scale and speed of growth are accelerating. The time from 0 to 1 million users in the B2C space has massively accelerated. Even the time in the B2B space to go from 0 to 1 million ARR, has massively accelerated.

I think if people agree with those three things then it leads to a lot of questions about whether the process, people, team, and org structure that we’ve used historically are still the right model. Or does it need to look a little bit different? It also helps explain a lot of the friction most organizations experience. The most notable one is the friction between marketing and engineering, where marketing’s always begging engineering for support to get things done and it’s really hard.

I’ve always tried to say, “Growth is not about the terminology or the tactics, it’s about a change in our mentality, process, and team.”

All that being said, here’s where my current definition fits. There are three things that, ultimately, you need to grow:

The first thing you really need to build is core product value. You need to solve a real problem and build an amazing product. Product clearly owns this piece of the pie. Growth teams should not be working on building that core value.

The second early thing you need to grow is the percentage of your target audience to experience that core value as quickly as possible. I say growth owns most of this number. The pieces I wouldn’t put in that bucket are anything that would be considered the soft elements of marketing. Branding, PR, buzz. Those things I consider in marketing, outside of growth.

The third piece we really need to grow is getting customers to experience that core value as often as possible. And this is where I think it’s a mix between growth and product, and where the lines are especially unclear.

What a growth team looks like

Brian: How this plays out in an exact team structure is going to vary widely depending on your business model, your customer acquisition channels, your company culture and overall strategy. A few ways it’s played out today is that, you see in B2C companies like Facebook and Uber, growth will have full ownership over an entire product where they’re reading to throw more fuel into the fire. The latest example is the growth team at Facebook taking over Messenger. The core product team got it to product-market fit. They had really developed some core value and they really need to figure out how to grow it. The growth team got in there and started to grow that thing like crazy.

The second way this plays out is that you have growth as a service model where they almost act as consultants. Or you they could like an independent SWAT team that looks at where the biggest friction is in the growth model and then jump in and apply their process and skill sets to solving that problem.

The third model that you often see is kind of a hybrid. You’ll see the growth team own the quantitative and tech pieces of marketing, such as high scale SEO, email push, stuff like that. As well as major pieces of the product that heavily touch growth, such as onboarding or new user experience or referral or virality.

I always try to come back to the fundamentals of those three things: building the core product value, getting the target audience to experience that core value as quickly as possible, and getting that audience to experience that core value as often as possible. And layering that with where you’re at as a company and customer acquisition channels, and culture, and strategy, to figure out exactly what the right model of growth is for your company.

Des: Right. Do you think people can occasionally adapt this to wherever their problems are? Like if your onboarding sucks, that’s a great area for the growth team to start, whereas if your onboarding is pretty good, but your engagement or stickiness is pretty bad, do you think it ends up being the company will ultimately deploy the growth team to work on wherever the biggest challenges/low hanging fruit is?

Brian: When you get really developed with a large team like AirBnB or an Uber, you have full time teams dedicated to teach one of these areas. There is a full time onboarding and new user experience team.

At that size and scale it makes sense. I think in the earlier days, what you really find is that you can’t have dedicated teams for all that stuff, so what you end up with is this SWAT team model where they do look for the biggest friction in the funnel and they’re deployed to that area.

There’s pros and cons to all of these situations and the biggest con is that you end up sometimes crossing ownership lines. You work on pieces that other teams feel like they “own”, and you’ve got to facilitate that conversation and how that works, culturally, within your company. The pro part of this is that the growth team can remain, sort of, independent. They don’t take on things that acquire technical debt.

They can run a really rapid experimentation process. There’s pros and cons to all of these models, for sure, but like I said, it really depends on where you’re at, as a company in terms of a stage and strategy.

What a growth team should focus on

Des: A lot of the most popular growth stories, the ones you read about, are the sort of famous ways that big companies get their first users. They all seem to focus on a particular hack, trick, or tactic. Is that what growth teams should spend their time finding?

Brian: I think, unfortunately, way too much time is spent writing and reading and thinking about those things. I often phrase this as I often get a lot of emails saying like, “What is the one little tactic that I should do to grow my business?”

My response to them always is, one, I don’t know, because I don’t know anything about your audience or your product or your business model, or any of that stuff, so for me to try to prescribe a single tactic to you is impossible. Two, you need a way to discover those things on your own. What I really try to focus teams on, that I work with, is on the process.

What is your process and mentality around taking ideas, prioritizing them efficiently, testing them efficiently, learning as much as you can from them, and then feeding the learning back into the process of your prioritization? Because if you really focus on those fundamentals, and getting really good at that process, it will guide you to a collection of tactics that start informing and forming your strategy over the long term.

I really try to focus people on saying, “Don’t focus on all of these tactics out there. You’re just going to end up, basically, thrashing in all of these different directions, and really focus on that process of building these things from the ground up and validating them for you, and you, specifically.”

Des: Aside from tactic chasing or trying to find whatever your equivalent of the mythical Dropbox invite a friend thing or whatever, are there other obvious mistakes that you see growth teams making?

Brian: Yeah. One of the biggest ones is the shotgun scatter shot approach. What I see a lot is people who don’t know what will work, and therefore decide, “I’m going to try everything at once.”

That’s a recipe for failure. You’re just not going to get anywhere with that. A big recommendation for is to spend some time in, what I call, a zoom out phase. Looking at your growth model, a mixture of quantitative and qualitative data, as well as mixing in your own insight, and really figuring out what is the highest impact area that you can focus on right now, given limited resources.

HubSpot is a public company now with a billion and a half market cap. A thousand people in the company. Even then, my team has limited time, people, and money. That constraint never goes away. You’ve got to get really, really good at choosing the highest impact area you can focus on right now and really nail it, really dig deep because to make really meaningful progress takes focus and depth.

One thing I always say is that in a lot of cases, the order of operations does matter. Especially in highly competitive startup spaces.

Des: Do you mean the specific order in which you pick your projects?

Brian: Yeah. It’s one of these things–you feel like you need to scale up acquisition in a given channel because one of your competitors seems to be acquiring a bunch of channels, but you might have a bunch of retention problems. You probably only have enough resources to work on retention or scaling up acquisition. You scale up acquisition before you fix that retention problem.

You’re leading yourself to nowhere anyways. You might not know. That competitor might not have great retention either, so ignore all that kind of stuff and really focus on where you can have the highest impact in your model today. Think about it not just in the short term, but in the long term.

When my team looks at growth models, we drag that model out a year or two in advance, to see whether or not if we made this change to this piece of the funnel in the next 3 months, would have a relative effect over the next year to 2 compared to all of the other things on our list. Because certain things do have bigger impact over a longer period of time than they do in the short term. You’ve got to evaluate that stuff. It’s cliché to say this, we’re all in this marathon of sprints, series of sprints. It’s not a single sprint that we’re going after.

Des: It’s an interesting point that there’s obvious longitudinal, multiplicative data play. Doubling your conversion rate for next month seems great, but over a 2 year time plan, doing that versus knocking 20% off churn, it might not be the smart thing to do, or certainly not the smart order in which to do them. It’s a non-obvious point. Most people, come up with their project plan. They think of what we can do or how long each thing will take, but I don’t think they’re actually thinking about the benefits of time sampling each of the returns that they’ll get.

Brian: Yeah. Growth isn’t a science, there’s some art in there. You can never look at a perfect quantitative model and have the answer staring at you or anything. You really got to evaluate these priorities from multiple perspectives because they all aren’t created equally. To your original question, what are other mistakes? I find that people heavily underestimate the depreciation of a particular tactic or area.

We might, like you said, focus on doubling conversion rates on some landing pages. I guarantee you, those tactics fatigue over time. Every tactic has a life cycle that fatigues over time and some things have shorter life cycles than others. The deeper in the funnel you go, into things like retention, the longer life cycle it will have. You’ve got to take that into account and the only way you can take that into account is how do you look at this thing over the course of a longer time horizon than what’s staring at you in the face in the next couple months.

Avoid vanity metrics

Des: An interesting thing that occurs to me is that, you say if you go deeper into the funnel, the tactics tend to decay much slower. I think a lot of that comes back to what’s meaningful work and what’s more superficial or surface level work.

Is there danger in chasing, what I would consider meaningless growth? The sort of people who are chasing vanity statistics or stuff that makes for a good news story.

Brian: I think so. Not to pick on them, but the hot story recently is Homejoy. They were touting these massive top of funnel growth curves in the press and stuff for the longest time, and then out of nowhere, they close up shop. Forty million invested. A lot of speculation as to why they weren’t retaining customers. Even though that top level growth number looked really good for a while, underneath the surface, it wasn’t true, authentic growth because these guys weren’t sticking around. There’s this really viscous cycle in our industry, and I call it the wheel of meaningless growth, which is, early in our life cycles, we do whatever we can to get press, so we tend to pick the biggest number we can out of our business, but that number tends to be a vanity metric. Whether it’s views, or downloads, or registrations, or something else like that.

You put it in the press. You’re celebrating it externally and you get caught in this trap. Every time you want to do that, you have to come up with a bigger and bigger number, so all of the sudden, you find yourself focusing on this meaningless number. You follow it into a real trap.

A company that’s done amazing at avoiding this is actually Facebook. Since the earliest days, they’ve almost reported on active users. That definition of active users not only included how many new people they were acquiring, but how many people they were retaining, as well, over time. It can be really dangerous if you set the wrong metric for people to go after. You can end up in a situation you don’t want to be in.

Des: It’s easy to see how that happens, as well. I think lots of young companies don’t want to give away any meaningful data because they don’t want to send early signals to competitors or investors or whoever. But almost by definition, if they don’t give away meaningful data, what they’re going to give away is meaningless data. Then, they’re going to be held to them and the next thing you know, they find themselves following these metrics down a chasm of, “Well we need to boost number of app downloads,” or whatever.

They can’t just stop reporting on it because then it’ll sound like, even that’s a news story. “Oh, they’re not even reporting how many active ones they’ve got anymore. It must be really bad.” It’s kind of funny.

Where to find meaningful growth ideas

Des:When you talk about going for meaningful growth, you were speaking earlier about picking out different ideas or different projects or whatever. What’s the best source for coming up with these projects? Is it the data? Is it intuition? Or is is it look? What is it?

Brian: I don’t think it’s one thing. We draw ideas from a lot of different places and we have set exercises on our teams that we use to generate those ideas, but at the end of the day, how those ideas come about or how they’re informed is blends a few different things. It’s quantitative information, qualitative information, and a little bit of our own insight. You always have to blend those three things because if you just look at one, they don’t tell you the whole story.

Quantitative information is really good at telling us what’s going on. It only tells us a little bit of the why, and that’s where qualitative information comes in. You can really get a little bit better on the why from talking to users and gather other qualitative information, but even those 2 things combined don’t tell you the whole story. The cliché saying is customers aren’t going to tell you exactly what feature you need to build. You need to interpret it using your own insight and your own vision. I think that’s very much true.

A good example of this is, we were working on our onboarding and activation rate for a tool called Sidekick. It’s this email tool where you install a browser extension or an Outlook plug-in and it adds all of these features to your email that those clients don’t already have.

We were working on first week retention. We gathered all of this quantitative data to look at all the steps and what can influence that first week retention. We found this big group of users that were only using the product with one email and then quit.

Those were the users that were churning and that didn’t make sense to us. Then, we switched to the qualitative information. We pulled a list of those users and we sent them one on one emails asking them why they decided not to use product and we got a bunch of responses back, categorized them, looked at the language, looked at how they were describing the problem, and a big thing they were saying is, “I just didn’t have time to figure it out.” That was really confusing to us because once you have this thing installed in your email, you just have to start sending emails. There’s nothing to figure out.

That’s when we moved to just drawing on our own insight and what we realized was that on our landing pages, we talked about an email product. In our onboarding, we had you install this email product, and then the first page that you landed on wasn’t in your email. It was actually within our web application. There was this huge disconnect and it was why they were saying they don’t have time to figure this out because they thought they had to figure out the web app, not just go to their email.

We designed a really quick test and at the end of the onboarding process, we wouldn’t even let them in the web app. We, basically, gave them a screen that said, “Go to your email and start sending emails,” and it worked. We would have never figured that out if we wouldn’t have combined all three of those pieces of information.

Des: I see. That makes sense. It speaks a little to your earlier comment about it being not just a science, but also an art, too, right?

Brian: Yeah, and that’s hard because when you think about hiring for these teams, people tend to really fall on one end of the spectrum or the other. It’s very rare to find people who are really good at both, so you really have to make sure you have that blending of creativeness, as well as a very firm understanding look at quantitative and qualitative information.

Don’t just follow the outliers

Des: We’ve talked about how other people’s tactics don’t necessarily work for you and maybe chasing down one particular tactic is maybe not the right approach. One thing I’m curious about is, when you see a company that’s growing insanely fast. Today, for example, it’s like a Zenefits or a Slack. What can you learn from other fast growing companies?

Brian: On the outside, not much. We fall in love with the outliers. A lot of the time, the outliers are not the best people to learn from. Everything requires a little bit of luck, but maybe they required more luck and in this answer they just hit at the right time. I really like to look at the people who had to fight their way through it to get to success because you can ask them a lot of questions about how they went about identifying the problems, coming up solutions to the problems, implementing those problems. It’s all this stuff that leads up to that inflection in the growth curve. All those little things that they did and learned, and how they learned them, where I think you can gain a lot of learnings, and all of the things they did afterwards to sustain that momentum.

There’s always outliers out there that just hit something at the right time with the right chord and they just take off. I’m not saying all of these outliers are just due to pure luck. A lot of it is due to other things. It’s just really hard to learn about those things from the outside, from their blog, or just looking at their product, or just the press. If you are going to try to learn something from them, you got to dig deep. You got to talk to people at that team and really focus on the underlying fundamentals in the things that they went through and learned to achieve that growth. Not just trying to figure out what is that one thing they did at that inflection point of their growth curve.

Des: Right. That’s a really interesting and pretty fresh take. I think it’s also true that when a company is growing pretty fast, it’s really easy to mis-attribute. There’s 2 things people do wrong. I think they misread success, they’re like, “Oh they’re popular because they did X,” and then the other thing they do is like, “Oh, we should copy their support, or their marketing, or their sales,” or whatever because they’re successful when you don’t realize the 2 things might be pretty much decoupled internally.

Brian: Yeah. That’s why I like a lot of the content that you guys are putting out. I like the stuff you put out on how your product team operates. It’s much more about those underlying things that educate that process and methodology for you guys. It’s all the inputs rather than the outputs.

Des: Right. Yeah. That’s an interesting philosophy. To think about things you actually can learn from other companies. It is analyzing the ingredients much more than the final out. The guts of the idea, whatever it is, workplace chat or whether its communications platform for web business or whatever. That bit bring its own standards for growth trajectory and virality and all these things to the party, but it’s way more interesting to look at the things that are layered on top of it rather than to post-rationalize the success of the idea itself on to all the tactics that follow.

Brian: Yeah. Absolutely.

How to start a growth team

Des: For our audience, who are a lot of growth stage and some early stage companies, who have yet to build out a growth team, what’s a good first step towards doing this? Do you put one person on it? Do you make it a part of the product team’s responsibility? How do you get started?

Brian: There’s a few recommendations I have. Go into it with the mentality that this going to be a team, not a person. Solving these types of growth problems really requires a mixture of engineering, product, design, and marketing skill sets. It’s almost impossible to find that all in one person, there’s only so much that one person can do. I see a lot of companies setting their growth efforts up for failure because they read all of this stuff and it’s like, “Ah, I should be doing this. You, Jack Miller, over there, you’re now our designated growth hacker. Go growth hack.” That’s just a recipe for failure. I’m not saying that you’ve got to have a team of 4 or 5. It could be, initially, a team of 2 to 3, as long as you’ve got all those skill sets represented in some way, shape, or form in that little small team.

So, the first thing is take a team mindset to this. I think, the second thing, looping back to the first question, is there’s very many flavors of this and so you have to think about things like projects and ownership lines, and team and org structure. Is growth going to be its own thing or is it a part of product? Some of these key questions.

My recommendation is, no matter what, a couple of things. One is this has to have top-level support, CEO support, because to implement this, it’s a new methodology, it’s a new process and there’s going to be inevitable friction in things that you guys need to figure out, as an org, and it’s too easy if you hit that friction for people to throw their hands up and be like, “Ah, we’re going back to what we know,” the old way. You really need that mandated support.

Two, you need to educate across all of the executives. If you’re going to bring in a growth leader that reports to the CEO, you need education and some buy-in from the VP of product, and the VP of marketing, and the VP of engineering, and open that dialogue of how we’re thinking of developing this over time. Once again, like all those things, ownership lines and a few other things.

Three, find ways to get early wins. Going back to that low hanging fruit conversation is that if there is any sort of skepticism across the model or this team format or whatever, spend some time figuring out where the highest impact is, given the skill sets and resources you have. Get some really early wins and show those wins through data. It’s very hard to argue with data and you want to spread those learnings and those results across the organization because that will help things get even better as the growth team grows and starts to take on more things.

Those are probably the top three recommendations I have on a long list.

When to build a growth team

Des: Do you think there’s a right time? Is it before product-market fit? Is it after product-market fit? Is there a specific time when you think it’s the right time to onboard it or does it make sense all the way through?

Brian: I think the perfect time is you’ve got some really solid data around product-market fit and, to me the most solid data point you can have is getting a somewhat meaningful amount of users or customers that look like they’re retaining very well. Retention is the key there. The perfect time is right after there’s definitely data around clear product-market fit and you need to start formalizing the growth model and the growth strategy.

That’s really the perfect time. Any time before that, you’ll probably, in a small enough company, where everybody needs to think about growth to some perspective. Or they need to think about this balance between traction and proving out that product-market fit. You can’t just build a product and to come to product-market fit, you need to have a certain flow of users and a certain amount of growth to get that data around product-market fit.

Typically, at those points, you’ve got a sense where everybody is responsible for growth in some perspective, but even then, you probably still want one person who’s thinking about it full time. They’re trying to understand it, and doing so before finding a product-market fit is not the right time to bring in an entire team. It’s definitely after product-market fit and, ideally, in that stage, like I talked about, if you get too far past that stage, I think it gets really hard to implement a growth team, in a very mature company. There’s too much culture, the machine and the play books are so developed and so built that it’s very hard to implement this a whole new methodology and process and stuff into that mix. That ideal time, once again, is probably right after some clear product-market fit data at which you’re starting to formalize the growth model.

In funding terms for startups out there, that’s typically around the Series B point. End of Series A, beginning of Series B. That’s typically where that’s at, but it really depends.

Des: Well, this has been a really fascinating conversation. I’d love to keep going, but we try to keep these things relatively short.

How can our listeners keep in contact with you? I know you’ve published a lot.

Brian: Yeah, so pretty much all my writing is on my site at coelevate.com or you can just Google my name and that site will come up. Another awesome other growth person, Andrew Chen, and I are co-hosting an 8-week lecture series. If you’re interesting in going deep into metrics, marketing, and growth, we’d love to see you sign up.

Des: Excellent. Thanks so much for your time today, Brian.

Brian: Thank you.