Why (and how) we started a growth team at our 28-person startup
It’s a startup… shouldn’t everyone work on growth?
At the start of the summer, I stood up in front of the whole SoapBox team and told them we’d be starting a growth team. Now, four months later, I wanted to share a look at how (and why) we took on this challenge, and some of the learnings from along the way.
What is a growth team, anyway?
A growth team is all about rapid iterative experiments designed to improve your product or business. At SoapBox, we talk a lot about our product as a bucket. Ideally, marketing and sales pour lots of potential customers into the bucket, and it retains all of them. But if your bucket has a bunch of holes, you’ll be leaking out customers as quickly as they come in. Growth is all about experimenting with ways to fill the bucket and plug those holes.
Why we started a growth team at SoapBox
A startup pretty much is a growth team. (As our friends at Intercom put it, “shouldn’t everyone in a startup be working on growth?”) At such a small scale, it makes sense that everyone is connected and working together to benefit — and grow — the business.
For us, it was more about making sure it’s a cross-company thing. In the past, it would be the founders and senior leaders of the company talking about growth — and that’s it. What I wanted instead was shared accountability for the entire product delivery, across the entire company. We wanted a formalized team of product, engineering, support, marketing and other business people that we could bring together, set them up for success and hold them accountable to growing our business.
We also know that a growth mindset is not as commonplace in the Toronto startup ecosphere as it has been elsewhere. So we wanted to help inject that into the DNA of our teams, by having people do a kind of “tour of duty” with the growth team, and then take what they’ve learned back to their functional areas.
The first 3 things we did to start a growth team
1) We got our docs in place
Before you can really dive into full-funnel growth, there are a few docs you need in place.
Why? Our goal was to scale the growth mindset to an entire team. In order to do that, we needed to have shared understanding of the problem we are solving and who we are solving that problem for. We also needed a shared place to document our learnings and reference them while running our experiments.
This is something that as a founder, is just stored in your head and validated with your gut. To scale your business, you need to share this information and externalize this validation process. That means you need to start codify these processes.
First, you need to define your ideal customer profile. This is who we’re going after, why we’re going after them, what specific problems they have, what competitors might entice them away…it’s like a very light weight lean canvas.
Second, you have a buyer’s journey. The path to get them from “I’ve never heard of you” all the way to “Oh man, I love your product so much I’m telling everyone I know.”
(One note on these first two documents: they shouldn’t be set in stone. You’re essentially creating a first draft and then using your growth team to validate everything that’s in them. It’s important to come to that first meeting with a draft in hand, so you can start the growth team on right direction and not have them go off track.)
And finally, you have your metrics. We use pirate metrics, which organizes our key numbers into acquisition, activation, retention, revenue and referral. These metrics give you data to back up decision making when deciding what the Growth team should focus on.
These documents become the basis for our learnings. We create a hypothesis and run an experiment targeting a specific pirate metric, validate what we know to be true from the buyer’s journey and ideal customer profile, review the results from the experiments, and document the learnings all in a fast paced weekly cadence.
Our thinking is that, by doing this process over and over again and as fast as possible, we get better at focusing and refining our experiments, and ultimately understanding what messaging and product experiences best resonate with the customer. Even if the results of the first 4 weeks of experiments are failures, we are still learning faster than had we remained isolated in our functional areas. All this shared learning should result in success, and better results in our core metrics.
2) We set up our tools
It’s fairly simple. We need:
- A meeting room with a whiteboard and TV
- SoapBox
We knew we’d gather once a week to meet about growth. But that leads to talking, and not execution, accountability, and transferred knowledge. We knew we needed a place to collaborate, store, track and evaluate which experiments were working/failing and document our learnings. We put SoapBox into place for that purpose: a shared place to track, discuss and collaborate on ideas ahead of each team meeting. We also used it to store decisions, next steps and learnings.
3) We created the team
We asked our growth team to spend 20% of their time on these experiments (50% for engineers). That’s a big commitment. So we needed to make sure we picked the right people to join.
But what are the characteristics of a good growth team member?
- They embody the “Let The Best Idea Win” section of the SoapBox Handbook: they’re positive, they know that customer is king, they are good active listeners and they have empathy.
- They’re curious: they love to think, explore, figure things out, ask questions, experiment, learn.
- They have a hunger to build/measure/learn: We’re asking team members to run experiments each week with only 20–50% of their time? That’s hard. But that’s what’s needed to out-learn our competition.
In a small team like ours, it was relatively easy to pick the people that would join the growth team. In most cases, there’s only one person doing each type of role. In the cases where we had multiple people in similar roles, we went with the person with more experience, ensuring they had full brand knowledge.
Engineering was trickier because it’s a bigger team and I didn’t want to just pull people. There were people that I wanted to be on the team, because they’re very product-minded, always asking inquisitive questions. But I didn’t want to make it forced. So I told them I needed two volunteers — and the ones I expected to join were the ones that volunteered. So that was perfect.
How our growth team works
Each week, I look at our pirate metrics and pick a drop-off area. Asking myself: Where are we losing people? What are the biggest holes in our bucket of acquiring and retaining customers? Are people getting to our website but not clicking the “Sign up now!” button? Are people clicking the CTAs, but never finishing the sign-up process?
I pick the biggest drop-off and tell the team that this is their target area. acquisition was the focus for the first couple months. Then we moved into activation.
Once we have a focused metric the team is set loose with a few guiding principles.
First, we have a few basic rules to abide by:
- We must always take the buyer journey and customer persona into account when designing the experiments
- Everyone should contribute equally (As there are no shortage of experiments that can be run)
- We must stick to rapid learning with experiments that are small in scope — ideally, it should take no more than a week to build and deploy, and a week for the experiment to run and capture enough measurable data.
From there, the team builds out their hypotheses. All experiment pitches must:
- Have a stated goal that the experiment is trying to achieve
- Have historical data that can be used as the “before” or the “control” of the experiment
- Have some sort of explicit metric being tracked that will indicate if the experiment hit the intended goal
Our process
Ideally, a growth team meets once weekly to talk about two things:
- A retrospective on the past week’s experiments, and
- A planning session for the upcoming experiments where we the team brings their experiment ideas and pitch which ones should be built out.
In the beginning, we broke this into two meetings every Thursday: a one-hour morning session, and a one-hour afternoon session. It wasn’t ideal, but I knew we’d need more time to get into the process and thought a single two-hour meeting would be too long.
Now, a few months in, we’re streamlined enough that we can do it all in a one-hour session. That happened by everyone becoming more diligent in their reporting and succinct with their pitches.
The most important part of the meeting is planning the experiments. When a growth team member comes forward with an experiment, they present the following information in our SoapBox:
- The current state that we’re in
- Why going after this problem is going to change the business or change a key metric
- Some details of what they will actually be doing
The team talks it through, makes sure that everyone that can contribute to this experiment is part of the process. This is where I become a feedback machine. I’ll spend 30 minutes prior to the growth meeting reading over everyone’s experiment ideas in our SoapBox, and I’ll come with questions. And at the beginning, it was a lot of, “you’re missing the goal again.” Or, “you’re thinking too big-picture.”
Now, as we get more into the swing of things, it’s more like gentle reminders. What are you trying to validate? How does this affect our buyer’s journey? And so on, and so on. I’m the broken record that keeps reminding them to follow the rules and guidelines we’ve put in place.
But still, every meeting, there’s always something big that someone wants to run. And I have to say, “Hold on, why are we doing this?” And they’ll say, “Oh I want to see change in the product.” And I’ll say, No. You want to learn first. The goal is to learn what’s working, and what’s not. Once we are confident in that, the results will follow. So I’m just constantly reiterating that.
Then, the experiments gets built. They go live, we let them run for a week and then we start to look at the data. We’re looking for full success or full failures. It’s completely fine to fail, as long as you learn something from the failure. For example, if we change the language on the pricing page CTA and we see conversion rates drop by 5%, that’s a full fail. We can learn from that experiment about the language that our core customers respond to, and we can leverage that intel across the site and product to get the right kinds of customers.
The final step is disseminating these findings to the rest of the company. So far, we haven’t done anything big to facilitate this. I wanted to see if it would happened naturally. After all, we’re a 30-person startup. We all work in the same room. We use our Company wide Demo Day to share key findings, but otherwise I just let the growth team members share their findings with their teams. We’re still seeing if that works, so far it seems fine.
Who’s the leader?
I started the growth team, but I don’t think you necessarily need senior leadership leading it.
You need buy-in from upper management so that everyone in the company sees that it is important. And, as CTO, being in the room definitely helps to shape the conversation better, since I’m able to share all the key information they would need. I’m aware of what is happening with our investors, our board meetings and up-to-date on customer feedback. I’m aware of what our marketing positioning is, so I can pull all that information out for the team. Though, if we’re more diligent, that information should be made more accessible to everyone in the company and discussed at our town halls and demo days.
And even though I started as the leader, I’m no longer leading the meetings. I sit back and react to the meeting and the experiments discussed. That’s been the best change. So I would say, don’t put a person from leadership on the growth team. At least not to plan, organize or present. Their job is really just to be the feedback machine.
Ideally, you’d have a couple leaders at any given time, depending on the target you’re going after. For example, when we were focusing on acquisition, our Digital Marketing Manager was leading the meetings. And as we transitioned from acquisition to activation, our Product Manager took over. That transition is where cross-leadership will evolve.
Something I’ve been asked is whether I need to be in the meetings at all. My goal is to help everyone on the team become disciplined in their experimentation process. So, I’m not going to leave the team until they are. As soon as they are, I’m out. But for now I’m still in there nudging, nudging, and nudging.
The future of our growth team
Now that we have a good meeting cadence set, we’re going to cycle the membership of the growth team. Why? For one thing, since we are a startup and everyone should just naturally operate on a growth mindset, we’re taking our first “School of Growth” graduates and unleashing them back to their functional areas with a growth mindset in their DNA and a good understanding of how to include cross team collaboration in their day-to-day.
It can also be hard to come up with new experiments week in and week out with the same membership and I can see the team start to feel the pressure, so some fresh members with a new pirate metric to focus on (on to retention!) will help keep the pulse of growth going at SoapBox.