How to Build a Successful Growth Team with Jeff Chang @Pinterest

Dec 4, 2018 5:31:55 PM
Author: Kieran Flanagan

What we cover in Episode 21

Pinterest is an incredible success story. Founded in 2010, they now have well over 250 million monthly active users.

Pinterest is also known for having an excellent growth team who have documented a lot of their work online.

In this episode, we catch up with Jeff Chang, who is a tech lead on their growth engineering team to talk about how you set up a growth team for success.


Enjoy the show, and happy growing.

1. Autonomy is the best way to scale experiment ideas

Ideation is an important subject at Pinterest. They work to empower individuals to not only come up with different ideas for experiments but to also be responsible for taking that idea from concept to execution.

There are many reasons why Pinterest approach ideation in this way:

- Pinterest has a large growth team, with 50 to 75 engineers. With a team that large, it's difficult for one person, or a couple of people, to come up with the best ideas.

- You get a better diversity of ideas

"Different people have different points of data and might be able to see things that others might not."

- You get a better quality of ideas because the person who comes up with the idea is the expert in that area, so they're the best person to turn that idea into a real growth initiative.

Pinterest encourages everyone to come up with their ideas, and they review all of them every two to three weeks.

"Every two to three weeks we have a meeting called 'Experiment Idea Review.' Everyone attends that meeting, engineers, PM's designers. Each person who is submitting an idea fills out the same template based upon the hype framework."

"They write down their hypothesis, the investments they'll need, the precedent, e.g. have we done anything like this before, and in return, they'll get feedback on it from the community who says yes or no if the idea is worth pursuing."

If the idea gets a thumbs down from the community, that person is still free to pursue their idea. Pinterest cares about autonomy.

It's a great way to teach people about the importance of decision making and how to be accountable for the choices you make.

It's also a great learning experience. Being wrong about something teaches you a lot more than being right.


2. A good idea is one that removes friction or increases user intent

Most ideas in Pinterest fall into two buckets:

"They either make something easier to do, or they make the user want to do more of something."

A good example of removing friction is:

"I might want to reduce the number of steps it takes to signup for our product. One way of doing this would be to add a social login feature, so you have a one-click sign-in. Also, I could provide help for people trying to signup, and do things like automatically fix typos. Each of these will remove friction from the signup process."

To encourage users to do more of something, I can better educate them on the value of my product.

I can demonstrate why they should use it, what features that will be of most benefit to them, and how it can help improve their daily lives.

3. Great ideas that require a lot of ongoing maintenance might not be worth doing.

When deciding on what ideas are worth dedicating time and resources to, you need to consider what the maintenance of change is?

If you build something, how much work is required to maintain it continually?

"We've had many examples where we didn't pursue ideas because the cost to maintain them would be too high. For example, we saw a lot of success from implementing our social authentication methods, like Google and Facebook."

"We can probably get a lot of value by allowing you to signup through platforms like WhatsApp, or Twitter."

"The initial cost of building them isn't that high. However, the cost of maintaining all of those different signup paths is very costly. You would get a bunch of additional account issues by adding new ways to signup."

A lot of people only consider the upfront cost of an idea and not how expensive it would be long-term to maintain.

Adding authentication methods is a small upfront cost, but an expensive long-term price.

4. Create an idea portfolio with a range of different risks associated with them

Growth teams are goal and metrics focused teams.

But how do you attach a goal or metric to something if you have no historical context for it? What if the thing you want to do is brand new?

"For new initiatives, I would recommend deferring goal setting until you've had a few results. The best way to tackle a brand new initiative isn't by making a grand plan that covers the next 6 to 12 months, it's to run a few experiments and gather data."

"Once you've gathered some data via experiments, it's going to become a lot easier to come up with a reasonable goal."

Those experiments will be run as minimal viable tests. That allows you to use a minimal amount of resources to gather enough data to make good decisions on what initiatives are worth pursuing.

But what happens when the minimal viable version is something that requires you to invest a lot of time and resources upfront? In these cases, there is a lot more risk involved as there is a greater cost of gathering enough data to prove something is worth further investment.

"If you have something that is a more high-risk experiment, you want to hedge against that risk with a lot of smaller experiments to decrease your exposure.
You want to have a mix, where you're investing in some big initiatives and a lot of smaller initiatives where you're more confident of success."

"In that case, at the very worst if the big initiative doesn't work, we'll still have success from other things.That's the best way to manage risk, with diversification."

"That's how we decrease the cost of MVTs as much as possible, and the max will spend on an MVT is 3 months."

5. A team should have a goal that they can be fully accountable for

When setting goals you want to ensure the team can have clear accountability for it. That it's both relevant and achievable.

The worst goals to set a team are when they can either miss or hit their goal through the actions of others.

For example, if an SEO team is goal'd on traffic, and a CRO team is goal'd on conversions, the CRO team can hit or miss its goal dependent upon how much traffic the SEO team generates.

Another common example in growth is where one team may own user signups, and a separate team may own monetization and retention of those users.

"There are some achievability and relevance issues with weekly signups as a goal. What if someone signs up and doesn't activate or if they sign up and don't retain?Those sign-ups are probably not valuable for the business.

A better goal might be weekly sign-ups that retain or weekly sign-ups that activate."

The podcast provides a more in-depth look at these topics, so if you enjoyed reading the above, please do give it a listen.

And until next time,

Happy Growing!

Subscribe Rate and Review

>> Apple Podcast

>> Stitcher

>> Google Play

>> Spotify

Follow us on Social Media

>> @Searchbrat

>> @SJTousley




Topics: Growth, Podcast, Leader

Join the Newsletter

Get my thoughts on repeatable playbooks for marketing, growth, hiring, and leadership.