Maker Hunt’s AMA with John Egan from Pinterest

John Egan is currently working on the growth team at Pinterest where he’s the engineering lead for the Retention & Resurrection team (R&R) which focuses primarily emails & push notifications. At Pinterest, he’s personally spearheaded several projects that cumulatively have helped increase WAUs by a double-digit percentage. Previously, John led the growth engineering team at Shopkick, a cross-retailer mobile loyalty app funded by Kleiner-Perkins & Greylock Partners which was acquired by SK Planet for $200MM. At Shopkick, he helped grow the userbase 8x from 1MM users to 8MM users primarily through invites, geofencing, & push notifications.

If you want to know more about Pinterest and how to scale fast-growth startups, this is the AMA for you. Get your questions ready. After John introduces himself, we’ll get right to the questions.

You can introduce yourself now. The floor is yours. Then, the questions will begin.

My name is John Egan (blog: http://jwegan.com). Eric covered most of my background ☺. I’ve worked a lot on emails, push notifications, invites, and activation. AMA.

Q. Can you tell us about some of your most successful campaigns at Shopkick? — Eric Willis

For invites, when I joined SK the invite program was get 50kicks for every person you invite. You could rack up some serious amounts if you invited a lot, but no one would bother to do the math in their head. We tried probably 10–20 different incentive schemes before finally landing on get 2000 kicks (worth $8) for inviting 3 friends. This program more than doubled the number of signups we got from invites. The main reason is the 2000 kick incentive was much more appealing so we got more people inviting, the minimum requirement of 3 people joining also increased the average number of invites sent/inviter and those two effects just compounded

Q. I would love to hear about how you do “Resurrection”. We’re thinking about that, but are afraid to look spammy. How do you make sure people won’t be annoyed if you contact them while they haven’t been active for months? — Olivier

Since content is Pinterest’s main draw, our resurrection efforts focus on getting content in front of users that we think they will like. We generally keep a steady stream of emails at a rate of at least 1–2 to inactive users. However, we stop sending emails if you have not been active in 6 months to help maintain our IP reputation for email. I think if you strike the right balance in terms of the number of emails you send to less active users, you can avoid being seen as spammy

Q. How much personality do you think should be put into your customer communications? — Michael Buckbee

That really depends on your brand. At shopkick we put a good amount of personality, at Pinterest we only put personality in our editorial emails. I do think personality can help engender brand loyalty and make you stand out (especially for early stage companies), but you also need to balance it and make sure your personality doesn’t also alienate certain segments of users.

Q. Do you think email is here to stay? Does that thought come into your strategy at all? — Ben Tossell

I do think email is here to stay. I’ve read posts declaring the death of email for the past 5 years, but email is still going strong and drives a lot of activity back to Pinterest. Another thing to keep in mind, is email is now a mobile channel. Over half of our emails get read on mobile and drive traffic back to our mobile apps. Email is important because on iOS only 30%-60% of users give you the push notification permission and you need a channel to re-engage those users that don’t give you push notifications

Q. What do you look for in “growth engineers”? Why is it so difficult to hire growth engineers? — Eric Willis

Hiring growth engineers is hard, because the field has only become popular relatively recently so not a lot of engineers have prior experience. So, we don’t try and look for engineers with prior growth experience (although we jump on them when we find them). We generally look for engineers that are full-stack (this is because we want to be able to execute growth projects without dependencies on other teams), who have some product sense so they can help generate ideas for growth projects. Finally, it is also important to understand what motivates that engineer. Some engineers are motivated by working on hard technical problems or optimizations, some engineers are motivated by building great user experiences, some by impact to the company. Generally the ones motivated by company impact are the best fit for growth.

Q. How is Pinterest solving the Web-to-Mobile conversion problem? — Eric Willis

For web to mobile conversion, we have some upsells to get the mobile app on desktop web. For mobile web we have pretty aggressive upsells for the mobile app. I think for mobile first companies, they should consider building out a desktop web experience though, because you always lose a significant percentage of people when trying to make the jump from desktop web to mobile.

Q. How would you recommend we at product hunt think about growing based on what you know about us? / What questions should we be asking? — Erik Torenberg

I think you guys are doing a great job. Network effects drive a lot of traffic because people ask friends to upvote. You have pretty good emails (although you may want to think about automatically switching people over to the weekly digest instead of the daily digest if they don’t seem to interact with the email that much). I don’t know how much you guys are thinking about SEO, but that could be a good channel for you since you probably build a lot of backlinks from people linking to specific posts and you have a lot of content on the site

Q. What’s the most common growth related mistake that you see startups making (like you go to their site and wince)? — Michael Buckbee

Most common growth mistake is worrying about growth channels/etc before they’ve reached product market fit. You can’t grow a product that people don’t get value out of. You need to make sure your product is something people really need. Sean Ellis likes to survey users about how disappointed they would be if the product were to be shut down in order to gauge if they’ve reached product market fit or not.

Q. Do you have some blogs/books/people to follow that people who want to learn more about growth hacking should check out? — Eric Willis

Ya, Andrew Chen, Sean Ellis, Andy Johns, Brian Balfour all write/have written great content on growth. I also have found some of the episodes on growthhacker.tv very education and insightful

Q. What are thoughts on wearables like the Apple Watch? Is that something Pinterest is really thinking a lot about at this point? — Eric Willis

With the Apple Watch & other watches, I personally am not sure the products currently has enough value to normal everyday consumers that it will lead to mass market adoption outside of the tech/geek community. It is certainly something to keep any eye on, because every new innovation is a potential opportunity. At Pinterest we currently don’t have a watch app, but it is something we are evaluating

Q. RE: push notifications, like you mentioned, mail is also a mobile thing now, since you get mobile notification for mails. Any idea if/how we should ensure to not send “too many” notifications (like for twitter, when a fav + reply will trigger 4 notification per device, push + mail). Or is it actually a good thing? —Olivier

If you haven’t optimized your notifications they will tend to have a very bi-modal distribution. A lot will receive none and a lot of people will receive a ton. You need to find useful notifications to send to people receiving none, and put systems in place to rate limit/aggregate notifications for people receiving a lot. There are things you can do to manage the volume. For instance Twitter will badge users without sending a notification. Facebook also does not send notifications to most hyper-active users, but if you’re a less active users it will send you a couple notifications a day

Q. Are you using inhouse tools for measuring growth stats like Daily Net Change or are you using some type of SaaS product? — Eric Willis

We use in-house tools. Most companies at Pinterest’s scale do since generally you want all your data in your own data warehouse. It is hard to build general saas tools for thise because everyone has their own table schemas, they have their own custom things they need to do to the data (ex: at pinterest we have to filter out spam users from all our metrics), etc

Q. How closely do you think growth + product need to work together and influence each other? — Michael Buckbee

Product changes are hopefully helping growth since they should be improving the product to provide more value to the user. The growth team should be in sync with product so they can spot new potential growth opportunities. An example is Pinterest recently launched Messages and this feature can be used to both potentially acquire new users and resurrect dormant users.

Q. Do you have a set of questions or good formula to request valid customer feedback ? — Sébastien Barrau

I like really open ended questions. You might got a lot of 1–2 word responses that aren’t valuable that you need to filter out, but in general it is a good way to get some relatively unbiased customer feedback. Examples of ones I ask are “How do you use Pinterest” to gauge product comprehension. “What feature do you like the most” to figure out what areas user’s are getting value from


Next AMA is with:

Tamar Weinberg — Author of The New Community Rules: Marketing on the Social Web — 3pm EST 28th May


If you liked this AMA please Recommend ☺. Keep an eye out on our Twitter — @MakerHunt for details/updates.

Special thanks to John and everyone who participated. If you’re a maker with a product on ProductHunt, be sure to sign up to participate in the AMAs and connect with others.

Already a member? Get your questions ready for the next AMAs! For any questions, don’t hesitate to reach out to: Eric, Jonas, Remco, Cat or Ben.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.