[WMD 2016] Advisor to Pocket, Airbnb, Darby Smart; Former Growth Lead at Pinterest & GrubHub >> Casey Winters "A lesson in retaining users"
•Download as PPTX, PDF•
2 likes•1,382 views
The document discusses improving user retention through constant product improvements, onboarding and education, emails, notifications, customer service, new platforms and products. It emphasizes focusing on reducing friction in the core product over adding new features. It provides examples from Pinterest and Grubhub of how they improved retention through localizing recommendations, lowering fees, and optimizing the multi-platform experience. The document advises testing engagement strategies like promotions carefully and designing loyalty programs to drive profits rather than costs. Overall it stresses the importance of understanding what causes users to reject a product and embracing that feedback to enhance retention.
Recommended
More Related Content
What's hot
What's hot (20)
Viewers also liked
Viewers also liked (20)
Similar to [WMD 2016] Advisor to Pocket, Airbnb, Darby Smart; Former Growth Lead at Pinterest & GrubHub >> Casey Winters "A lesson in retaining users"
Similar to [WMD 2016] Advisor to Pocket, Airbnb, Darby Smart; Former Growth Lead at Pinterest & GrubHub >> Casey Winters "A lesson in retaining users" (20)
More from 500 Startups
Recently uploaded
Recently uploaded (15)
[WMD 2016] Advisor to Pocket, Airbnb, Darby Smart; Former Growth Lead at Pinterest & GrubHub >> Casey Winters "A lesson in retaining users"
- 1. Constant Rejection: A Lesson in Retaining Users Casey Winters Growth Advisor Former Growth Lead @ Pinterest First Marketer @ Grubhub Casey Winters | @onecaseman | caseyaccidental.com
- 2. Retention is like teenage sex Everyone says they are doing it Nobody really is doing it (well) Casey Winters | @onecaseman | caseyaccidental.com
- 3. You may think you’re doing retention well “I’m sending emails…” Fuck your emails Casey Winters | @onecaseman | caseyaccidental.com
- 4. What happens if you don’t work on retention? Casey Winters | @onecaseman | caseyaccidental.com
- 5. So how do you improve retention? Learn what causes rejection Learn to love rejection (AKA feedback) Casey Winters | @onecaseman | caseyaccidental.com
- 6. So how do you improve retention? 1. Product Improvements 2. Onboarding & Education 3. Emails 4. Notifications 5. Customer Service 6. New Products 7. Promotions (but be careful!) 8. Loyalty/Engagement Programs Casey Winters | @onecaseman | caseyaccidental.com
- 7. 1. Product Improvements Retention is driven by a maniacal focus on improving the core product That is more likely to mean reducing friction in the product than adding features to it Casey Winters | @onecaseman | caseyaccidental.com
- 8. Product Case Study #1: Grubhub • Increased Restaurant variety • Lowered minimums and delivery fees • Buried guest ordering Casey Winters | @onecaseman | caseyaccidental.com
- 9. Product Case Study #2: Pinterest • Localizing recommendations • Localizing key actions • Connecting men to topics instead of friends (who were likely women) Casey Winters | @onecaseman | caseyaccidental.com
- 10. 2. Onboarding & Education • Get people to the core value your product provides as fast as possible (but not faster) • Don’t be afraid to educate (contextually) Casey Winters | @onecaseman | caseyaccidental.com
- 11. 3. Emails • Test manually, then automate and personalize • Stop fucking sending email like a marketer. Send email like a personal assistant –Right content –Right time –Right amount • Subject lines and calls to action matter; design doesn't as much Casey Winters | @onecaseman | caseyaccidental.com
- 12. 4. Notifications & Badging • Start with transactional, and take baby steps to test other types of messages • Notifications are hard to unsubscribe from, so people will just delete your app • Copy matters, as does landing experience • Badging is severely under-utilized, especially on Android • Pinterest built a completely new experience to leverage badging and notification landing pages Casey Winters | @onecaseman | caseyaccidental.com
- 13. 5. Customer Service • Studies show that people who have a bad experience that is made up for by the company are more loyal than people who never had a bad experience • Go out of your way to make sure you repair any negative experience • Train your team to spot them and solicit for them • At Grubhub, we trained every customer service rep to use Twitter and Facebook Casey Winters | @onecaseman | caseyaccidental.com
- 14. 6. New Platforms • Going multi-platform –Not only web to mobile, but mobile to web • Every company I’ve seen sees LTV at least double when the user goes multi- platform –You can incentivize this like we did at Grubhub –You can optimize this like we did at Pinterest Casey Winters | @onecaseman | caseyaccidental.com
- 15. 6. New Products • Understand core product limits to engagement before investing in new products –Snapchat hit a limit with core product, so they created Stories to get an even bigger market to engage, then Discover for even broader appeal. Casey Winters | @onecaseman | caseyaccidental.com
- 16. 7. Promotions DON’T DO IT (unless…) Casey Winters | @onecaseman | caseyaccidental.com
- 17. 8. Loyalty/Engagement Programs • These programs should be profit drivers not cost centers, unless building a moat Casey Winters | @onecaseman | caseyaccidental.com
- 18. Find your bucket to focus on • Apartments.com was definitely infrequent, non-loyal • For Grubhub, we had to figure out biggest opportunity • Yummy Rummy Casey Winters | @onecaseman | caseyaccidental.com
Editor's Notes
- I’m going to butcher a quote from Dan Ariely and say that retention is like teenage sex. Everyone says they are doing it. Nobody really is doing it well.
- You may think you’re doing retention well. People will say I’m sending emails… Fuck your emails. They probably suck. They probably have a promotion too, training people to get your product for cheaper.
- What happens when you have effective acquisition, but not effective retention. Let’s look at Twitter. Twitter’s activation rate into MAU looks to be about 15%. Since it’s so effective at acquiring new users, it’s basically run out of the internet population to grow its MAU number. So it stays around 300 million MAU, with now over 1 billion dormant accounts. Re-engaging people who already tried you is way harder than activating new accounts. It would have been better for Twitter not to sign up these users until it improved its retention.
- Okay, so you’re bought in that retention is important, but you want to know how exactly you do that. That brings us to the lunch section.
- Okay, so you’re bought in that retention is important, but you want to know how exactly you do that. That brings us to the lunch section.
- So what improvement did Grubhub make to its product to increase retention. Focus on restaurant variety: The more restaurants people could order online from, the more conversion rate improved, as well as activation and frequency 2) Lower minimums and delivery fees: Restaurants set their own minimums and delivery fees. We had to educate and coerce restaurants to lower these so that repeat order rates were higher. Restaurants almost always more money in the long run through higher volume. 3) Burying guest ordering: On first order you could create an account, connect with Facebook, or continue as a guest. We started noticing activation rates of users who placed an order aguest on first order were dropping year on year, and # of people choosing that option was increasing as well. These were all given equal weight in the UI. So, we tested using that same page and talking about the value of an account, and burying the guest option way below the fold with a link that said. “No thanks, I hate convenience. Continue as a guest.” New order voume did not drop, but # of people choosing guest dropped from 50% to 15% with no drop in retention for account holders.
- Localizing recommendations: Pinterest recommended content based on global popularity, biased toward English Localizing key actions: The Pinterest metaphor of the “pinboard” wasn’t as familiar to people outside the U.S. Changing to the local words for “save” and “collection” improved activation rates. Connecting men to topics instead of friends (who were likely women)
- Onboarding and education may be a big or small opportunity given how complicated your product is. At Grubub, 85% of people ordered the same day they got an account. So we never talked about it. At Pinterest, it was a huge focus of ours. So, what’s the goal of onboarding? To get people to the core value your product provides as fast as possible (but not faster) So, if you look at this screenshot of Pinterest, this is what happens after I click those topics from a few slides earlier. I am presented the core product experience, a feed of content that matches my interests. It’s not a blank screen. This is called addressing the cold start problem. You hear a common idiom thrown around startups that if your design requires education, it’s a bad design. It may sound smart, but personally, I think it’s bullshit. Don’t be afraid to educate if it helps your users connect to the core value, but try to show not tell. You’ll notice here Pinterest is educating users on what they can do in a contextual way. It’s not making you go through five screens that tell you everything Pinterest can do. As you experience the product, it’s telling you what things are for and what can you do. After you scroll, you see this, including these pulsing circles indicating the content is clickable.
- All right let’s talk email. Emails are a key driver of retention. They won’t solve your retention problem, but they will certainly help if you do them right. At every company I have been do, people hated email and didn’t want to send them to their customers. When they finally did, they saw lifts. You are not your customer. You get more email than they do. Emails help them if they’re connected to the core value of why they use your product. At Pinterest, I made this mistake. I set up campaigns with emails that explained all of the things Pinterest could do. People don’t care about what Pinterest can do. They care about seeing cool content related to their interests. So we replaced those emails with popular content in topics of interest for each user, and our retention increased. So, if you’re starting emails, you want to try to do some things manually to see what works. When I started at Grubhub, they weren’t sending any email, so I emailed the entire user base asking what they wanted to hear about from us. They said new restaurants to order from and new deals. We didn’t have many deals, so I started manually curating new restaurants once a week by city to prove it was valuable. Once it was driving thousands of orders, we started a personalized program for each person’s address with multiple email types that rotated over time. The goal is to get a problem that sends like emails not like a marketing interrupting you, but like a personal assistant making your life easier. When do most people think about ordering food? 4PM. So that’s when Grubhub sends suggestions. Subject lines: they matter. One of our engineers tested 4,500 variants of subject lines for our emails. Hundreds of thousands of additonal WAUs resulted from this. Some emails’ open rates went up 40%. We did a major redesign of our emails before this project, and no metrics moved meaningfully.
- Notification are a bit more sensitive. I suggest starting with transactional message, and then baby stepping into other messages. Notifications are hard to subscribe from, so people will delete your app Copy matters a lot for push notifications, so test variations. And don’t just drop them into the app. Drop them some place that matches the intent of the push. And utilize badging of your app icon. It drives re-engagement, and no one is doing it on Android because they don’t realize the APIs are manufacturer specific. We did a long term holdout for badging, and saw a 4% sustained lift in DAUs after a year.
- The biggest misconception of loyalty programs is that they are actually rewards for customers. They are there to increase the profits made by the company. Now, what’s important to realize about engagement programs is you have to know what behavior you’re trying to change. I group customers into four buckets, and see what the biggest opportunity is for the business. For frequent but non-loyal users, you have to figure out a way to incentivize the usage they are doing not on your platform to come to your platform. For infrequent, but loyal users, you need to incentivize additional use cases for them with your product. For customers that are infrequent and non-loyal, that’s a bigger issue. You usually need to create additional products, services, use cases, and incentives. At GrubHub, our biggest opportunity was here. People ordered delivery frequently, but not always with us.
- So, Apartments.com was definitely in the infrequent, non-loyal segment. People look for apartments a max of once a year. That’s very infrequent. So, when they do a search again, they usually forget how they found the last. What Craigslist did is provide that kind of value in tons of other areas so people used it regularly. That meant when they needed a new apartment, it was already top of mind. So, all we could do at Apartments.com was build in some moving services and some get to know your new neighborhood services through partners. We tried to create an apartment living section, but it was not that important a need. For GrubHub, we needed to learn which segment to focus on. So we surveyed users to understand their loyalty, and data mined our user base to understand frequency. Once we had an understanding that we were frequent, non-loyal, we had to understand the use cases that people used delivery, but not Grubhub. Once we talked to a bunch of users, we got some major reasons. Then we surveyed our non-loyal users again to get statistical representation of what biggest reason for ordering delivery, but not using Grubhub was. Was it a product issue, a coverage issue, or just habit. It was just habit. We thought we could change habits, but the unit economics of loyalty programs suck for marketplaces. When someone spends $30 with Grubhub, Grubhub doesn’t make $30; it makes $3-5. So you model out what you can spend and how much you can influence. In our case, traditional models didn’t work. 25 orders would get you a free drink. Not great. So we got more creative. We made it a game you played on every three orders with a 25% chance of winning and a variable reward. You generally want to shift to variable rewards in engagement programs instead of static awards, because people become accustomed to static rewards, so they cease to influence behavior. You should get a lawyer if you do this, as there are so many weird gotchas. Like having to post a bond for each state, like being classified as a sweepstakes, which means you need post bonds for each state you operate in, you can’t AB test, and you need to be able to play without purchase. So at GrubHub, we were classified as a sweepstakes, so we couldn’t AB test. So we could either do it in one state and not another, or pulse it, like the McDonald’s monopoly game. We chose the latter.