Growth Lead Michelle Sun on Buffer’s Success at Converting Free Users to Paying Customers

Posted by Anant January 13, 2023

Michelle is the Growth Lead at Buffer, and in this interview she talks about converting free users to paying users and why Buffer has been so successful at it. We also discuss what it looks like for Buffer as they follow Sean Ellis’ Growth Pyramid.


→ A Buffer is a tool for scheduling social media updates

→ Launched in 2010 and been growing for almost three years

→ The user base has grown from 500,000 to 700,000 since the current growth lead joined the company

→ Her growth lead at Buffer describes her role as unique and without a typical day

→ Involves building internal dashboards, analyzing A/B testing results, and uncovering insights from data

→ Explores the use of exploratory data analysis to identify core metrics and user behavior

→ Explains how Buffer segments users and uses data to track usage of new users and make sure they activate the product and move towards becoming core users

→ She learned to have a data-driven culture at her previous job and brought that philosophy to Buffer to inform product decisions

→ Her previous experience as a data analyst at Bump Technologies

→ What lessons she brought to her role at Buffer

→ And a whole lot more


Michelle’s Personal Website

Buffer App



Bronson: Welcome to another episode of Growth Hacker TV. I’m Bronson Taylor and today I have Michelle Son with us. Michelle, thank you for coming on the program.

Michelle: Thank you for having me.

Bronson: Yeah, we’re so glad to have you here. Now, Michelle, you are the growth lead at Buffer. So tell us real quick, what is Buffer? What does it do?

Michelle: Yeah. So Buffer is a tool for people to schedule their social media updates in a smarter way. So maybe I can provide like my, my own use case a little bit. And so, for example, in the morning, like, I usually spend 10 to 15 minutes reading some articles that I saved the day before and I was loaded onto my buffer. And so I don’t overload my followers on Twitter for ten articles. And the first thing in the morning and Buffer will schedule those updates and space it out throughout the day.

Bronson: Yeah, that’s actually how I use it as well, except I use the nighttime right before I go to bed or right after I put my son to bed. I go through Feedly. I look at all the the articles that I want to retweet, and then I’ll put them in the buffer because I know a bunch of people don’t want ten articles at midnight from me. So exactly my use case. So when did Buffer launch? How long has it been a product?

Michelle: Yeah. So it was launch late 2010. So we’re approaching our third year now and we’ve been pretty lucky about our achieving product markets is pretty early. Our founder, Joel launched the product as a two page MPV with a pricing option. And with that, within the first week, we actually got the first paying customer.

Bronson: Yeah, that’s great. You can’t be that. So it’s been around for almost three years. How long have you personally been there? As the as the growth lead?

Michelle: Yeah, I joined before February this year. And I actually know Joel and Leo, the co-founders of Buffer since early 2012. Back in Hong Kong, they were traveling and working out of Asia and I was actually from Hong Kong. And so it was a pretty good story there. And since I joined Buffer, our user base has grown from 500000 to 700000 and is still growing pretty strongly.

Bronson: Yeah, that’s great. So as the growth, we kind of walk us through a typical day because you have a unique job, you know, overseeing growth at Buffer there. What do you do? What do you not do? How do you spend your 9 to 5? What does your day look like, Michel?

Michelle: Yeah, I think that’s a great question. Bronson You know, like, actually, I can’t say that there is a typical day, which is probably one of the best things I like about my role. You know, given any given day, I can be building some internal dashboard for the team to look at our data and access things more easily. Or I can look into our IP testing results and communicating with products and making sure that we’re moving the needle the right way and we can. I can also be looking at exploratory data and diving into some bigger questions about how the user is engaging with our product.

Bronson: Yeah. Let me ask you a few follow up questions that you mentioned, building dashboards. You actually have a background in programing as well, is that right?

Michelle: Yeah.

Bronson: Yeah. So you can actually build the dashboards based on the data for the rest of the team to to look into. And then you also mentioned exploratory kind of looking at the data that way. What do you mean by that? Do you just mean kind of looking at data to see what you might uncover even though you don’t know what’s going to be there? Is that what you’re getting at?

Michelle: Yeah. So that’s a great question. Basically, the first couple of months that I joined the team like that, that was basically my main focus, which is the data exploratory data analysis. The reason is because even though in the beginning we wanted to have a dashboard for the whole team, one thing that we ran into is that we actually need to find out what is the core metric we want to focus on first. So we took a step back and really dove on the data. And some of the examples that we borrowed from is, for example, the Facebook case of seven friends in ten days. Or I think there was a similar one from Twitter that we were curious about our own users usage of buffer. So that actually yielded pretty interesting results where we were able to split our users into more casual and inactive users. And the core users. We found out that it’s basically users that have interacted with over five days within the past month.

Bronson: Okay. And so when you separate out the users, does that change how you message them, how you treat them? Does that change kind of the product? Yeah. So what are some ways that that would change the product? Because I think that’s where a lot of people don’t find those kind of, you know, camps of users. But even when they do, they’re not really sure what to do with that kind of information. It’s hard to know what the next step is. What are you guys doing with that now that you know the kind of write down?

Michelle: Yeah, to be fair, actually, we we’re still kind of fine tuning it and learning about it ourselves, too. But what we found helpful is to say segmenting out these users, we’re able to kind of make sure, you know, that if they reached a core user’s camp, they’re very unlikely to churn. So so right now we’re tracking the usage of new users and making sure that, first of all, most people, like more users, can activate our product first and we will gradually figure out more experiments to move them from not just active users but core users.

Bronson: I just use it as kind of a method to move them through the channels to get them to that core place. Now, before Buffer, you also worked as a data analyst at Bump Technologies, and we actually had the founder of Bump on a month or so ago. What are some of the primary lessons that you learned from both leave carried over into your role at Buffer? Because I know they had kind of an explosive growth. They had a lot of really great things happen there early on. Were there any takeaways that you carried over early?

Michelle: I had an incredible experience at Bump, and I was working closely with the data scientist and the product manager there. And I think what I took away from both is that it’s a very data driven culture and a lot of the products and systems and just the entire company culture is driven by data, at least from my experience. And and when I joined software, I really brought that philosophy. And I was planning to also help the team use data to inform product positions more frequently.

Bronson: Well. Is that where you kind of got the first taste of really using analytics to inform our product? Was that where you kind of cut your teeth?

Michelle: Yeah, I mean, of the user base is so huge and it’s, it’s like a really perfect playground for anyone that’s interested in data. Like, imagine 100 million users and all the data was tracked really well and it was just an incredible experience to get my hands dirty on lots of data.

Bronson: Yeah, that’s great. Let’s talk about Buffer a little more. You know, it seems like everywhere you turn online, people are talking about the growth of Buffer and how great Buffer is doing. And you just mentioned earlier that you guys are now with 700,000 users, is that correct? Yeah. Yeah. Zero, 700,000. Talk to me about the growth curve. Has it been kind of steady, you know, up into the right? Has it been, you know, ups and downs and, you know, fits and starts? I mean, what does the curve look like for you guys?

Michelle: Yeah, I would say the overall. To the growth rates, which when you think about it, the larger the numbers, the more difficult it is to maintain that number. So I think that’s really encouraging for us in terms of the deeper metrics we look at the monthly active users and daily active users and the growth rate is actually outpacing the overall users. And I think that’s even more encouraging from my standpoint that the users are not only signing up to our service but also actively using it, and we’re probably also doing a decent job in retaining them as well.

Bronson: Yeah, I don’t know if you guys disclosed this number, but I’ll ask it anyway in case you do. What percentage of buffers users actually convert to a paying plan? Because I know early on when they were raising money for Buffer, that was one of the metrics that got VCs excited was the the freemium to paid conversion rate. Do you guys disclose that?

Michelle: Yeah. So yeah, the most recent number that I’ve got is back in April, 1.8%. I think it’s pretty steady in the past couple of months. It’s like.

Bronson: Yeah, and like, you know that you were open about that because it gives people watching this a real idea for the numbers because they may have a product with a 1.5% conversion rate to the premium usage and they may think that’s bad or something because they’re comparing it to a different kind of product that has a different sort of model. So 1.9 I mean, that’s great because at 700,000 users you can do the math and 1.9%, that’s a great conversion there. What do you think has allowed you guys to have that kind of conversion rate? I think I know what you guys are doing from the outside looking in, but I want to hear from you guys. How are you getting people to upgrade?

Michelle: Yeah. I think that’s a that’s a good question. And I think from the very early stages, the two founders have been really focused on starting with building a product that solves a real pain point. And that’s really important where if we save our users some time every day or self help them do something painful for us, it’s like scheduling social media updates. Then we can find people that see this pain point the most acutely and actually pay for our service. And then another thing that we also focus on is to fine tune the limit where the free product can offer. So it’s actually a tracking point where how much should we offer for the free users? It shouldn’t be too late until that they would never reach that limit. But we also want to offer them enough of the peak of the entire product and how powerful it is to entice them to upgrade. So we went through a couple of rounds of fine tuning and in the early stages we actually have a very limited set of features for the free users which limited our growth for a little bit. And we now can be more comfortable with the set of features that we offer to the previous and the premium users.

Bronson: Yeah. When you gave more to the free users, did you actually see the growth or the conversion increase? Is that what you’re saying? Yeah. Why do you think that is? Why is it actually giving the free people more made you convert? Because it seems like, you know, it would be the opposite. You give them less and they’re forced to upgrade.

Michelle: Yeah, I think most users actually have a warming up time to a product just like Evernote. I like if if it was a paid product to begin with, I, I would believe a lot of users would actually like turn to free, free products or even less well-designed or well executed products and just having the opportunities for the user to actually interact with the product and, and develop like a almost emotional attachment to the product. Like give us more time to convert them.

Bronson: Actually, yeah. No, that’s in my experience as a user because I was using it and you guys asked me to upgrade at just the right time because it had become a habit. So I was in the process of using it every day for a certain amount. And then when it was time to upgrade, I already saw the benefit of it and how much I needed it and I was already connected. So going to have to do that, I have to upgrade. Oh, you guys, at least in my experience, the way I was using it, it was at the exact right time where I was like excited about it and needed it, wanted it, and then I needed more. So I think it was good. And just so people know, kind of the way it works is, you know, you add a tweet to your buffer and then eventually you added too many tweets. And so that’s hard because I like to work ahead. So if I got, you know, ten, 20, 30 in a row that I’m ready to tweet like I don’t have to come back tomorrow and then do my research again and find those articles again or find those tweets again. So it just it made it where I was going to have to upgrade.

Michelle: Yeah. Nice. I didn’t know you were a paid user. Is awesome to hear that.

Bronson: Yeah, absolutely. So now you guys recently wrote a blog post or you recently wrote a blog post about the transition to growth at Buffer. And I use the phrase transition to growth in a technical sense because you’re referring to Sean Ellis’s growth pyramid kind of as your structure. So real quick, kind of walk us through the three stages of Sean’s growth pyramid because we actually haven’t talked about on the show before.

Michelle: Well, yeah, within Buffer, we’re all big fans of Sean Ellis and I read his blog religiously. So from my understanding, like the three stages of growth, according to John Ellis, is that the first one, the foundational of the pyramid is achieving product market fit. And once you hit that stage, you’re well positioned to transition to growth, which is the second stage. And lastly, the top of the pyramid as well. Yeah, for Buffer, as I mentioned, we’ve been lucky to achieve the product market fit fairly early on and we’re working on the transitioning to growth stage.

Bronson: So Michelle, you said that, you know, your transition to growth, which means you think you found product market fit, what are some of the internal indicators that makes you all believe that? Because it’s hard sometimes for startups to know, should we move on? Does the product fit enough or should we stay there and really tweak the. Product more. So how do you guys know that it’s time to move on?

Michelle: Yeah, actually, our founder, Joel, has been a follower of the lean sort of philosophy for quite a while and one building buffer. He has followed through on our list methodology in validating his product market fit actually. So Sean had a blog post about if you want to check if you have achieved product market fit, you can send a survey to your existing users asking how would they feel if this product does not exist tomorrow? Hmm. And from his experience in working with various startups, including Dropbox, he found that around 40% is the safe margin for if 40% of the respondents say that, they would be very disappointed that you have a cheap product market fit. Yeah. And so Joel actually sent out this exactly question, the exact question nine months into the birth of buffer with around 30,000 users. And the result was that 55% of the free users and 70% of the users said that they would be very disappointed. So we we beat the margin by quite a bit.

Bronson: Yeah, that’s great. So you guys feel like you found product market fit and you said that, you know, that early on they sent out an email to 30,000 users after nine months. How did you guys get to 30,000 users? Because part of getting the market fit is having some early users that you can poll and ask them how much they need the product. But to get that, 30,000 is no small feat. So what did that process look like for you guys?

Michelle: Yeah, I think you’re totally right. And I think we’re thankful for Leo, our other co-founder in the very early stages. The first year, he’s been working very hard at writing a lot of blog posts, both on our own blog buffer that’s locked up for and guest blogging for a lot of social media sites. And that has propelled the growth of buffer within the first year by a lot. And one of the blog posts that he wrote detailed on this content marketing strategy. And within the first year, we estimate that over 70% of the users sign up because of that content marketing.

Bronson: Wow. So of all the traffic he has received early on, would you attribute most of it to your own blog or the guest blogging that Leo did? Or was it just a mixture of both, really?

Michelle: Yeah, I actually don’t have the actual figure like tracing back so early in the beginning, but I’m guessing that it’s a mix. And in in the early days, we also kind of get our fame foreign pool by guest blogging a lot and driving organic traffic to our blog. So I think definitely that’s there’s a virtuous cycle happening there.

Bronson: Yeah, that’s great. So you talked about, you know, that your transition to growth now. What does that actually mean? So you kind of leaving product market fit. So where are you going to start doing that you weren’t doing before now that you’re transitioning to growth? Walk us through that a little bit.

Michelle: Yeah, I think there’s two fold. So when in this stage of transition to growth, we’re focusing on first running experiments in faster pace. So where we’re committed to running any test consistently, having one or two tests running and adding more in different parts of the product in any given point in time. And secondly, which involve a larger effort is to start tracking internally our user metrics. So what are the means by use of metrics is that we try to trace each user interaction with the product back to each users. So instead of aggregating the numbers, for example, oh, this month, how many people were were active and we want to know Bronson is Bronson active? And how how does Bronson use Buffer and why does why is Bronson more active? And what is he doing differently than some other inactive users? So that’s been a really educational effort for the entire team to be able to dove back into the users actions every day.

Bronson: Yeah, no, that’s good. You mentioned, you know, kind of going back into the actions because it kind of leads into what I want to talk about next. You have a strong back. Analytics. I mean, you actually have a degree in economics, which, you know, is very related in a lot of ways, in terms of, you know, just dealing with numbers all the time. How do you currently track metrics that buffer? So you talked about, you know, now that you’re transitioning to growth, you’re trying to go back and track the individual. Users are using all the Shell products to do that. I don’t even know they exist to the level that you need them to to do that. Are you building stuff in-house? Is it a mix? What does that setup kind of look like right now?

Michelle: Yeah, right now we actually are building everything in-house. We were using some external tools as well. But I think we’ve reached a point where we wanted more flexibility and accuracy in slicing our user data. So we decided that it was the right time to transition to our own tools.

Bronson: Yeah, that’s great. So let me ask you about how you’ve grown in your understanding of the data. I know you’ve only been there, I think, four or five months now, but, you know, the team has been there for three years in total. So I’m sure they’ve kind of gone through some cycles of really, you know, letting the numbers speak. Are there things that you track now that maybe you didn’t track before? Have there been any surprises in the data that you didn’t expect? Tell us the story of the numbers a little bit.

Michelle: Yeah, I would say one of the surprises that we had kind of going back a couple months before was a major redesign of our Web dashboard. And it was it happened late last year. And with the goal in mind to have more people install our from our browser extension and we changed the user flow to engage them to install their Chrome extension very early on right after they signed up. And we actually found out that this change has decreased activation for the new users significantly up to half, half the previous vote. So that’s that’s been something that’s a difficult lesson to learn. But definitely we we kind of pick up very key learnings there to make sure activation and retention can be treated in different stages in a user’s lifecycle.

Bronson: Yeah, you mentioned activation and retention. There are those the key metrics that buffer really optimizes for. Is there any other kind of key metrics that you guys are looking at every morning when you wake up or what are the main numbers for you guys?

Michelle: Yeah, I would say activation and retention are very key metrics that we look at. We also look at the traditional, you know, acquisition numbers and activation retention and revenues. And so we have a real time dashboard on the sign ups, upgrades, downgrades and how many updates people are sending today. So these are all the numbers that we try to keep our minds on top of.

Bronson: Yeah. So what advice can you give to a startup that hasn’t yet set up their analytics? Any lessons you’ve learned the hard way? Because, like you said, you’re the one building the dashboards, actually programing them. You’re the one deciding kind of what to track and what have you learned the hard way that maybe you could pass along in terms of tracking analytics?

Michelle: Yeah, I mean, I would say start doing it. So I think kind of taking a step back when startups are thinking about growth and I think it’s most important to identify the stage that the startup is at in the growth pyramid that we talked about earlier. The reason why is that for startups, if they experience low growth, sometimes it’s because they have not achieved how the market fit. And when when you’re in the stage of achieving product market fit, it’s really important to, you know, get out of the building and talk to customers and really understand why they use the products and what their actual needs are. And these are actually things that not necessarily data can tell us and it can be misleading to rely on, you know, just a small set of data to to make product decisions. So and once the product market fit is is achieved and moved quickly and track user metrics because it takes time to kind of accumulate enough data to to draw decisions upon. And you want to start that early. And for example, our lesson on the dashboard tells us that we can have the data more real time then we can probably. Kind of resurrect like more users in the early stages.

Bronson: Yeah, that’s great. Let me ask you this kind of a higher level question. If you had to pick three reasons why Buffer has succeeded in growing, what would they be?

Michelle: Yeah, that’s a that’s a tough one. I think the three ones. Let me give it a shot. The first one would be the product focus of even from the very early stage, Joel has a strong focus on the product methodology and that really help the team make sure that we’re solving a real problem for customers. And secondly, we moved to focus on marketing quite quickly in parallel with the product, and that has really helped us engage with users and get the word out and build our brand. And the third one would be the emphasis on customer support. And I think we’re actually we can talk a little bit about that as well as buffer and customer support is really in the blood of the entire team. So we currently have eight people and two of us are actually full time customer support and even four engineers. We have one full day in the week every week that we work on the support tasks. That includes working on some more technical emails in the inbox or responding to tweets. And for myself, I’ve been really enjoying the days that was and the support tasks, mostly because I think as an engineer it’s really easy to just focus so much in your head into the code and you just forget that there’s actually real people that are using the product and you know, you’re you’re helping someone to achieve their their work or their life goals. Right. And and probably some bugs can be really frustrating to them. And so I think having that once a week, logging into the customer’s shoes is really, really helpful for the entire team to stay in sync with serving our customers.

Bronson: Yeah, sounds like a great practice. We had one other founder come on who does the same thing and he talked about how much it really informed their engineers. And ever since I heard him say it, I thought, what a great idea. So I’m glad to hear that. You know, you mentioned you guys changed the user flow and it decreased activation by about 50% at one point. And then you had to kind of figure that out. Let’s talk about the opposite of that. What’s maybe one of the single best things you’ve done that you’ve implemented that really increase usage? Is there any been like any small changes or any, you know, thing you’ve done was like, wow, the numbers really went up into the right with that one change. What would you point to?

Michelle: Yeah, I think if I have to point to one single or growth pack, if there’s a there’s one at buffer, I would still think of it as the customer support side. I think having the kind of feelers and all the users telling us there are problems and even customer complaints have been really useful for us to retain customers. We’ve had numerous cases where, you know, angry users get in touch with us and because of our support, they actually kind of change their mind and say, okay, I’m not canceling this anymore and I think this awesome and thanks so much. And so I think our philosophy as Buffer has been not only to acquire customers, but also making sure that they stay with us. And even if they leave buffer at this point, because our product didn’t fit their needs, we want to make sure that they are they keep us in mind and it later on, if we have certain features come out, we can get back in touch with them. And so we’re taking a more long term approach in terms of tracking the growth, if there is one.

Bronson: No, it’s good. That’s good. Well, Michel, this has been a great interview. Thank you so much for coming on Growth Hacker TV, for sharing your insights and for giving us real numbers. Yeah, you’re telling us, you know, what percentage of converting users that how many users you have now, how many users they had after nine months. And that’s kind of transparency really educates other startups and helps them immensely. Just get a handle on what numbers could be or should be. So, Michel, thank you so much for coming on.

Michelle: Thank you.

Ready To Grow Your Startup?TVicon

Get the strategies, motivation, and in-depth interview with all the details every week!


Popular Blog Posts

26 Simple Websites that Prove Innovation Doesn’t Have...
Learn How to Achieve Startup Growth with Matt,...
Learn How Unleashing the Power of Product Marketing...
Learn How to Drive Viral Growth with Sangeet...
How Adam Berke Made AdRoll the Internet’s Largest...

Share On

Are you an
who is trying
to grow a

Get the strategies, motivation, and in-depth interview with all the details every week!