Is Marketing On Marketplaces Worth It?

Links:GitHub Student Developer PackSeth GodenAWS MarketplaceHeroku MarketplaceGitHub MarketplaceMicroconfRailsConfHoneybadger Blog

Full Transcription:
Ben: Speaking of Macs and stuff, software kind of being weird, my son, my younger son over this week has been playing with Raspberry Pi again.

Starr: Oh, that's cool.

Ben: Yeah, he wanted to do some funky USB stuff, and he's like, "I think I'll try it with the Pi." And so he's getting all into that, and then he's coming to me with questions like, "okay, how do I start up something on system boot?" And I'm like, "Oh, well let me introduce you to System D."

Starr: Oh my goodness, Ben. You've been waiting for this day, haven't you?

Ben: Yes, it's pretty awesome.

Starr: Oh, that's so great. That's so great. I'm so glad. 

Starr: Did you buy your kayak yet? You said you were buying a kayak.

Ben: No, I didn't buy the kayak yet. I almost bought the kayak, but I decided against it because we had the expenses for working on the house and sending Addison off to college and so on. So yeah, hesitated. I'm going to wait. Maybe this summer. Maybe in the summer I'll get that kayak finally.

Starr: Yeah, that's the reason I was thinking about it. I was thinking about... I was like, "Oh, has Ben got his kayak yet?" And then I was like, "It's really cold to be kayaking." Do people kayak in this weather? I mean, probably, because it's Seattle and people are crazy, but-

Ben: Yeah. Well, just this week some dude crashed his plane in a lake nearby and a kayaker rescued him. So yes, kayakers apparently do paddle out this time of year.

Starr: Oh, well thank goodness for these kayakers who want to be freezing, I guess.

Ben: I know.

Starr: They saved that guy's bacon. All right. Today I think... Well, our astute listeners will know that Josh is not with us. He's had a bad cold and yeah, we just haven't been able to... He hasn't gotten well in time to make this happen. So we're having another one of our Honeybadger fireside chats, which is our branded marketing effort for when one of us is gone. And so it's just two of us kind of chatting, and-

Ben: It's like everything is better when it's a branded marketing effort. You know?

Starr: Exactly. Exactly. That little trademark, like that little TM sign. That's like the salt. The salt in your cooking, you know?

Ben: That's when you know you're getting a high-quality podcast.

Starr: Exactly. We looked up the Unicode character for that and everything.

Ben: Nice.

Starr: Today we're going to be talking about a number of things. They're all sort of loosely related. First we're going to be talking about GitHub Student Developer Pack, which is this sort of marketing effort that I don't really know about, and so I'm going to use this as a chance to sort of ask Ben lots of probing questions and figure out what's going on. And since that's kind of like a referral marketing-type effort, I don't really know what the official type name is for it. We're maybe going to talk about other things that we've done in the past. What is the GitHub Student Developer Pack, and why have we done that?

Ben: Yeah. So GitHub started this a while back where they offer a variety of products and services to students as part of a outreach thing. And we just recently started being involved in that, because I think it's a good idea to get students familiar with our product. It's the Microsoft strategy that they did way, way back in the day. Well, even Apple before them, right?

Starr: Oh yeah.

Ben: You get students using your product, and then as they grow up, they eventually buy your products. Right? I think Apple has been very successful in that in the education market. And then Microsoft came and did the same thing, and then Google did that too with Chromebooks now. Anywho, yeah, the GitHub Student Developer Pack is all about helping students get access to a variety of services and products to help them in their development. So GitHub's like, "You know what? We should make students better at developing stuff and give them tools to help them do so." And so you can go to their website and get all kinds of goodies if you happen to be a student.

Starr: I study the Blade. Does that count?

Ben: Hmm. Well, maybe, but only if you happen to have an email that will certify you as an actual student.

Starr: Okay. An email. Really?

Ben: Yeah, yeah. If you have a .edu.

Starr: Oh, an email address. Okay.

Ben: Yes, yes.

Starr: I'm sorry. I was like, so they think... Anybody who can use GitHub can forge an email. Like, come on.

Ben: No, they actually do some verification to make sure you're a legit student, but once you can verify that you are... And if you don't have the .edu kind of email address, if your school isn't cool enough for that, then you can upload your student ID, and presumably some human somewhere will look at that and say, "Oh, yep, this is an actual student," and then flip the bit on your account that says this person is a student.

Starr: Awesome. We've done various things like this in the past, and I think our thinking about these have changed a little bit. And this was maybe prompted by that Seth Goden course that you and Josh did. In the beginning, we were really looking for... I don't know. We were just kind of tracking these efforts very seriously and being like, "Oh, okay. How many people are signing up as a direct result of this versus like..." But then, yeah, at some point we were like, "Oh, this is just kind of like... These are students. These are seeds we're planting for years to come, and there's literally no way we can tell." There's no real direct way to track that, right? Is there? I mean-

Ben: Well, yeah. Like you said, we in the early days were really interested in being able to track all the conversions and having measurable results of all the efforts. And over time, we've gotten more comfortable with the idea of brand and marketing, and doing things that will pay off in the future. In this case, I think it's a bit of both, because we do want our brand out there. We do want developers to be exposed to us early in their careers, but also we can track it pretty well, because we are of course signing them up with identifiers saying, "Oh, this person came from GitHub Student." We know of the however many signups we have in a given week, how many of those are coming in through that offer. And over time, we'll be able to find out just how well that converts. Now of course in this case, the time horizon is a bit longer than our usual, because our particular offer for that pack is one year of free service. We'll have to wait a little bit to find out if they convert at all, but we'll see.

Starr: We'll see. Siri, remind me to check back on this in a year. So you mentioned tracking. I assume that there's some sort of development effort involved in setting this all up, because it seems like this wasn't just like going into Stripe and creating a coupon code, was it? Or was it more complex than that?

Ben: Yeah, it was a bit more complex than that. I mean, it did involve creating a coupon code in Stripe, so that was a good start. But we did have to build a custom little app that the student can come through to get access to the offer. What GitHub asks you to do as a partner in this scenario is to allow a user to come in. The developer comes to your site. They click a button that says, "I want to get this offer." And you ask them to log in via GitHub so you can find out who they are, and then you check with a GitHub private API whether this person is actually a student. And then the API tells you yes, no. If yes, then you allow them to come through and claim that offer. And if not, you say, "Yeah, I'm sorry. You can't have this offer," or whatever.

Ben: We had to build that little app. And when I say little app, I guess I should say micro app. In our case, it's just a couple of serverless functions that run on AWS Lambda.

Starr: Oh, nice.

Ben: Really, all it does is do a redirect to GitHub's 0Auth. It's very, very minimal, but it was a little bit of code.

Starr: Yeah, and that's great. I feel like if we're going to be attracting students, we really need to be using the trendy stuff like Lambda. I think that's a good call on that.

Ben: Totally. Yeah.

Starr: They would see us using just Rails for that and they'd be like, "Come on, man. What's that all about?"

Ben: Okay boomer. Yeah.

Starr: Yeah, yeah. I'm not a boomer though. I get to say that.

Ben: That's true. Neither am I. Generation X is totally the best generation ever, if we actually cared.

Starr: I don't even really think I'm Generation X. I'm that sort of in-between generation where I get to pick, so I picked millennial.

Ben: Cool, cool.

Starr: Because that means I'm younger, and I will die less soon.

Ben: We do have a coupon, and one thing that I'm still not sure I did the right way or... Right way. I mean, I'm not sure I did it the way that in the long-term I'll be happy with, because the offer that we give is students can have a free year of our small plan. We have a coupon that they can claim that puts them on our small plan, and it's free. But the thing that I'm not sure about whether I actually did it in a way that I'll be happy with in the long-term is they're on our normal small plan, and once that year is expired, then it'll act like they just had gone a year without paying us. The normal trial stuff isn't really in place, so I'm going to have to build something, I think in about, I don't know, 10 months or so when these guys come to you that does some smartness, like checking on their coupon or checking on their referral source, and then sends them a special onboarding kind of campaign.

Ben: Like, "Hey, we noticed you've been enjoying Honeybadger as a student for a year now. It's time to buy up. I hope you want to do that," or something. I don't know. It may have been better to put them on a different plan in Stripe, and then we could have had a different trial kind of thing. Because in Stripe, you can configure the trial period, and so our default trial period is like 15 days. We actually had to work around that if they had this particular coupon to not give them the trial expiring message that we usually give people. Maybe it would have been better if I actually would have created a different plan in Stripe, and it could have had a year-long trial, and then we could have had our normal kind of trial onboarding sequence happen. I don't know, but I have time to figure it out.

Starr: You know, that reminds me, is it okay if I go off on a little tangent?

Ben: Please.

Starr: You're talking about creating new plans. This is one thing that always struck me. I don't know if there's a better way to do it, but we have so many plans in our database. And dear listeners, I'll tell you about this, because if you go to our website and you look at our pricing page, you will see a nice ordered set of a couple plans you can choose from. But on the back end, this is what, our fourth or fifth revision of those plans? And we don't-

Ben: At least.
Starr: Yeah. And we tend to grandfather people in, so we've actually got several versions of each plan, and then we've got special one-off versions of each plan. If a customer comes to us and they're going to pay us a lot of money and they want a special deal, we're not going to say no, so Ben will go in and create another plan. And I'm not saying this is the wrong way to do things, but when it comes time to change a plan, it's kind of difficult. And I mean, if you're going to grandfather everybody in, I don't know any other way to do it other than to keep the old plan stuff around. But I don't know.

Ben: Yeah, it's become a bit of a operational burden, but it's not something I'm crying over in my milk. But yeah, because we do pricing experiments from time to time, and that's how we got those five or six or whatever versions. It's like, "Oh, let's create a new set of plans." The Stripe ID is kind of like small V2 unlimited tests.

Starr: Yeah, yeah. I remember a long time ago I had to go in, and we originally were looking up the plan by the plan name which wouldn't work, because we wanted to keep the same plan names but just change the options of the plan, like the different parameters.

Ben: Yeah.

Starr: Yeah, I had to go in and make real a plan name. This is the plan name, but this is the real name that really defines the plan.

Ben: Right. Yeah. And that's actually why I hesitated to create a new plan for this offer instead of go with the coupon route. That was the main consideration, I think. But on the other side, I'm kind of concerned what is the return going to look like in 10 months when these students stop using Honeybadger. Right? Because they're effectively falling out of their trial, but it's not a trial because they've been out of the trial for a year now, because that trial was only 15 days. You see what I'm saying?

Starr: Yeah.

Ben: It could really mess with our numbers, so I'm trying to think through that and say, "Well, do I need to change something between now and then?"

Starr: Yeah, that's a good point. I mean, ideally we wouldn't really count them in turn, because they're on a completely separate track, maybe.

Ben: Right, right. Yeah. Something to figure out. But we believe in just-in-time development here at Honeybadger, so I don't have to figure it out for quite a while yet.

Starr: Well, here's an option. Instead of writing that code that you were talking about that goes in and doesn't bill them for the past year, maybe we just do build them for the past year, and we threaten to send them to collections. I know that when I was a student in college, I didn't really understand how any of that worked, so we might actually get some of them to be able to pay up. Yeah, I mean, this is capitalism. I think we're way too honest about things.

Ben: Well, they're used to being in debt, right? I mean, they are students after all, so it wouldn't be a surprise to get another collections call.

Starr: That's true. That's true. I don't know. Let's see. We've got a couple other sort of efforts maybe in the works, maybe not. I've heard them talked about, but I don't think we've actually done them yet. And that is GitHub has a marketplace too, which is different from the student pack.

Ben: Yeah. Mm-hmm (affirmative).

Starr: And also AWS Marketplace.

Ben: They do. Yeah. And we've been on the Heroku Marketplace forever, almost since day one.

Starr: Yeah.

Ben: And that's been a good source of customers for us. And so we've looked at doing the AWS Marketplace and the GitHub Marketplace to add on, basically having our brand, our product be in front of people when they go shopping for stuff at these good places, right? And that's the idea. And the thing is, the reason we haven't done it yet is because every one of these marketplaces has a different way of interacting with your system and telling, "Here's a new user that wants to sign up, and here's how you build that user." Right? Each one has their own way of doing it and it's like, "Oh, I got to write yet another provisioning thing." Right?

Starr: Yeah. And they don't want to standardize, because they all want to own the space. They want everybody to use their standard, probably. Or actually, they don't want other people to exist.

Ben: Yeah. And they have their own opinions about how things should be. I mean, every developer does, right? They say, "Well, we'll do it this way, and you should follow that way." And that's great, except now we have to have three or four different ways to provision an account in our system.

Starr: Yeah, that's true. And that's true for a lot of sort of affiliate or partnership-type plays where... We've had talks with other companies and stuff. It's, "Oh, we should integrate our products. We can send customers to you, and it'll make it more useful for our customers if they can link their error data or whatever." And it's like, "Okay, that's great. We'll put a developer on that for the next three months," and we don't really have that developer. And also, a lot of times the deals aren't really that sweet. They're not really enticing enough for us to put a developer on it for months. Yeah. It's like, "Okay, so we're just going to be a name amongst all of our competitors in your list, and we're probably not going to get any real sign-ups." I don't know. I can be too pessimistic about things like that, and I own that.

Ben: I think it's situational in that for us, early on, we supported Heroku. GitHub Marketplace didn't exist yet. AWS Marketplace, I don't think it existed or if it did, it wasn't really relevant to our audience. Heroku really was the only option when we launched.

Starr: Yeah.

Ben: And so we did launch early in our lifecycle with Heroku Marketplace, and that was a great way for us to get our early customers, because we were unknown and that's how we got some discovery. But as time has gone on, we've been at this for more than seven years now. The number of people coming to us through Heroku has just decreased over time. I mean, we still have plenty of customers that have our service as a Heroku add-on, but it's definitely been on the decline for years now. And I think part of that is we have enough brand out there aside from Heroku that people just come to us through word of mouth or through advertising or whatever. They're not discovering us so much through Heroku. If we were brand new, I think it would be more of a priority for us to get on every marketplace under the sun to get our name out there some more.

Starr: Heroku seems to me like it was sort of a special case though, because everybody was getting on Heroku at the time and everybody was using Heroku extensions. You know what this reminds me of?

Ben: Yeah.

Starr: Remember how years ago, everybody was using Facebook apps? And I know you developed a few Facebook apps for your clients when you were a contractor.

Ben: I did.

Starr: Yeah, but everything was going to be Facebook apps.

Ben: Yeah.

Starr: The Horoku add-ons kind of remind me of that. I mean, they're actually a bit more useful than Facebook's apps, I feel. It's actually kind of a good user experience for some things. But yeah, it just seemed like everybody sort of wanted to have this sort of thing where it's like you'd go to one place and you'd get your hosting, and then you just press some buttons in that control panel and everything is set up for you already.

Ben: Yeah.

Starr: And that's kind of gone by the wayside a little bit, or maybe it's been replaced by... I don't know. I don't know. It seems like people are more into full-on dev ops these days, and if they're going to go to one place, it's going to be AWS and whatever.

Ben: Yep. Yeah. I don't think there's any right answer. For several months now, I talked about adding AWS Marketplace and GitHub Marketplace. And like you said, it's like, "Okay, that sounds interesting and could be useful to us, but we got to put a developer on that, and we just had other things to do." I guess if someone is desperate for us to be on the GitHub Marketplace so they can have their consolidated billing and they want to use Honeybadger there, yeah, reach out to us and let us know. But alternatively, we will be getting around to that eventually someday, because yeah, it's good to be out in various places. Why not?

Starr: Let me just pitch an idea to you, Ben.

Ben: Okay.

Starr: Okay. Honeybadger Marketplace.

Ben: Yeah.

Starr: Right? You come to us. You can purchase your AWS services through us. We'll get a very reasonable cut, and it will be great. We'll make lots of money.

Ben: Totally. Yeah, yeah. Have a friends at the Honeybadger program. And then we'll sell credit card processing, because that's where the real money is.

Starr: Oh, is it?

Ben: Oh yeah. You get that 2 percent of that 1 percent of that transaction, and man, you can just make bank. Yeah. Ask me how I know. I have a friend that's in that business. It's pretty awesome, actually.
Starr: Really? How do you get in the credit card processing business?

Ben: It's actually really easy, because the credit card-

Starr: That's kind of scary the way you said that. "Oh, it's easy."

Ben: Yeah, yeah. In fact, actually, I still get checks from when I was reselling credit card processing. I'm not getting big checks anymore, but 
yeah.

Starr: Oh, okay. Okay. So you're talking about reselling, you're not talking about setting up a credit card processing company.

Ben: Right, right. Yeah, right, right. Being a reseller. Yeah.

Starr: Oh, that's great. Your friend does well for himself?

Ben: Oh, yeah. He's done very well for himself. He was in it in the early days. Back when the iPhone launched, basically, he built an app that you could use to run credit card transactions, and this is before Square.

Starr: Okay.

Ben: And he found quite a successful business by selling that app to people who are selling stuff at flea markets who needed a way to handle transactions on their mobile device. And he also set himself up as an affiliate for the credit card processing, so he got a cut of their cut of each transaction. Right. Yeah. Pretty sweet business. So yeah, we should totally do that.

Starr: All right. Yeah. Let's workshop that a bit. We'll do some improv sketches around it and see what happens.

Ben: Yeah, yeah. Yeah, we're not going to do that. But we did recently start kind of an affiliate kind of thing, where we started having resellers. I guess reseller might be a bit more glamorous than this program actually is, but we do have a very down-low kind of referral system that we've been... I don't know. It's been, I don't know, four or five people, I think right now. It's early days, but it's been interesting to see people who get to make money off of us. Right? That's kind of cool.

Starr: It is. I think I remember at lunch at MicroConf last year, you were talking with some people from some company which I don't remember about that. Is this sort of like an outgrowth of that?

Ben: Yeah. Yeah, that was a interesting chat where they were talking about affiliate programs. We didn't actually end up going with a ready-made affiliate program, because I mean, we're developers. We get to build stuff, right? So we built our own. But also, we didn't want it to be something, at least at the outset, something that was really big and that could overwhelm us. So we started off really small and built our own little thing. Basically used sign-up links. We have some affiliate tracking code that lets us know when someone has come through one of our affiliates, and then we just remit a percentage of their monthly or annual billings back to that affiliate. It's pretty simple. It didn't take us too long to build that.

Ben: And so far, we've had one who's done well with it, who's really worked at promoting those links and figuring out what works. We've had a couple of others who have showed lukewarm interest, and then one who's like, "Yeah, I'll do that," and then it's been radio silence since then. It's an interesting experiment. It's still early, but we've had new customers coming in because of at least one affiliate, so it's been interesting.

Starr: Well, that's exciting. It's early days. I guess if you really want to get on the ground level of this, I don't know. Send Ben flowers or something.

Ben: Yeah, hit us up.

Starr: Do you prefer flowers, or cakes or cookies or-

Ben: I'm more of a chocolates kind of person.

Starr: You are?

Ben: Yeah.

Starr: Wait a second. But our new code of conduct says no bribery, doesn't it?

Ben: That is true, but we can still accept gifts. We carved that out.

Starr: Oh, we can accept gifts? Okay. We're like senators. As long as there's no quid pro quo, it's okay.

Ben: Yeah. I wrote our code of conduct this week, and as I was writing it, I was using some examples from other people and their code of conduct. Because I mean, who wants to write something from scratch, right? And what I found out there, they touch on bribes, accepting and giving bribes, and how you shouldn't accept gifts from people that would influence your business decisions, like partnerships or vendors and that sort of thing, right? But as I was looking at this, ironically enough, considering our topic for today, I thought about the GitHub Marketplace work that I just did a few months ago, because as a thank you, GitHub sent us t-shirts. Right?

Starr: Oh yeah?

Ben: Yeah. So in the code of conduct, I made sure I did not say that you could not accept gifts, because I mean, it's a t-shirt, right?

Starr: It's a t-shirt. It's not-

Ben: You got to be able to accept a t-shirt, you know?

Starr: Yeah. I don't think you're going to sell us out for a t-shirt.

Ben: No. No. So yes, we can still accept those gifts.

Ben: If you ever think about doing that, then you just come to me, because I've got some t-shirts that will just blow away anything GitHub could give you. I promise.

Starr: Right.

Starr: I'll double their t-shirts. I'll double the amount of t-shirts you get. You wear a medium now? You're going to get an extra large.

Ben: Nice. I'll remember that.

Starr: Let's see. Is there anything else you'd like to talk about, about the marketing stuff before we sort of tease our RailsConf 2020 event?

Ben: Ooh. No, I think that pretty much covers it. I mean, we've done other random things in the past. We've tried all kinds of stuff, but the things that we're trying now, we'll see how they turn out.

Starr: Yeah. We'll keep everybody up to date. Yeah, we've got an exciting RailsConf 2020 event. It's going to be in Portland, right?

Ben: It is going to be in Portland.

Starr: And we're not going to mention what it is, because we're going to be those people.

Ben: It's going to be a surprise.

Starr: It's going to be a surprise. It's not going to be announced until it's announced.

Ben: But it's going to be awesome.

Starr: It's going to be announced. It's going to be a very Portland thing. If you live in Portland or have spent time in Portland, you'll hear it and you'll be like, "Yeah, that's a pretty Portland thing."

Ben: Right. Yeah. I mean, we had to do something special, because Portland is one of our hometowns. Being a hundred percent remote company, we don't have just one hometown. We have multiple ones. And yeah, we thought like, "Hey. Portland, RailsConf. Yeah, we got to do something awesome." If you are on the fence about whether you want to go to RailsConf, you should totally come because we're going to have something awesome.

Starr: Yeah. And I feel like this is kind of coming full circle too in this nice way, because the first RailsConf I went to with you all was in Portland, and we didn't have the money to buy tickets, so we just hung out in the hallway. And so it's nice to go and throw a little event for people.

Ben: Yeah, yeah. It'll be great.

Starr: That's going to be nice. It's like, "See how far we've come?"

Ben: Honeybadger is growing up.

Starr: Yeah. Let's see. Other announcements. We are always looking for developers who are interested in writing for us. And for more information on that, check out our blog. In the sort of header, there should be a link. I said there was a link last week, and I really thought I put a link there, but I guess I didn't deploy it or something, because the link isn't there. Maybe somebody took it away to gaslight me. But yeah, it'll be there. And yeah, if you like this show, please go onto your favorite podcast place and review it. And if you want to say who your favorite Honeybadger person is, that's all right. We're not going to stop you. Nobody's going to be mad at you if we lose the office pool. All right. I guess that's it. Anything you want to add, Ben?

Ben: Always a pleasure, Starr.

Episode source