The Code and Bootstrapping Podcast

Using Heatmaps and Surveys

(When your audience is tiny, you need a microscope)


[00:00:00] Hello and welcome back to the Code and Bootstrapping podcast. This is Mark and today I'm going to talk about heatmaps and surveys.

[00:00:11] In the early days of Alchemist Camp, I started off with just a YouTube channel and then my next step was to make a single-page Web site. And I don't mean a fancy JavaScript framework. I mean a single page of HTML with basically, just whatever was spit out by It's a thing that turns markdown into HTML. And, I just had links to the YouTube videos which were screencasts that teach the Elixir programming language and I was getting a handful of visitors a day. I wasn't offering any paid product. It was just a YouTube channel with free resources for learning a niche programming language and then a web site that listed those videos that I had made and linked to them.

[00:00:55] Since I only had a few visitors per day, there wasn't really any way that I could use Google Analytics and figure out what people were looking for. Google Analytics stopped showing what search result people use to arrive at your page a long time ago and with just one page on the site, there really wouldn't have been anything interesting to go off of. So, I used a site called Hotjar that has a number of tools for marketers, and the two that I used were surveys and heatmaps. It was totally free. So, a heatmap, basically, lets you see how your users interact with your web site.

[00:01:33] You can see their mouse moving around, you can see what they try to click on or what they do click on, and you can see how long they stay on the page before they leave. With just a few visitors, well, that's a little bit of information at least. So, I set up the heatmaps. Then after a week or so, I could see where people had clicked on the page. I could also watch someone go to my page, sit there idly for twenty five seconds, then scroll down to the bottom quickly and leave or whatever, you know, each individual user did. But that's pretty boring and not very informative most of the time. But the heatmap, just seeing like how many times people clicked and what areas, you know, they'd be like light blue, if it only got a few clicks... they'd be brighter if they got more clicks.

[00:02:20] I saw that people were clicking all over on a few things on the page. They were clicking on specific topics because I had the videos all on a table and for each video I had the name of the video and then next to it, a link to that video on YouTube, and then next to that, a few topics that were covered in the video. And I could see people really clicking the heck out of the topics because, you know, maybe they thought it was navigation. Maybe they wanted to filter down to just the videos that are teaching how to use regex or how to, you know, how to do pattern matching in Elixir.

[00:02:56] That gave me some ideas of what people might want to learn. And once I had a few ideas, based on where people were clicking, I made a YouTube video where I basically said, "Hey, I'm thinking about building out some more things in Alchemist Camp and maybe making a premium offering. I want to know what you're interested in and please go take this survey. It only has two questions. You can do it in 10 seconds, literally, and then if you're interested in whatever future thing I make that might be premium, leave your email."

[00:03:30] And, the results I got from the video were fantastic. At least in the first few weeks of the video being out there, I believe only 50 people saw it at all. And I got more than 20 responses. I got, I think 13 people who signed up to be notified when and if I launched something premium. Actually, a lot of those people ended up converting a couple of months later when I did. I believe I got nine out of those first 13 people. And this was from a YouTube video that was initially seen like 50 times. So despite it being a tiny audience, the conversion rate was fantastic.

[00:04:10] So what did I do to make that happen? Well, I think with the heat maps, it's really kind of a gut feeling like you just want to see what people do, see what people click on... A lot of times people will click on things that aren't links, and that's kind of an indication that they're interested in what's going on there. After getting just some... basically the heat map will give you some questions to ask and then after getting those questions, then you can make a survey. And what I like to do, what I did do, is I only asked two questions.

[00:04:46] The first question I asked was, "what topic would you like to see recorded first?" and these topics were all things related to learning the Elixir programming language and the popular frameworks that go along with it. So, one of them was the Web framework, Phoenix. One of them was a database wrapper Ecto and then other topics were distributed OTP apps, Phoenix channels, WebRTC, Elixir protocols, short explanations of language features in general, or bigger projects, or one final option, which was "other".

[00:05:23] And then the second question I asked was, "of the following features on the site that I'm building out, Alchemist.Camp, which would you like to be prioritized?" and I gave them the options of source code, like being able to see the source code from the screencast on the site, recording videos faster, organizing lessons by topic, or creating summaries for each episode. And actually all of those have been implemented, except maybe recording videos faster, although the videos have definitely improved in quality.

[00:05:57] So, this was just two questions. It was just "pick one thing you want out of the content, like what content you want most" and then "what site feature do you want most?" So someone could literally go to the survey, just read the options really quick, click, click and then hit submit and they're done. There was one third field, which was "any other requests for comments?", and that was mostly there just so that someone who clicked "other could answer what they did want me to do. But, I got a lot of random feedback there that was... it was pretty useful and pretty insight,ful for only having 20 people fill out the survey and for only having about 50 people see the video at that point.

[00:06:41] Looking at it now, over a year later, there is still only 112 people that have ever seen that video. So, it was one of the highest return activities I've done. Just a quick, six minute video, with a request for feedback and that request was so easy that almost half the people who watched the video did it and about a quarter of the people who watched the video signed up to be on the early bird subscriber list and most of them ended up buying. So, it was really a great way to get some direction and get an idea of what my very most engaged students wanted to learn and what I could provide them. And obviously, this isn't limited to an info product business.

[00:07:30] There are lots of things you can start up where you've just got a tiny, tiny, tiny amount of traffic at the beginning and you can use a survey to get meaningful feedback from a group of people that are too small to use traditional marketing analytics to understand. And you can also use heatmaps if you start with such a small number of people that you don't even know how to figure out what to ask on the survey. So... it was really useful for me and it's definitely something I'll do again in the future.

[00:08:03] One possible stumbling block that I would watch out for that I've actually seen a few of my friends hit is, you don't want to ask too much. Like, you really want it to be something where someone can do it with very little effort, because the more effort it takes, the fewer people do the survey. And worse than that, the people who take the effort to do the survey aren't going to be the same kinds of people who saw the survey in general. So you might have, you know, like say a 12 question survey where people actually have to type things in, in some parts, as opposed to just making multiple choice selections.

[00:08:42] And if it's enough of a hassle that, say, only 1 percent of the people to see it finished the whole thing and submit it, well, that 1 percent you're seeing is going to be the people who have some super strong opinion about something in the survey or people who just enjoy writing surveys for the heck of it. So, it's not going to be that representative and not that useful. You can do other things like offer incentives to fill out the survey. But again, that's going to lead to a different group of people than your... your viewers, your audience or your customers, in general. So. I really, really like to keep the surveys as short as possible, unless it's something that the users have been already putting a lot of time into or putting money into and like I know that I can get a good percentage to actually fill it out without any external incentives.

[00:09:39] So that's it for this time, and thanks for listening to Code and Bootstrapping.

Show notes

In this episode Mark talks about how heatmaps and surveys were useful in the very early days of Alchemist Camp, how he used them and a common pitfall when taking surveys.


Back to index