Designing Products for Existing Behaviors

The alternate title for this post is “Lessons from Mixel That We Applied to Kidpost.” It’s longish.

The best bit of feedback that we’ve gotten to our Kidpost public beta is the recognition that Kidpost aims to be extremely compatible with—even deferential to—existing user habits. When people express their delight with that fact, it makes us feel great. This is true for me particularly, because it directly relates to one of the hardest lessons I learned from my failed startup, Mixel: Asking users to adopt new behaviors or even modify their existing behaviors is very, very hard.

At Mixel, one of the things that doomed us was the very thing that infatuated us so much from the outset: the amazing possibilities inherent in creating a wholly new kind of content that merged aspects of visual collage, photo sharing and remix culture. We thought it was a brilliant idea, obviously, but the cold truth was that only a small number of people agreed with us. Actually, it’s more accurate to say that a majority of people just didn’t see how it fit into their own lives. The things we asked people to do with Mixel were fundamentally foreign to the way most people conducted their days, to the way that they were already using their devices, to their existing habits.

I don’t mean to say that you can’t change user behaviors, or that you can’t engender new ones. Clearly, you can if you build a “what the world was waiting for” kind of product. Instagram is one of the best recent examples of that, in my opinion. It identified a sweet spot at the intersection of social sharing and taking pictures with your phone, and it turned millions of people into iPhoneographers, or whatever you want to call them. That insight produced a new behavior, one that many people have taken to with obsessive dedication. It’s a major achievement, but it was a behavior that people were practically primed to adopt.

It’s also entirely possible to build products that change the way people think about the world, and if you’re successful millions of people will change their habits to complement what you’ve built, instead of the reverse. This is what Twitter did, though of course that story is much more complicated than just dropping a huge, epiphanic truth bomb on the masses and having everyone suddenly recognize the life-altering benefits of tweeting, retweeting and hashtagging.

Products that change user behaviors, which is what Twitter always has been and in retrospect is what Mixel was trying to be, require tremendous effort. They almost always take lots of iteration, lots of manpower, and lots and lots of money. Actually their success tends to be just one part product and, say, nine parts everything else that makes for a great company, particularly marketing. That’s not to take away from the intrinsic importance of the products themselves, but rather to emphasize how much credible, widespread and persistent persuasion is necessary for people to change their existing habits.

Twitter has always been an amazing if imperfect product, in my view, but among other things it benefited tremendously from its outrageously successful marketing—the countless hashtags splayed all over televisions and movies, the endless number of celebrities hawking their Twitter accounts, the innumerable ways that the media found to constantly remind us that Twitter was changing our lives. That phenomenon had its roots in genuine merit, it’s true, but it was not entirely organic either. Twitter-mania was the direct result of millions of dollars of venture capital and untold man hours dedicated towards convincing us to give Twitter a try, and if it didn’t take the first time, then to give it another try, and another, and another, until the service had amassed hundreds of millions of users who had all decided to modify their existing habits to fit Twitter into their lives.

Mixel just wasn’t ever going to become that kind of phenomenon, as much as my acute case of Founder’s Tunnel Vision insisted that it would be. There were many reasons for that, but one of the biggest was that in order to effect that kind of mass market change, we would have needed to think about Mixel at a much bigger scale than we did. We should have raised much more money and amassed much more talent much more quickly, and we should have spent a lot more time thinking about marketing and how to change user behaviors, how to convince everyone that Mixel was something they needed to incorporate into their lives.

I’ve spent a good deal of the time since Mixel shut down licking my wounds and trying to figure out what lessons I could take away from the experience. Mixel’s failure never dimmed my desire to make products but it did turn my attention towards ideas that can enhance things that people are already doing, because to build something like that, you don’t necessarily need tremendous amounts of capital and marketing and manpower. You need to work really, really hard still, of course, but your job becomes significantly easier if you build things that complement use cases that already exist in the world.

So when I hit on the concept for Kidpost and shared it with Matt and Mike, we spent a lot of time honing it down to just the bare essentials; our guiding principle was to ask for the smallest changes in our users’ behaviors as possible, if any at all.

That meant Kidpost should be a “set it and forget” kind of application. We wanted a very easy setup process that was an end in and of itself. Once users completed a brief onboarding process, we didn’t want them to ever have to return to the Kidpost interface. They should be able to use it every day and enjoy its full potential without having to come back to the site even once.

We also wanted our users to be able to keep posting their pictures exactly where they were already posting them (for now that means just Facebook and/or Instagram, but support for more services is coming soon). The only modification we were willing to ask for, and even this made us a bit nervous, was for users to add the hashtag “#kidpost” to their pictures.

The hashtag drives the Kidpost emails, of course, and for the email subscribers we imposed on ourselves an even stricter limit on required technical expertise—if you’ve ever played the role of tech support in your family, you’ll sympathize with this. Kidpost doesn’t ask your friends and family to install new apps, doesn’t force them to check new sites, streams or feeds, and definitely doesn’t require them to create, manage or remember new passwords.

Once we had established this basic framework, it freed us to build something lightweight, allowed us to focus on real user goals, and helped us articulate a clear, sensible roadmap without a lot of detours and distractions. We also felt like we were free to bring something into the world that didn’t bear any obligations to change the whole world, or to scale at breakneck speed, or to satisfy the demands of investors—we’ve bootstrapped everything for Kidpost.

All of this was particularly satisfying for me after the highs and lows of Mixel because it was a kind of anti-Mixel: something deliberately small and that real people that I know will actually use. Well, that remains to be seen of course; we’re working our way through our beta right now (you can try it out at and will get to an official launch within a couple of months…which is when I suppose I’ll figure out if all of these theories I’ve developed have any truth to them at all.


Do you have thoughts about this post you’d like to send Khoi? Your remarks will not be published.

Thank you! Your remarks have been sent to Khoi.