This is a transcript of episode 280 of the Troubleshooting Agile podcast with Jeffrey Fredrick and Douglas Squirrel.

Flickr started as an online game, but the company quickly learned that they had much greater value on tap as a photo site, and “unsold” their initial customers instead of sticking to their original idea. By being curious, you can learn what the market really needs, jettison the custom-built “products” that are holding you back, and keep the right customers delighted. Listen for more on the art and practise of “unselling”!

Listen to the episode on SoundCloud or Apple Podcasts.


Listen to this section at 00:11

Squirrel: Welcome back to Troubleshooting Agile. Hi there, Jeffrey.

Jeffrey: Hi, Squirrel. I’ve got to ask you about this, I was looking at your Twitter feed and you used a phrase I haven’t seen you use before, and I really want to hear the explanation. You said “In tech sales, you’ve got to unsell.” What do you mean by “unsell?” Selling I know about, selling what you have I know all about. Unselling, that’s a new term on me.

Squirrel: Absolutely. So here’s the thing that we’re always talking about. If listeners have got any sense of what we care about at all, I hope they’ve got the sense that the thing we’re really interested in is being curious and learning things. And one of the most important things that happens when you build new software is you learn stuff. Typically that you were wrong, and that’s one of the most painful kinds of learning. It means that the organizations that I work with—and it’s not just startups, it’s much larger organizations too—build software with an assumption in mind that this is how users are going to use the software. This is going to be the valuable piece, this is exactly what we’re going to need. And then they discover something completely different that customers actually need. I can give two examples that our listeners will know about. One of them is Flickr, and the other is Discord. Do you know how both of them started?

Jeffrey: Um, no, I don’t. But I’m curious to hear about the connection between them.

Squirrel: The connection is that they both started as video games.


Listen to this section at 01:42

Jeffrey: What!?

Squirrel: Yeah, exactly. The organization started by building a game: Flicker was a game. I’m told that in the Flicker code base, you can still see artifacts of this, that when you want to look at a picture you actually have to go get the monster who is the owner of the picture, because what they did is they built a game and the game involved having pictures of the monsters in the game. And users could create their own pictures. They realized that people weren’t playing the game, they were just uploading pictures, and they said, “Oh, it seems like we built a really good system for looking at pictures.” And that’s where Flicker came from. Discord was similar. I don’t know as many details, but I read recently that they started similarly creating some kind of game, and then they said, “Well, actually people are really using our chat system. Maybe we should use that and be associated with the gaming community,” as they are. At any rate, these kind of sharp right turns happen all the time.

Jeffrey: You know what? It actually wasn’t Discord. It was Slack. They’re both started by Stewart Butterfield. Not only were they both games, they were the same game by the same person.

Squirrel: That one I didn’t know! I think Discord actually was like this too, because I was reading about Discord. Anyway, listeners can correct us. But the point is that lots of software that you have heard about and used and know and love, is software that did not start out being what it is now. And that creates a really significant problem for sales people, because the poor sales people listen to the developers when they first come on to start selling whatever software the developers have built, and they go off and play golf and make PowerPoints and do all the stuff that sales people do, on the assumption that the developers are right. And then the developers learn that they’re wrong. And the important thing that you need to do at that point is not keep selling whatever it is that you started with, because you will wind up with a few sales to organizations or people or whoever it is you sell to that are on the old platform. It would be as if Discord or Slack or Flickr were still trying to create a game, still trying to run it. Thank heavens they killed their games, right? Those weren’t the things that they needed to continue with. Similarly, I have lots of clients who come to me because they have continued with whatever it was that they started with, and they haven’t noticed they’ve made a sharp turn. It doesn’t have to be as violent a turn as going from video game to a photo site. But even if your turn is gentle, it’s still a turn, and it means that you wind up with a software system that’s trying to satisfy what you assumed your software was going to do, the original need, and what the software actually does, where the actual value is. And that’s where unselling comes in. Am I making sense so far?

Jeffrey: 100%. This is actually what I found most compelling about your tweet, because you said “If you don’t unsell, you create a long tail of loss-making ‘products’, with associated misery.” That’s the part that really resonated with me, because so often I come across applications and they have features that essentially fit exactly what you’re describing. There was a feature created with the idea that it was going to be a very major thing seven years ago, and then three big accounts bought into it. One of the clients still uses it heavily, and no one else does.

Not Exactly Mass Adoption

Listen to this section at 05:11

Squirrel: Exactly. Welcome to being a consultancy, being an agency. Because what you wind up with is you’re trying to build a software product, and you have a piece of your software product which exists only for one customer. This happens very often in enterprise sales where you’re selling to big businesses as you were illustrating. It can happen in retail as well. There you have a class of users nrather than a business, who you’re trying to serve, badly, and you’re holding on to them because you think that’s what you’re obligated to do. But in fact, you’d do better to cut them loose or give them an alternative. And that is the process of unselling. I remember a client I had, you want to talk about a long tail? It was literally a 50 foot wall. I measured it because it was so long and on it were all these little cards and every card was measuring what custom thing they were doing for which customer that they had held on to because they had built something custom for them. And of course they had a huge development team, but they had no capacity to do anything new, because they were busy manipulating all these cards. And the unselling process was one in which we helped them, and they did a yeoman’s job of converting those clients all onto something standard. So instead of having loads and loads of custom items, they had to go back to them and say, “You don’t need this custom thing we built for you, this special feature that only you use actually won’t help you. It isn’t helping you. Here’s what you need instead.” That’s a selling process. Salespeople are good at that process. They just need to understand that what they’re selling is not doing something and instead doing something else. And the something else could be “use our standard product,” “use our software,” “hire us to build you an even better thing that actually fits into our product plan and that will actually solve your problem instead of this weird workaround that you’re using now.” Or it could be you’re selling them on not using your product, and you tell them, “Look, this great contract that we have, it’s really nice. Let’s find a way out of it. Let’s find a way for you to get better service elsewhere because it doesn’t make sense for us to continue. We’re losing money. You’re not getting the best service. How can we make this happen?” Salespeople can do that if you help them. If you never tell them that your product does something new and you keep letting them believe that it does the old thing as well as the new thing, You are in for a 50-foot wall of misery.

Jeffrey: Wow. This is fantastic hearing you describe it, because as you do, I think of examples across decades, and the trade-offs and alternatives that people make, because there are cases and business models where doing that custom thing is what you want to do. I remember talking to some people, I was trying to understand why ClearCase like, who is ClearCase? Do you remember the ClearCase version control system?

Squirrel: Very vaguely, but it’s certainly been eclipsed, hasn’t it? Keep going.

Jeffrey: Absolutely. But it was back in the day when you would often have someone whose job it was to manage your version control system, which seems like a crazy idea today. But at the time you’d have a ClearCase administrator, because ClearCase was a version control system that was built on a custom file system. So really you had to sort of design your repository. Anyway, I was trying to understand what was this for? Who would need this kind of complexity? And someone actually had a good answer, which is they worked with a company that made customized software for high net worth individuals-

Squirrel: “Rich people” is the translation of that.

Jeffrey: Yes, exactly. Very, very wealthy people and their “family office.” You know, how they personally manage their private money—or rather their team of people managed an individual’s money. And so there would be value in having something very custom to describe certain property types or something like that, that were one-offs. You wanted to have this modeled, but you didn’t want it as a general feature for all your clients. So there was a sort of common core and then these extra products. And the key thing here was they knew they were in the business of doing custom software. That was their business model. And to have those things custom and one-off? They charged appropriately for it.

Squirrel: Exactly. I hope everybody noticed these are rich people. These are people who have to decide which yacht to take today. You know, that’s the kind of problem they have. So paying an awful lot for their software was not a problem. I suspect most of our listeners do not have the problem of having too much money and having too much resource to work with. They probably have the opposite problem and therefore they do not need a 50-foot wall.

Jeffrey: Exactly. And by contrast, I remember Netflix for a while had a very rich community interaction where you could follow friends and see what they were renting and their queues and their ratings and whatnot. And they just decided, “This is not adding enough value,” and they killed the whole thing. They unsold it to their retail client base. And what you’re describing is something kind of in-between. You know, you’re neither totally one-off, you’re not totally mass-market, but you do need to be curating down like, “What is our actual product? What’s worth us investing in and maintaining?” And just because we thought something was a good idea at one point in time doesn’t mean it remains a good idea out into the future.

Squirrel: There we go. Okay, So I hope we’ve blown listeners minds sufficiently by introducing this idea of unselling. Thanks, Jeffrey.

Jeffrey: Thanks, Squirrel.