035 – Spotify bugs bite back

Posted by  Anecdote International —November 27, 2018
Filed in Business storytelling, Podcast

People tend to be resistant to change. So, what is the best way to keep people interested in a new product? This story suggests finding a balance between new and familiar features is the answer. 

man listening to his phone

It’s now Shawn’s turn to share a story. The story he has chosen suggests the success of new products relies on finding a balance of new and familiar features. Spotify’s ‘Discover Weekly’ is a personalised music playlist of 30 songs, chosen based on the user’s listening patterns. Matthew Ogle designed the playlist so that it would only include songs which the user had never heard before, but it became popular as the result of a bug in the software. The bug allowed some songs the user had heard before to be included.

The story contains a number of meaningful business points, including: people are more receptive to change where there is familiarity; don’t put people too far out of their comfort zone; and mistakes can lead to new, important insights. 

For your storybank

Tags: bugs, familiarity, insight, mistakes, product development, software

Matthew Ogle spent 10 years developing and perfecting a music recommendation engine before he received his big break. In 2015, he got a job with Spotify. His music recommendation engine would be applied to tens of millions of users. 

His first task was to create a product called ‘Discover Weekly’ – a playlist of 30 songs which the user had never heard on the platform, but which were selected based on the user’s listening patterns. The playlist was automatically updated every Monday morning.

After Matt and his team had released the product, they realised there was a bug in the system. Some songs that users had heard before had slipped in. 

They rushed to fix the bug so that the playlist only contained songs which were new to its users. When they did, the product’s popularity plummeted. 

They reintroduced the bug, and discovered in the process that their users desired some familiarity. The product’s popularity rose again. 

Podcast Transcript

Shawn:    

Welcome to Anecdotally Speaking- a podcast to help you build your business story repertoire. Hi, I’m Shawn Callahan.

Mark:   

And I’m Mark Schenk, and Shawn your turn this week. What story have you got for us?

Shawn:

Well what I want to do is put this is a bit of context of what might happen in an organisational setting. So, imagine you have one of your team mates who has just come up with a product and everything about that product is new.

There are new features, new functions, it’s going to be totally unfamiliar to the end user. I know this is a bit of a risk, so this is the story I could potentially tell just to temper if you like, the approach. So, I could sort of say to Bob, let’s call them Bob.

Mark:         

So, Bob has kind of fallen in love with his idea.

Shawn:     

Yes, and he’s a top, smart guy who’s doing good work, I just need to tweak his thinking a little. So, I might say, ‘that’s fantastic work you are doing there but just be a little bit careful about how many new things we provide to our customers.’

‘Cause there’s a guy called Matt Ogle who for ten years had been working on perfecting the music recommendation engine that sits under all those great platforms like Alpha music and Spotify and he’d been really started honing it and then in 2015 he had his great break to really apply it to tens of millions of users and that was when he joined Spotify.

The first job he had was to create a product which they called Discover Weekly. The idea behind Discover Weekly is that it would be thirty brand new songs, everyone different and never heard by that user before, presented to you on a Monday morning.

Mark:  

But based on your listening history?

Shawn:    

That’s right. They go you know those songs, I’m not going to show you any of those songs, and you’re going to listen to these.

Mark:

Completely new.

Shawn:   

Totally new. And he thought this is great, so they got it out there with lots of listeners and as they did, they realised there was a bug in the software and some songs had slipped in that the users had actually heard before. So, they thought, gee we have to fix this bug, so they come back and do the software development and fix the bug so it’s only new songs. When they did that, they noticed the engagement of that playlist just plummeted and all of a sudden people weren’t as interested.

Mark:     

So, Discover Weekly started dropping off in popularity.

Shawn:   

That’s right, exactly and so they looked back at it and they realised that those few songs which were familiar just created a bit of trust in those songs so the other new songs, you could feel them in a sort of comfortable way. So, they reintroduced that bug into the software and off they went, and they had a great music recommendation engine.

Mat Oval has now gone to Instagram which is interesting to sort of see what’s going to happen there. But this idea, this principle of don’t have too much, Bob. It’s been known for some time, take Raymond Loewy, the great designer in the US, he had the idea of Most Advanced Yet Acceptable MAYA acronym.

He knew that if he was going to come up with something new, he had to have stuff in it that was known and acceptable and familiar. He did things like lots of the great logos in America. Crazy things like he fitted out and designed all the elements to Air force 1- Robert Loewy. He did trains, I mean this guy was a designer extraordinaire.

‘So, Bob, in your design maybe what you should do is just find a few familiar things to put in so when people see this new product, they can sort of jump into the new stuff but also feel comfortable, feel safe, that they can trust what you’ve got there.’

‘Yeah Shawn that’s a great idea; do that, so that’s how the work really works, is it?’ Okay that’s my story.

Mark:     

Fantastic, so I liked the way that you set that up in a business context because we haven’t done that in the past with the stories. We’ve kind of just launched straight into them and talked about it later, how you can use them, but I think that’s something we might do a little bit more.

Shawn:    

Actually, we should ask listeners if you like that as an idea to give us feedback. Pop some messages on the comments section in the podcast that would be great.

Mark:

Indeed, so we get some idea of whether contextualising it in a business scenario adds value in terms of understanding how to use that story.

Shawn:  

Yeah, but in terms of elements in that story, what worked for you?

Mark:      

I guess the insight that came from the mistake because we all make mistakes. Making mistakes is a familiar thing although I was a bit surprised that there was a bug in their code.

Shawn:    

What, you haven’t worked in the software industry? There are bugs every day.

Mark:

And the surprise that came from the mistake, the lesson, the surprise and of course one of the things we know is that it’s not a story unless there is a surprise. If that story had just been, ‘and Discover weekly was very successful’, it’s not a story

Shawn:   

Something unanticipated has to happen.

Mark:    

Yeah.

Shawn:  

And in fact, I’d go one step further and say without a surprise it’s not interesting which I suppose it’s very similar and that’s where the insight comes.

There are so many great accidental discoveries in science in the world, so I think it’s one of those as well.

Mark:    

Another thing about the story that works is that most people are familiar with Spotify. I would imagine a lot of people would have experienced Discover weekly.

Shawn: 

Yes exactly.

Mark:  

So, it adds to the relate-ability of that story.

Shawn:    

I think there’s a nice thing too there about the fact that a general principle comes out of it that you can apply to a whole range of situations. It’s not like you are providing a very narrow band that that story only applies to that thing. You’ve actually come up with a story that has this big principle in it to be applied to just about anything.

Mark:  

When we communicate, we’ve got choices about how we communicate those principles. You could just say to Bob, ‘Look Bob it’s really important you have an element of familiarity about the product that you are producing so have a bit of a think about that’.

Shawn:

Yeah, I agree and that would be actually quite common.

Mark:   

It would be, but poor Bob would walk away going, ‘I’ve no idea what you are talking about’.

Shawn: 

No, exactly.

Mark:

He’d have to interpret

Shawn: 

Or the research says you have to have something in there that’s familiar, Bob. And again, he wouldn’t know what that meant.

Mark:     

Absolutely, and that takes me to another thought as to why that story works. It’s of course the credibility that comes from having the names in it. Just a very simple thing gives the story credibility.

Shawn:  

And how fast does it take these days to whack that into Google; find the Wikipedia page, find out a little bit about Mat Ogle.

Mark:

Find out about Sir Raymond Loewy

Shawn:

Exactly, so it’s got all those elements to it. I gave one scenario of how to use that story, are there any other ones that jump out for you in how you might be able to use that story?

Mark:  

I guess one of the things that happens all the time is change in organisations. So, when people are involved in a change programme you might say it’s important to retain some element of the familiar. Don’t put people too far out of their comfort zone.

Shawn:    

It’s so true. I actually notice this a little bit when we are designing strategy stories for organisations. You must have had that situation where you are trying to talk about the past and your tendency is to say the past was bad and now, we are going to go into this future which is good.

But the reality is whose people in that organisation were there during the past so if you say just the past was bad blanket statement you are more or less saying to those people, you’re not worth your salt. Whereas if you can find some familiarity, things that worked things that really set the pace for them that made them different and keep that in the new world that you are creating, you get that familiarity.

Mark:

It’s not very good to stand up in front of an entire organisation and dis the whole place.

Shawn:  

It doesn’t get you very far does it?

Mark:

That’s not a good start to a change initiative.

Shawn: 

That’s good. I’m glad you mentioned that. I saw an article just this week talking about that, about how important it is for change initiatives to have something familiar and something that people can work off.

Mark: 

It’s context. People need context and if you put them in a situation where they have no context it’s very difficult for them to operate effectively, at least for an initial period. And you’re going to cause a lot of discomfort.

Shawn:

This may not be totally relevant but let me throw it out there.

Mark:      

Something from left field.

Shawn:

I caught up with a friend of mine from university (I did a degree in geography & geology).

Mark:

Tidal movements in the Auld River.

Shawn:   

Yes, that was my honours thesis, but I did a lot of work in archaeology, digging things up and the tedious elements of documenting and all that. My friend Gideon was saying that (talking about context) it’s interesting going to a national park to see a cave painting.

And the sign more or less interprets the cave painting at the beginning of the walk before you get to the cave painting. And now, you’re separated from the signs so you’re looking at the cave painting and it makes no sense to you because the context is in the sign, which is 100 metres up the track in the other direction.

He said the best interpretation is where you have the sign next to the cave painting. It’s just another example of what you’re saying around how important context is. You’ve got to have that context right there, built in. Anyway, that was a nice reminder of my days in archaeology.

Let’s give this a rating.

Mark:   

O.K. I’m going to give this a 6.

Shawn:   

Oh wow, that’s lower than I expected. What’s holding you back on this one?

Mark:     

Actually, no, I’m going to give it a 7. I think the addition of the Raymond Loewy MAYA principle makes it much more usable. I can see myself retelling it adding that piece in. One of the things is I can’t remember the name of the guy and I would need to do some work to remember the name of the Spotify guy.

Shawn:   

Just think of a doormat with 2 big ogling eyes looking at you.

Mark:   

Mat Ogle—mmm, we’re wired for images, not words.

Shawn:      

I would probably give it a 7 as well. It’s a nice little story for a very specific situation. I don’t know if it would actually change Bob right there and then. I rarely see a story in one fell swoop change people’s minds; sometimes it’s multiple stories.

Actually, that’s not true; I have seen stories doing that. I look at Mark’s face as I say that. But more often than not I find when somebody is really entrenched it requires a few stories, but this is a good one to have in your back pocket.

Mark:

I agree with you when it’s a matter of a really strongly entrenched story. If there is a strongly entrenched view it can be difficult to shift. But I’ve seen a lot of situations where a single story has changed someone’s’ mind. Sometimes they don’t even know that’s happened and they just move on.

Shawn:  

Yeah, that’s good. Fantastic—that’s a good place to wrap up things. Thanks again, guys, for listening in to Anecdotally Speaking and tune in next week for another episode of how to put all these great stories to work.

About  Anecdote International

Anecdote International is a global training and consulting company, specialising in utilising storytelling to bring humanity back to the workforce. Anecdote is now unique in having a global network of over 60 partners in 28 countries, with their learning programs translated into 11 languages, and customers who incorporate these programs into their leadership and sales enablement activities.

Comments Off on 035 – Spotify bugs bite back

Blog