Deezer #Startup Lessons Episode 1: Pivots & Product/Market fit

This is a serie of articles on the story of Deezer as I lived it (Deezer is a global music streaming service / competitor of spotify). These articles will hilight key principles that have been theorized to help startups succeed. Please note that this is a retrospective analysis, we did not know about these principles at that time! I was the 12th employee of Deezer in 2008, I built and lead the mobile team there. Deezer apps are now used daily by millions of people worldwide.

Pivot

​Evangelized by the book Lean Startup, the concept of pivot is much important in the startup world. Basically, a startup’s goal is to start from a plan A and find a plan that works along the way. Most of the time, the initial idea / plan “A” fails, but part of it has been validated. Pivoting is the action of changing part of the business (preferrably the parts that do not work!!!) to maximize chances of success.
It’s very similar to the pivot position in basketball or handball: you keep one foot on the ground and you move the other foot. You keep parts of the business model, and you move the other parts.

There are various types of pivot, the most frequent one being the switch from B2C to B2B. A startup struggles to reach success in B2C, but the team has detected that their product was valuable for a category of businesses in B2B. In that case, the startup does a customer segment pivot (there are other types of pivots), by targeting a different type of customers.

Anticipated pivots VS constrained pivots

Anticipated / chosen
Constrained / forced

I’’m full of admiration for startup teams that succeed to anticipate or to choose their pivot, or even when to pivot. For example, I saw WePop pivot 2 times before becoming JulieDesk (great team, they’re doing pretty well and I’m glad I had the chance to coach them at Paristech Entrepreneurs). Each time they decided to pivot, it was far from obvious, but prove to be a good choice over time.
 
That’s not what happened at Deezer.
We did not decide to pivot. It’s just that at some time, the company was about to die if we didn’t change something. And each time, we found another way. Richard Branson once told “Entrepreneurship as a lot to do with instinct of survival”. And yes, we really fought to allow the startup to survive!
Did I mention already that during my 3 years there, there were probably 15 times when Deezer was about to die within months? (sued by Universal Music, fundraising harder/longer than expected, fights between co-founders, …).
Whatever, there’s no rule to success, but not pivoting when it’s needed is a sure way to failure.

Pivot 1 — From Ad Funded website to Premium Mobile Apps

​​Deezer got traction very early. Millions of users registered within months.
Huge traction.
At one time, I figured out that there were roughly 200K registered users per people in the company! (25 of us handled 5 million users)
The revenue model was quite usual at that time: website was accessible for free, and Deezer was monetizing its (huge) audience through ads.
And Deezer was paying a fixed fee to record labels (and the artists through them) for each music track that was listened on the website.
 
Regarding mobile strategy, Jonathan Benassaya (CEO / co-founder) was a strong believer that this was where we really could bring disruption. The apps were planned from the very beginning to provide unlimited downloads, but accessible only within a Premium offer / paid subscription model. And it was already obvious that it could become a great revenue driver over time.
Let’s pass on the technical hurdles we had to cross to make it real, they prove to be nothing compared to the issues with labels and the challenges we faced to obtain the proper deals to launch the unlimited downloads feature. On that part, kudos to Adele Zangs (legal affairs), Julien Simon, Bertrand Habart & Ludovic Pouilly (artists & labels relations), and… Spotify!
Indeed, I really think that this revolution of music industry happened because there were 2 actors fighting against the issues and the status quo (eg CDs, iTunes, piracy…). When one actor negociated something, the other could follow on and negociate something else.
I’m not sure Deezer alone or Spotify alone would have succeeded…
 
In the meantime, Simon Baldeyrou arrived as CFO. Recomputing everything, he quickly came to a conclusion: the ad funded model was a sure way to hit the wall. Ad revenues were simply not covering the costs.
“No issue Simon, we’re about to launch Deezer Premium with unlimited downloads in mobile apps, it’s gonna be great and we’ll make a lot of money soon.”

Nov 2009: Launch, big announcement, big party, huge PR reviews, …​

Failure… but a proxy of Product / Market fit!

​At that time, there were about 10 millions registered users on Deezer. Quite quickly, we succeeded to reach the milestone of 10K paying users. And it stalled. It didn’t grow. At most could we reach 15K subscribers…
It was a failure and Deezer was still about to die within months.
 
However, we also quickly discovered that the issue was not related to our product. Most people were simply not willing to pay for music, and kept going with our main competitor: illegal music download / piracy.
 
Product / Market fit is a much important concept in the startup world. Basically, it’s a virtual milestone that validates that a startup has found the right product that satisfies its market. The startup can find people that love its product, adopt it, stick to it and pay for it. Then the startup can grow.
An interesting way of figuring out if P/M fit is reached or not for a startup during the early phases is to take a sample of power users and simply tell them that “it’s over, we have to stop our service”. If 30% or 40% of them yell and scream and protest into the streets, it really means you’re onto something.
For the 10K subscribers of Deezer Premium offer, we discovered that 86% of them made it their main music consumption channel. 86%!
 
Unfortunately, it was not a real Product / Market fit (hence the term “proxy”), because PM fit involves that (enough) people are ready to pay.

​Pivot 2 — Switching to B2B2C

​During our market research, Clément Cézard (CMO) and I had spotted an interesting mobile music company called Omnifone. They were doing B2B with carriers, and provided them with a white label music platform. And they were making millions in revenue!
However, since they did not have a B2C business, their product was not user centric, and it relied on DRMed downloads on aging mobile standards (eg, it sucked!).
 
Because we were user centric, our product was far better.
But there are strong challenges to be able to work with carriers and convince them to work with you. At that time in France (beginning of 2009), iOS and Android were just starting to grow, so the most challenging issue was being able to support a huge range of mobile phones, including Blackberry, Sony Ericsson, Nokia, Samsung, … Most of these phones were J2ME phones (Java phones), and along with Android expertise, this was the core expertise of my team — namely Florian Dauly (Mobile Expert), Matthieu Gorvan (now VP Engineering) and I — indeed that’s when I worked on my last piece of ‘production’ code!
Despite strong expertise, it was still a huge technical challenge to run Deezer on mobiles like a Sony Ericsson W595 or a Blackberry Curve 8900, but we succeeded. iOS was at that time developed with Sophiacom (Florian Galanti arrived later in 2009 to internalize iOS).
Needless to say we also had a team of rockstars on platform side, with Daniel Marhely (CTO / co-founder), Aurélien Hérault (probably one of the best innovators France has ever made), Guillaume Pelat (Backend rockstar) and many others.
 
So, we were set up and ready to partner with carriers. But would they want to partner with us?
SFR was ready to give it a try. They decided to provide Deezer as a loyalty incentive. But only for 3 months, then user would have to pay.
User’s feedback was clear: “It’s really great, but I’m not ready to pay for music. Could it be included in my mobile subscription?”
Trial with SFR was sort of a failure (again), but we learned a lot (again).
Also, we had now really good assets to demonstrate Deezer was a very serious mobile player — not only a web platform. So we applied for the Global Mobile Awards at Mobile World Congress. And Deezer got awarded “Best Mobile Music Service 2010”.

At that time, this award was something really serious in the mobile industry, because the jury was made of key mobile actors. And it’s a very good asset to convince carriers that you’re a serious mobile player and you can work with them.
A few months after this, discussion with Orange got really interesting — orchestrated by new CEO Axel Dauchez.
Orange decided to replace its whole existing music offer with Deezer, and include Deezer Premium into their mobile fees, starting with France.
The cool part? Since it started with Orange France, it was decided to adopt a grey label approach “Deezer Premium with Orange”, rather than the Omnifone white label model. Deezer brand was highly valuable, so it was kept.

Execution was orchestrated by Jeanne Bitker (Marketing / branding rockstar), and Orange provided huge distribution and marketing, in fact all you can imagine: offline media, huge presence in Orange agencies, events, street marketing, TV ads, …
Strategically, the benefits for Orange were both for acquisition (having unlimited music with your carrier’s fee is quite attractive!) and for retention (if Deezer with Orange becomes your main music consumption channel, you won’t change for another carrier).
Deezer quickly reached 500K subscribers, making it a suitable business. Deezer strategy has since been to deploy the premium offer in other countries with a local carrier (not necessarily Orange) or a key local partner, in a B2B2C distribution model. There are now millions of people using Deezer apps daily worldwide, and directly or indirectly paying for music. ​

More reads:

2nd episode is about growth hacking / the growth hacks that made Deezer successful: https://medium.com/@AlexJubien/deezer-startup-lessons-episode-2-the-growth-hacks-330b3f951756