Jean-Charles Samuelian-Werve
Co-founder & CEO @ Alan
22 déc 2020Jean-Charles' Newsletters

JCNews #51 - Netflix et ré-invention.

Chers amis, fans de bonnes lectures,

Bienvenue aux 96 personnes qui nous ont rejoints depuis mardi dernier ! Si vous lisez ceci et n’êtes pas encore inscrit, rejoignez vite les 2 7551 personnes qui ont déjà eu la bonne idée de le faire. C’est juste ici !

Cette semaine encore, je vous propose mes lectures les plus intéressantes, en mettant notamment l’accent sur un article principal et en vous partageant mon avis dessus. J’espère que cela vous sera utile.

Ecrivez-moi pour en parler, sur Linkedin ou sur Twitter. Bonne lecture à tous !

JCN#51

💡Must-read de la semaine

Chaque semaine, je publie un must-read, que vous choisissez.

👉 Designing a Culture of Reinvention with Netflix (a16z)

On how to receive feedback:

Getting feedback, and the pain, is like doing crunches or push-ups. You want to stop, you know it hurts, and you know that it’s the painful ones that make you stronger.(...) If you contain your ego, and if you can take the pain, you’ll get stronger, and you’ll get better as a leader.

Then instead of arguing with the person, I’ll say, “Tell me more? What else?” And just keep hitting those two, “Tell me more? What else?” And it’ll be amazing what comes out. It really hurts, but that’s what makes you better.

Recevoir et donner du feedback est parfois difficile : il faut savoir dire la vérité même si cette dernière fait mal, et il faut surtout accepter de l’entendre sans ego (et en demander plus). Un bon feedback a pour objectif d’aider les gens à grandir, pas relâcher votre énervement.

When people are talking about other colleagues, which is normal and fine, just keep pressing them with, “Oh, and what did they say when you asked them about that?” That stimulates directness.

Quand vous entendez vos collègues parler d’autres collègues, interpellez les : et quelle a été sa réponse lorsque tu lui as dit ça ? Il faut encourager la transparence, l’échange direct, c’est la meilleure façon de s’améliorer et d’éviter les frustrations. Cela peut être flatteur d’être la personne qu’on vient voir pour résoudre les problèmes, il faut éviter cela à mon avis et donner les outils aux gens pour qu’ils se parlent directement.

Optimising for long-term innovation:

Most companies over-optimize for efficiency. They want to get so good at their current market that they lose flexibility to adjust to the future. And the non-intuitive thing is it’s better to manage chaotically, if that’s productive and fertile.

You have to be very conscious as a leader how you’re optimizing for long-term innovation. Manufacturing has dominated the economy for 200 years. There’s this big influence from manufacturing, because it’s generated most of the economic wealth of the past couple hundred years, around the boss and the worker and the worker following the rules. You want zero variation. That’s nirvana.

And yet, if you’re an innovation culture, variation is essential.

Beaucoup voient l’absence d’imprévus comme le graal, du fait d’un biais cognitif assimilant cette situation à une bonne santé de l’entreprise. En réalité, savoir s’adapter, changer de cap, est essentiel et fertile.

On n’aurait pas fait face à deux confinements sans cette capacité d’adaptation :)

Google 20% did not work well:

And Google is a fascinating one because, of course, 10 years ago they were all about 20% time, and now that’s all gone. What did they learn? It’s pretty hard to do innovation one day a week

Il est difficile de quantifier le temps passé à innover, qui idéalement est constant, même dans les plus petites tâches. Chez Alan, on a préféré pousser l’équipe à aider à définir la stratégie, et avoir des hackathons réguliers qui permettent l’émergence de petits projets qui deviennent souvent gros.

What markets to approach:

I have always thought you want to go after the smallest market possible that can hold your 5- to 10-year growth ambitions.

J’aime beaucoup cette approche contre-intuitive : optimiser pour des marchés suffisamment grands pour nous permettre de croître et d’apprendre, plutôt que pour le plus grand tout court. C’est ce qu’on a fait pour l’Espagne et la Belgique, et ce qu’on va tester avec les nouveaux parents.

How they did the hard choice to move from DVD to streaming (and how they failed the landing):

So, partially goaded by the wonderfully radical Marc Andreessen who was like, “burn the boats,” we came up with this plan to separate DVD and streaming.

We made one tragic mistake in it. The pricing for the combined plan had been $10, and the separate plans we set were $8 for streaming, which was about Hulu’s price, and $8 for DVD. It was effectively a 60% price increase.

We sent an email to 20 million American families, so 20% of American society, on one day, saying the price is going up by 60%, and you get to use two services, two websites instead of one. Less convenience, radically higher price in the middle of a recession. (...) And no new feature. We were right to separate them. It’s just we did it an awful way.

On peut avoir raison sur le fond, et être mauvais sur la forme. C’est un exercice très périlleux, surtout quand on est convaincu de faire la chose “juste”. Se pencher sur les “second order consequences” (les conséquences des conséquences) est un outil intellectuel très utile pour imaginer comment les autres peuvent réagir à nos actions.

How did it happen?

Our leading people were too deferential. They were like, “Reed’s been right so many times. I think this is going to be bad, but he must see something.”

They didn’t know that each of them felt the same way. In hindsight, if we had said, “All execs, write down your level of confidence in this move, disaster to genius,” it would’ve come up that 20 people think it’s going to be bad. With the strength of their shared feeling, they would’ve realized, “No, we’re right. Reed is wrong.” Then, probably, we would’ve stopped.

We said, in the future, all major decisions, we will have everybody write down what they think +10 to -10 and why. We do that in a Google Sheet, but any shared medium is fine. Then everyone knows what everyone else thinks in writing. That little device has helped us avoid chaos and catastrophe

Il est essentiel de cultiver l’esprit de dissidence, qui n’a rien de naturel ou évident. Il permet par exemple d’éviter le biais de confirmation lorsque le CEO ou autre employé ‘senior’ défend une idée qui nous semble ne pas être la meilleure pour l’entreprise.

The importance of focusing on the right thing:

When we were in the DVD fight with Blockbuster in ’05 to ’07, we got distracted doing magic charms.

We did four things to make ourselves feel better. We went into selling used DVDs directly on our website rather than on eBay. We started buying some little films at Sundance for original content. We launched a private social network called Netflix Friends, and this was 2005, Facebook was barely out of Harvard. You could see each other’s queue and viewing history, if you gave each other permissions. And we sold banner advertising on our website, like Overstock used to do. It was four nontrivial engineering efforts to differentiate us against Blockbuster.

Fortunately, we also spent some time getting shipping more reliable, so our queue fulfillment rate went from 96% to 98%. In the end, once we beat Blockbuster, we realized the only thing that mattered was that queue fulfillment rate, the 96% to 98%.

We, as leaders, did not have the courage to stand before the employees and say, “We’re going to win because we can move this from 96% to 98%.” We needed these little magic charms to make ourselves feel good. It was totally bad management to get distracted by those magic charms.

Having the confidence to focus on the basics and doing the basics incredibly well is important.

C’est un des sujets les plus durs au sein de l’organisation : comment mélanger innovation et construction des “basiques”. Il nous est arrivé dans le passé chez Alan d’aller un peu trop loin dans le “nice-to-have” avant d’avoir fini les basiques. En parallèle, nous avons aussi vu que quand nous étions concentrés sur les briques fondamentales, les plus dures, cela faisait clairement la différence : comme la reconstruction de notre outil de gestion des sinistres et remboursement qui permet maintenant d’avoir 75% des remboursements en moins d’1h, beaucoup plus d’automatisation, et qui est le socle de nombreuses innovations internes.

About inclusion, and how to start it as early as possible:

A big example would be about four years ago, we added inclusion as a core value, and we’ve been working hard on it, and we’ve made real progress.

I wish that I had led that 10 years ago, but I didn’t. I wish I had led it at all, but it was brought to me as something that really needed to be done. I’m on board, and we’re driving it, but it’s a real improvement because we’ve got more people thinking about it.

Now, of our top 20 leaders, we’re half men, half women. We’re 25% leaders of color in our top 20. That’s been a great improvement that was really driven by all the new talent at Netflix.

Il faut vraiment être intentionnel là-dessus, et je pense que nous aussi sommes en retard, par exemple sur les recrutements d'ingénieurs du fait d'un sourcing plus difficile. On a mis en place de nouvelles politiques de recrutement pour continuer à s'améliorer en 2021.


🏯Construire une entreprise

En plus d’articles triés sur le volet, je partage un principe de leadership d’Alan par semaine. Le même que je partage en interne et à nos investisseurs tous les mercredis.

👉 La qualité est difficilement quantifiable (Healthy Business)

  • La qualité est un concept relativement abstrait, difficile à mesurer. Par exemple, comment faire tenir la qualité dans un OKR ? Difficilement. Et pourtant, le concept existe, et est très puissant.
  • Il y a une part de raison, une part d’intuition. Nous faisons confiance aux Alaners à ce propos, lorsqu’ils construisent le produit : ils comprennent le problème à résoudre, et portent un jugement sur l’exécution. - Cela va de la fréquence des images des animations à la précision des illustrations, en passant par la rédaction, et jusqu'à l'architecture même.

👉 Interview de Larry Page et d’Eric Schmidt en 2011 (Youtube). Ce qu’on y apprend :

  • Be crazy enough to do what the other companies don't dare trying:
    • Android were 20 people when acquired for €50m.
    • Chrome: people were asking why you need a new browser? We have already several existing ones.
  • When you have something growing quickly. You think it's small. And the next day it's huge.
  • Story of their near bankruptcy. The cash restriction program (CRAP) where they were not paying their bills.
  • Change your views often based on the facts.
  • When you grow the company (headcount) you should do more things. Add businesses as you add people.
  • Large companies are their own worst enemies.
  • Good slow decisions don't exist. Only good fast decisions. Avoid at all cost to slow down decision making.

👉 Comment Facebook a lancé deux produits majeurs en 2006(qui étaient très décriés) (Dan Rose)

  • In 2006, we shipped the 2 most important products in FB’s history: News Feed + Open Registration. A lot of smart people thought these moves would destroy FB. Instead, they transformed the company and cemented Zuck’s leadership.
  • The backstory:
    • News Feed shipped first. In 2006 there were no feeds (other than RSS), NF was a novel product idea. Websites were measured on page views back then, and NF was designed to reduce PVs by eliminating the need to click around profiles. Less PVs = less ad impressions, seemed crazy.
    • Mark described FB as a utility, and NF was central to his vision. It showed info you could already see on people’s profiles, but organized efficiently on the home page. And stories would be ranked based on what people found most interesting. This was a massive change.
    • Many users were disoriented to see their personal info organized this way. Within 24 hours after launch, 10% of our entire user base joined a group calling for NF to be rolled back (ironically this group had gone viral via NF). People protested outside our office, I was nervous.
    • Mark was calm and collected, and his conviction never wavered. He listened and made changes in response to feedback from the community. But he also looked at the data, which contrasted sharply with the protests: people were more engaged than ever before.
    • Within a week users had embraced NF, and we turned our attention to Open Registration. At the time, only college students could register for a FB account. 8k new people were signing up each day, while another 30k were rejected for not having a college email. That blew me away.
  • Some of the smartest people in Silicon Valley thought Open Reg would kill the company. We had created a modern day MTV (Viacom even tried to acquire us), and allowing old people would spoil the party. But Mark believed more users would make FB more useful to everyone.
    • 2 months before I joined FB, Mark had rejected Yahoo’s $1B acquisition offer. Everyone told him to take the deal, but he believed News Feed + Open Reg (and later FB Platform) could transform the value of the company. 18 months later, FB was valued at $15B.

🗞Dans l’actu

📱Monde des technologies

👉 Google s’attaque à la verticale paiement sous tous les angles (Casey Newton)

  • Open the new Google Pay and you’ll see the — shall we say — influences from other apps everywhere. There are easy peer-to-peer payments (Venmo, Square Cash). There are simplified group payments for dividing up bills (Splitwise). There are features for tracking your spending (Mint, You Need A Budget). You can even order food (DoorDash). And, per the iron laws of Google product development, there is also a messaging feature.
  • Connect your bank accounts and you can do natural language searches to find past purchases — type in “shirts” or “Mexican food” and Pay will find your relevant purchases using machine learning, even if those words don’t appear on the receipt.

👉 Snapchat lance une loterie d’1 million de dollars par jour pour renverser la manière dont le contenu est créé, et agrandir le pool de potentiels influenceurs (Casey Newton)

  • What Snap is asking with Spotlight, the company’s own take on short-form video, is whether you can build an engine for culture that benefits many thousands of people, rather than hundreds.
  • The company’s idea is to replace public follower counts, likes, and comments with something that more closely resembles a lottery — and at least through the end of this year, the company will pay out $1 million a day. Notably, Snap is placing its incentives at the level of the content, rather than the creator.

🏥 Santé

👉 Amazon va fournir ses services de santé à d’autres entreprises (Business Insider)

  • A secretive Amazon initiative intends to provide virtual and in-person primary care services to employees at other large companies.
  • The healthcare service is reportedly an extension of Amazon Care, a pilot program announced in September 2019 that currently provides on-demand medical care for Amazon employees in the state of Washington. Eligible Amazon employees can use a mobile app for prescriptions, virtual doctor’s visits, and to schedule house visits from medical professionals.
  • Amazon intends to sell a version of Amazon Care to large employers with workers around the U.S.. It plans to charge companies a tech fee, plus additional fees based on the number of employees that use the service each month, the report found.

👉 K Health lève $42 million pour développer sa plateforme de télémédecine basée sur l’IA (venturebeat)

  • Tel Aviv-based K Health closed a $42 million series D round to grow its telemedicine offering, bringing its total raised to date to $139.3 million.
  • “Since the onset of the COVID-19 pandemic, more than one million new people have turned to K Health for their primary care and mental health needs. K Health has grown 1,000% just this year alone”.
  • K Health’s smartphone app, K — which Bloch claims has more than 4 million members, growing at a rate of 10,000 to 15,000 new users a day — facilitates in-app visits from a roster of doctors K users can consult for a fee, assuming they live in one of the U.S. states where service is available. The doctors review an AI-assisted breakdown of the patient’s symptoms and then diagnose, prescribe, or refer the patient as appropriate.
  • K Health users can chat with a doctor for $9 per month, $19 for a one-time session, or $27 for three-month unlimited access. Mental health services are also offered, and prescribed medication can be delivered to users for $34 per month.
  • K sources from a 20-year database of millions of electronic health records (including physician notes and lab results) and billions of “health events” — like nausea, headaches, and vomiting — supplied by Maccabi, Israel’s second-largest health fund, as well as health systems in the U.S. and Mexico. Users start by downloading an app for iOS or Android and answering roughly 20 questions about their age, gender, body mass index, health history, and symptoms. A machine learning-powered backend uses the responses to build a private profile, which it compares to insights gleaned from a corpus of over 400 million clinical notes and charts.

💚Alan

👉 Maintenir l’équilibre entre hypercroissance et bien-être des salariés (La Libre Belgique) : je présente la proposition de valeur d’Healthy Business dans une interview avec Pierre-François Lovens.

👉 Les bureaux en 2030 selon Alan (Deskeo) : interview avec Frank Zorn, fondateur de Deskeo, et présentation de notre cadre de travail unique.


C’est déjà fini. Bonne semaine et à vous de jouer maintenant ! Invitez vos amis à s’inscrire, ici.

PS : pas de JCNews la semaine prochaine, on se retrouve le 5 janvier ! D’ici là, passez de merveilleuses fêtes 🎄❄️

La crème des articles alan

Dans votre boite mail. Garantie sans spam.

Populaires en ce moment

Populaire en ce moment

De la même catégorie