#2 – 🕵🏻‍♂️ Définir et tester le parcours clé de Festview

[Ce billet est le 2ème volet de la Saga Festview – Mettre en ligne un Wikipédia de la musique]

Dans l’épisode précédent, 1. Consolider la vision de Festview , on a vu que :

  1. Lors d’une session exploratoire de navigation, je n’avais pas réussi à comprendre le projet global de Festview car jusqu’à ce jour, il n’avait été présent que dans l’esprit du fondateur
  2. Que les business qui écrivent et diffusent leur vision ont de meilleures chances de réussite ; d’autant plus lorsqu’ils portent attention aux 6 paramètres majeurs d’une vision efficace
  3. Un atelier de 2h nous a permis de rédiger une plateforme de marque, contenant la vision, et de la publier sur le site
  4. Cette démarche avait créé un gain de confiance, de motivation, d’alignement ; ainsi que de nouvelles idées fortes pour Festview

Dans cet opus, je vais parler de l’urgence numéro 1 qui, à mon sens, doit préoccuper un Product Manager lorsqu’il prend un nouveau produit en main : identifier son parcours clé. C’est-à-dire l’expérience utilisateur qui porte la promesse du projet avec le plus de puissance. Je vais raconter comment, pour Festview, nous avons imaginé ce parcours clé, puis comment nous avons mené une grande phase de tests d’utilisateurs pour mesurer l’efficacité de ce parcours.

Le parcours clé 🎵

Après avoir clarifié le positionnement de Festview avec une vision bien rédigée et publiée, il fallait désormais proposer l’expérience adéquate aux utilisateurs.

Nous rêvions notamment de compléter le parcours de l’époque avec une brique supplémentaires : donner des récompenses aux utilisateurs qui gagnent des points grâce à leurs contributions.

Cette fonctionnalité n’était pas encore présente sur le site ; il fallait l’implémenter. Et cela allait prendre du temps à construire…

  • Permettre aux administrateurs de créer des récompenses – 1 semaine
  • Permettre aux utilisateurs de consulter, trier et filtrer des récompenses – 2 semaines
  • Permettre aux utilisateurs de consulter, liker et partager le détail d’une récompense – 2 semaines
  • Permettre aux utilisateurs d’obtenir leur récompense contre leurs points – 1 semaine
  • Livrer la récompense aux utilisateurs – 1 semaine
  • Trouver des partenaires business intéressés pour faire leur promotion à travers des récompenses – 3 semaines

Pour une expérience utilisateur complète et bien faite, ce chantier représente à lui seul une charge de travail de plusieurs mois. Or nous n’avions pas ce temps là.

Story mapping : se concentrer sur l’essentiel 🎵

Nous avons donc retravaillé ensemble la feature de « Récompenses » à l’aide d’un Story Mapping. A mon sens, c’est le meilleur outil dans ces cas là : il permet d’identifier ce qui est vraiment nécessaire pour que la feature fonctionne, versus tout ce qui est superflu. Le but est de garder l’essentiel, et de remettre à plus tard tout le reste.

L’exécution s’est traduite par un atelier de 2 heures :

  1. Introduction au Story Mapping, avec l’exemple de la routine matinale*. Super simple et efficace : en 20min nous étions parfaitement alignés sur la méthodologie
  2. Listing des grandes fonctionnalité et des sous-fonctionnalités du parcours
  3. Réduction du parcours aux fonctionnalités et sous-fonctionnalités indispensables

*Je publierai prochainement un contenu à propos de cet atelier « Routine Matinale », car je n’ai pas trouvé de ressources solides dessus. C’est un atelier bien connu qui consiste (1) à imaginer collectivement une routine matinale parfaite. Câlin dans le lit, café, tartines, journal, salle de bain, choix des vêtements,… (2) puis à la dégrader sous la pression d’une panne de réveil. Il en ressort « un parcours minimal » qui ressemble souvent à « je m’habille, je brosse mes dents et je file ».

Voilà le résultat de notre travail ce jour là :

Impliquez l’équipe 🎵

Construisez-vous les story maps de vos projets de fonctionnalités directement avec votre équipe de développement ?

Je pense qu’il le faut. Peut-être pas pour faire la décomposition de Tous les projets de features qui entrent dans le backlog. Mais au moins pour les features majeures ; celles qui mélangent du back-end, du front-end, du design,… Grâce à cet exercice, les priorités sont bien comprises par tout le monde. L’équipe – ici, l’unique développeur de la boite – peut se mettre au travail immédiatement en sachant exactement pourquoi elle développe chacune des briques de la fonctionnalité. En un mot :

Construire la story map d’une feature avec son équipe, c’est la promesse d’un passage fluide des idées au développement.

Ainsi, mon ami a entamé aussitôt la construction de la feature minimale que nous avions définis ensemble. La semaine suivante, cette dernière était livrée :

Tests utilisateur 🎵

Reprenons. Vision ⇒ Positionnement ⇒ Idées de parcours ⇒ Fonctionnalité ⇒ Dev ⇒ …. il restait plus qu’à savoir ce que les utilisateurs en pensaient.

Grâce aux outils adaptés (voir partie outils plus bas), en un jour nous avons booké 20 rendez-vous avec des proches susceptibles d’être intéressées par Festview. Pourquoi 20 ? La théorie dit qu' »entre 20 et 30 entretiens sont nécessaires avant d’être capable de savoir à l’avance ce que vont dire les utilisateurs à propos du produit. Une fois que c’est le cas, c’est qu’il n’y a plus rien à apprendre, et qu’on peut se mettre au travail.« (Ash Maurya – Running Lean)

J’ai préparé un script d’entretien, comme le suggère la méthode. Avec une intro, des questions et une conclusion. La répétition schématique du même script à chaque interview favorise la détection des propos intéressants des utilisateurs. J’ai également préparé des outils permettant de réceptionner correctement toute la masse d’informations que nous allions recevoir.

Le contenu du test utilisateur était relativement libre. Voici ce que nous souhaitions obtenir :

  • un ressenti général de l’utilisateur sur le concept. Que pensait-il de l’articulation entre encyclopédie, recherche de contenu, contribution, points, récompenses,…)
  • un ressenti sur la fluidité de la navigation
  • une détection des bugs les plus récurrents et les plus embêtants

Pour obtenir ces informations, nous avons opté pour une navigation semi-guidée. L’utilisateur était libre de ses mouvements sur l’app, et racontait à voix-haute ce qui lui passait par la tête, et de temps en temps, nous demandions d’aller spécifiquement sur une fonctionnalité pour compléter l’expérience.

Nous lançâmes, donc, cette série d’entretiens.

A la fin de chacun des 5 premiers entretiens, nous prenions toujours un court moment avec le fondateur sur notre méthodologie, afin d’affiner le script. Et au terme de chaque jour de travail nous débriefions des apprentissages majeurs de la journée.

Au bout de 5 à 8 entretiens, nous avions déjà détecté le plus gros problème à corriger.

Au bout de 15, ce problème devenait aussi visible que le nez au milieu de la figure.

Au bout de 21 (oui nous en avions fait un p’tit dernier pour la route), nous mourrions d’envie de passer à l’étape d’après pour régler ce soucis qui avait concerné 100% des utilisateurs.

Débrief des interviews 🎵

« Je ne comprends pas trop comment ça fonctionne »

Voilà, le problème c’est que les utilisateurs, ne comprennent pas immédiatement le fonctionnement général de l’application. [je reprends la narration au présent, car c’est là où nous en sommes !] Les différents verbatims que nous avons récoltés à ce sujet ressemblent à cela :

  • [depuis la page d’accueil, dès les premières secondes] « Okaaay, donc là je vais lancer une recherche, mais je ne sais pas tellement ce que ça va donner. Mais c’est beau en tout cas ! »
  • [depuis la page d’un artiste, après quelques secondes de navigation] « Hmm… en fait je crois que j’ai pas trop compris le système de points. Je suis content de voir qu’il y a des contributeurs et que c’est participatif, mais je ne sais pas comment faire pour en gagner moi aussi« 
  • [depuis la page des récompenses, après plusieurs minutes de navigation] « Aaaaah d’accoord, donc en fait on créé des pages, on gagne des points et ensuite on peut échanger contre des récompenses… ça marche ! Mais c’est trop bien ! Pourquoi vous ne le dites pas dès le début ?« 

Ce sont des retours très encourageants : les utilisateurs ont systématiquement approuvé le concept et donné des feedbacks spontanés très riches sur ce qu’ils trouvaient de positif dans l’application. Il faut simplement mieux mettre en avant notre parcours clé. Il faut plonger l’utilisateur dans ce parcours clé dès sa première visite, dès ses premières secondes de navigation, pour lui faire comprendre et apprécier tout le projet de Festview en quelques manipulations.

Je décrirai précisément le projet de feature que nous avons élaboré grâce à ces retours dans les épisodes à venir. Pour la fin de cet opus je préfère terminer de détailler la fin de la méthode de récolte de feedbacks, très importante pour la suite.

En tout et pour tout, nous avons récolté 104 insights différents avec ces tests. Par 104 insights, il faut comprendre 104 remarques différentes qui ont été formulées sur Festview. Par exemple, les 21 utilisateurs qui ont chacun dit « le positionnement de l’app n’est pas clair pour moi dès la page d’accueil« , cela ne compte que pour 1 insight.

Naturellement, un insight qui a été formulé par plusieurs utilisateurs paraît plus important qu’un insight qui n’a été mentionné qu’une seule fois.

Grille de User Impact Score 🎵

Mais, pour être précis dans la priorisation, chaque insight doit être pondéré avec un User Impact Score. C’est-à-dire qu’il faut prendre en compte l’importance qu’avait cet insight pour l’utilisateur qui l’a formulé. C’est nécessaire car si tous les insights avaient le même poids, on pourrait se retrouver avec des incohérences comme :

"Je trouve que la couleur du bouton ne va pas avec le reste" x 21 utilisateurs

↓ a le même poids que ↑

"Je ne peux pas créer la page d'un artiste car le formulaire ne fonctionne pas" x 21 utilisateurs

Dans la majorité des cas on souhaitera donner un poids total plus important au second insight de cet exemple. L’utilisateur semble bloqué dans sa tâche, contrairement au premier insight, où il semblerait qu’il ne s’agisse que d’une préoccupation esthétique.

Pour arranger cela, on va donc multiplier par un score allant de 0 à 3 à chaque fois que l’utilisateur mentionne un bug ou une suggestion d’amélioration, en suivant cette grille :

En appliquant cette grille à l’exemple utilisé plus haut on obtiendrait :

« Je trouve que la couleur du bouton ne va pas avec le reste » x 21 x 1 = 21 points

« Je ne peux pas créer la page d’un artiste car le formulaire ne fonctionne pas » x 21 x 3 = 63 points

Il faut noter que tous les utilisateurs n’ont pas le même regard sur les problèmes qu’ils rencontrent. Certains ont par exemple des exigences particulières en termes de graphisme, tandis que d’autres seront davantage dérangés par la syntaxe ou la fluidité de la navigation. Il faudra donc être capable de mettre en avant ces différences de perception, comme dans ce nouvel exemple :

« Les artistes ne sont pas triés par ordre alphabétique » a été dit par 14 utilisateurs au total.

Pour 7 d’entre eux, c’était très dérangeant. Pour 5 autres, ce n’était pas grave. Enfin, les 2 derniers ont dit que ça pourrait aller, si seulement le tri était aléatoire, sans être surs que ça leur plairait.

Poids total de l’insight = (7 x 2) + (5 x 1) + (2 x 0) = 19 points

Pour Festview, nous avons terminé la phase de tri et de priorisation avec un user impact score total de 620 points pour l’ensemble des 104 insights relevés.

Tri des insights 🎵

Dans ProductBoard, nous avons ensuite rattaché chaque insight à des projets d’amélioration. Ces projets sont découpés :

  • …en plusieurs composants (components)
  • …chaque composants est découpé en fonctionnalités (features)
  • …chaque fonctionnalité est éventuellement découpée en sous-fonctionnalités (sub-features)

🟥Components : 11

🔴Features : 104

🔺Sub-features : 17

Les components sont triés de bas en haut selon leur poids. C’est grâce à ce tri qu’on identifie les projets de développement prioritaires.

Mais quels sont ceux qu’il faut commencer à traiter dès aujourd’hui ? Quels sont ceux qu’il faut remettre à plus tard ?

Combien peut-on en traiter simultanément ? Et pendant combien de temps ?

Prendre trop d’items simultanément pourrait s’avérer contre-productif. Et en prendre trop peu pourrait retarder la date de la prochaine version du produit…

Conclusion 🎵

Vision ⇒ Positionnement ⇒ Idées de parcours ⇒ Fonctionnalité minimale ⇒ Développement ⇒ Test ⇒ Pondération et tri des insights ⇒ [priorisation]

Ce parcours de product management quasi-scolaire établit un cadre méthodologique très saint : grâce à cette méthode, on acquiert chaque jour un peu plus de certitudes quant aux choix que nous faisons et pourquoi nous les faisons.
Dans cette partie, nous avons vu comment faire ressortir des idées d’améliorations sur le parcours clé, suite aux retours des utilisateurs. 104 idées qu’il est impossible de traiter simultanément. Il faut donc désormais prioriser.

Dans le prochain opus, 3 – Définir les objectifs stratégiques de Festview, nous verrons comment ce travail de recherche auprès des utilisateurs vient s’articuler avec l’avancement général du projet. Comment l’équipe définit un panels d’objectifs à atteindre pour un temps donné. Car c’est grâce à ces objectifs qu’elle pourra sélectionner un nombre correct d’items à délivrer pour l’application de manière à servir ces objectifs. On passera notamment par la méthodes des Objectives and Key Results – OKRs.

Outils

Comme promis, voici les outils utilisés dans cette phase de définition du parcours clé puis de recherche utilisateur.

  • Atelier story mapping : Miro, permet manipuler des post-its, à distance s’il le faut. L’outil dispose même de templates pour ce type de travaux
  • Booking des interviews : Calendly, permet de déclarer des disponibilités dans un calendrier et de laisser son contact choisir le moment qu’il préfère parmi ces disponibilités
  • Trame des interviews : Google Form. On ne le présente plus. Pour les questionnaires quanti directement remplis par les utilisateurs, j’aurais utilisé Typeform, bien plus classe
  • Visio avec les ßeta-testeurs : Whereby. Ancien appear.in, on ne peut pas faire plus simple : le site génère un lien, les utilisateurs qui cliquent sur ce lien se retrouvent dans la même pièce virtuelle en visio.
  • Report, pondération et tri des insights : ProductBoard. C’est le seul outil de gestion des insights que j’ai utilisé jusqu’à présent, je le trouve parfait. A l’exception de son prix, trop élevé après l’expiration du mois gratuit : 49$/mois/user. Sachant que pour un projet qui démarre, on n’utilise pas ce genre d’outils tous les jours après une phase de recherche. On peut cependant utiliser la version gratuite et exporter tout le contenu avant qu’elle n’expire.

Spotify for pets ?!

Yep, you read it well hahaha! Spotify for pets. Two days ago, music streaming leader Spotify launched an automated playlist maker designed to please your beloved animal. They describe it as « An algorithmically created playlist based on your listening habits and your pet’s attributes, so it’s music you both can enjoy.« 
I’ve tested the setup, it’s quite nice!
What about the overall experience? Would you qualify it as a buzz feature?

For now, this funny tool is made for iguanas, dogs, birds, cats and hamsters. Once you’ve selected what type of room-mate you live with, you’re to answer 3 questions around their personality. So that, we suppose, the issued playlist fits even better with the atmosphere of your living room. The same attributes are used for all animals: energy, shyness and curiosity.

In the end, you get a 30-songs playlist that mixes your preferences and discoveries (50% of the artists/musics that were playlisted were already in my library).
But does it really adapt the playlist to the pet’s attributes?

Does it work?

Yes it does!

With the same CAT, I’ve tested 2 different sets of answers. The first one, qualified as APATHETIC, SHY and RELAX. The second one, qualified as CURIOUS, ENERGETIC, FRIENDLY. Here’s the comparison below.

Indeed, I would definitely agree that artists and musics in the two playlists have opposed style and atmosphere.
The first one stars tracks that I would qualify as relaxing, dreamy, blurry or chilly. Musically speaking, they can be defined with : low tempo (70 to 100BPM), poetic, dramatic, sad or elegiac musical phrases (e.g. Am-F-C-G)…
– But the second one is radically different and offers songs I would qualify as exciting, boosting, happy or groovy. Musically speaking, they can be defined with : high tempo (100 to 140BPM), smily, happy or positive chords sequence (e.g. C-E-G)…
These are the data Spotify relies on in order to qualify their music genre and suggest to the right playlist. We’ll try to dig that matter in an other article.

Note that there’s a smart function in the algorithm that prevents it from suggesting twice the same content with the same data set. I’ve tested it 2 times with the same cat and the same attributes: the playlists had the same style, but not the same artists and musics.

Why this feature?

Spotify has proven many times being king in terms of retention. The paid service churn rate barely crosses the 20%. And 250 millions active users are at stake all around the world. As we can guess, this buzz feature is not supported by scientific background or any promise that your pets will feel better with the playlist. This new service aims at activating pet owners of all Spotify subscribers and even pet owners that haven’t tried the streaming service yet.
Plus, could you imagine how easy it is for any digital marketer to identify pet owners on social networks. Engagement around pet content is very easy to spot. As a product manager, I probably would think this feature as a conversion or retention lever that would go along a digital acquisition campaign targeting pet owners.

Let me know what you think 😉 Cheers !
PS : here are the two playlists: Energetic, Friendly, Curious / Relax, Shy, Apathic

Pro goals for 2020 🎯

Happy New Yeaaaaar !! May 2020 bring you the best both in your personal and professional life 🥰

Here are few lines to have a look back on 2019 and talk about 2020. In mid-2019 I had my first official steps as a freelancer. Here are some major achievements that I’d like to sum up, as source of learnings and self motivation for next year!

Blogging’s so cooool

I launched this personal blog with a specific focus on music and tech related topics. 8 articles have driven 500 unique visitors. I had an amazing help from Product fellows and reader friends to improve both writing style, sourcing and argumentation, many thanks to you guys 🙏🏻.

As a short feedback, I confirmed this year that blogging is an amazing source of professional and personal new inputs. An amazing source of learning. It helps structuring your thoughts and your professional positioning. Even if readers don’t show: go for it, write and post. You’ll see what it brings later.

🎯I will write at least 2 times more article next year. In the same editorial style, I will try to dig deep into product, go-to-market and design issues related to music apps, softwares and devices.

I will also increase the pace of small publications to give you relevant and up-to-date piece of information around music-tech news. I will try to cover both major industry actors (Spotify, Soundcloud, Deezer, Pandora, Resident Advisor,…) and minor emergent products or trends (Soundtrap, Kraze, podcasts, Beat2beat,…).

Product Management

I got involved in the development process of 4 digital products:

  • Wookyhome.com: find amazing places to work for your team and enhance well being at work.
  • Fest-view.com: contribute to the greatest music encyclopedia and get rewarded with music products.
  • Kraze.fr: find the electronic music party that suits you the best in your city.
  • Appaloosa.io: deploy your mobile apps onto your employees device in a BYOD strategy.

My help on these projects had different angles. For some, the priority is to shape the vision and the business model. For some, the priority is to think a smart and cost-efficient go-to-market. For others, what I helped enhance the user understanding, the experience or the interface. And finally, some teams had me working on the agile delivery process itself.

🎯I hope that all these products will reach their business and experience targets in 2020.
My goal is to continue working on 3 to 5 running projects, focusing on industries that I love.

Design

I designed 3 logos for 3 beautiful environmental projects: Speaker4Earth, Coach4Earth, Business4Earth. They allow professionals to plant trees each time the work for their client. Collaboration was perfect, and projects are now starting really great.

🎯Plan for 2020 is to work a little more on design missions.
I would love to help other starting or growing business in the music or environment industry to communicate with efficient and modern materials.

Remote working

Finally, 2019 has been, for me, the discovery of the remote working world. I’ve spent 5 months working remotely from Germany 🇩🇪 (Hamburg 🥰), 1 month in France 🇫🇷 (Paris and Biarritz), a few weeks in Brazil 🇧🇷(Rio, Sao Paulo and Ilha Grande).

🎯I did it both as freelancer and fulltime team member. In 2020 I have 2 goals regarding this matter:

  • Spread the word and the best practices/tips around me. If you or your team want a feedback around that, knock here 😉
  • Continue working from the places that I love around the world 🌎

Kraze 💃🏻🕺🏽 The app you’ll need next Friday night

Do you remember last Friday night? 

Straight out of your working or studying day, you rushed into a bar for a couple of drinks with your friends. Happy as you were, you’ve warmed up together until you reached that key moment: « Guys! What do we do next?! ».

Kraz app help partygoers finding the right music event.

And this question is not so easy to answer… Today, if you want to find the perfect party for your crew, you have to patchwork some « I’ve-heard-abouts », with some Resident Advisor and Facebook events. If you care about the music that you want, then you must also pop over Soundcloud or Spotify to have a preview of the artists and DJsets. Eventually, you would also combine it with a public transport app to estimate and compare travel times… It’s a multi-interfaces, +15-minutes, hard-focus mission.

OR, to wrap it up the easy way, you can try Kraze.

Kraze app logo

Kraze helps partygoers finding the right music events around them.

This 2 years-old Paris-based startup has built a smart crawling algorithm that finds music events published on Facebook, and gathers them on a map. Just pick a date, and the app will show you what’s happening that day. As they’ve just released their V2, I thought it was a good moment to give it a thoughtful look. I’ve spent some time with one of its founders and CEO, Bastien Champ, to understand both the global mission that they want to achieve and the next features they think of.

☝🏻Before we get into the app, I wanted to share this vision that Bastien told me, and that made me veeeery curious:

« We want to bring cultural events into the in-real-time era, so that everyone can have fun when and where they want. »

A quick tour of the app

Kraze has a quite simple approach of its user experience.

Here’s the main flow that you can see above :

  1. After signing in, you’re asked about your favorite music genres. Just pick a few of them and allow app’s access to geolocation…
  2. … and then the city is yours 😉 The map points out music events happening in Paris today. You can swipe dates if you want.
  3. Each type of event (before, festival, after party…) has its color. Click on a dot to get more details about the event.
  4. A description of the event, a link to online ticket purchase and even a smart « How to get there » plugin help you making the right decision. Process it by saying that you will « Attend » the event.

☝🏻What did you want to bring with this second version Bastien ?

« We wanted to simplify the experience of browsing events. We have 200 of them every week ; this version should help users find their event more quickly.« 

1 app, 2 usecases

Remember our Friday night emergency « What do we do next, tonight? ». Well, this is the first use case that Kraze addresses. And there’s a second one that pops out very quickly: « What can we plan for next weekend? ». Indeed, some of Kraze users might prefer thinking about their weekend in advance. They like it slow and mindful. They like grooming event pages for hours to be sure that they’ll chose the best party of the weekend.
In both cases, still, the goal is to make a decision. The user goes from a list of possible options, evaluates them, compares them and eventually commit to one of them.

So let’s try to analyze the decision process of the two usecases.

Kraze app : 2 usecases

Here’s a suggestion of decision process breakdowns. It goes in 3 steps for the first usecase, and in 5 steps for second usecase.

What do we do next, tonight?

#1 – Timing and geolocation constraints

The number of event options is not so high in the beginning, because we’re only interested into coming-soon events, in a certain location. And since timing and geolocation constraints have strong power in this usecase, what you need more than ever is a clear and minimalist interface that points out easily the few events that might interest you. Meaning reachable events, starting soon and in a genre that you love.
You don’t want to be disturbed by any events happening next week, tomorrow, or in the opposite suburb of where you are…

🤔What does Kraze bring at this step?

Kraze app: map overview
  • ❤️A one-of-a-kind event map, bang at the opening of the app. The color code is helpful and fun to use ;
  • ✅If you prefer the list view, they have it as well ;
  • ✅A relevant and comprehensive suggestion of events when you’re into EDM. I’ve compared Kraze to Resident Advisor or other local event apps on the same day: Kraze has proven being more comprehensive several times.
  • ✅A focused minimalist UI.
  • ❓Could be even more focused. The date picker might be optional here. The time to get there could be pre-loaded and displayed. And the map actually displays all music genres ; even though I’ve given some favorite genres in the on-boarding phase of the app.

#2 – Filter, compare and choose

To reduce again the number of possible choices, users should have easy access to relevant filters.
Then, good synthesized information about the events will help them finding the one that they love. Short pre-listening widgets of DJ sets could be very useful for example.
Final choice could eventually be triggered by some live information about the venue, or the artists. « DJ Blabla has just began her set, join us!« 

Kraze app: 1st usecase, second phase : filter compare and chose

🤔What does Kraze bring at this step?

Kraze app: how to get there integrations
  • ❤️For each event, a one-click « How-to-get-there » Google Directions API integration (see screenshot on the left). Pretty smart and useful in this usecase.
  • ✅The list of genres that user can enjoy is relevant. They are actually EDM sub-genres that, I think, have the good granularity.
  • ✅Kraze also reaches the industry standard of showing other participants on the same event. We’ll come back on the social potential of this feature. Little tip: the size of the dots on the map vary according to the number of attendees. Cool right?
  • ❓Filtering was not so clear to me. I was confused in the identification of what was a filter, what was my personal preferences and what results they would affect.
  • ❓For now, live information has been set aside.
  • ❓What’s missing could be the synthesized music related content. DJ sets previews, links to Spotify or Soundcloud pages… It could really help partygoers taking a decision.

#3 – Commitment

Now that the event is chosen, the user is only a few steps away from getting there once and for all. What could you use now? First, you might need to purchase a ticket, if possible. This should be very easy to find. Then, maybe share to friends that might join you later. And eventually, a geolocated direction would be, again, very helpful.

🤔What does Kraze bring at this step?

  • ✅It is convenient to have a stable permanent button to online ticket purchase. Instead of having to dig a hidden link in the description.
  • Sharing features are also within easy reach.
  • ❓For now, online purchase goes through an external link. Some competitors host the payment directly in-app. Even though it could be considered as a non-prior evolution, let’s just say that it would simplify the final experience. Letting the user on a consistent and reassuring flow.
Kraze app helps you if you wonder what party you could go to tonight.

What can we plan for next weekend?

This use case is more complex. When they plan for next weekend, the users have time. So they can put more effort into their research. They can leave it and come back to it later. They can make a shortlist and groom their saved events day after day, as an entertainment.

Kraze app: sceond usecase : what can we plan for next weekend.

#1 – Narrow down possible options

Initial start is quite wide, since planification could cover several days of events. When you live in a big city, a summer Saturday can offer up to 30 options, with open-airs, before parties, warehouses and classic nightclub DJ sets. In this context, smart recommendations according to user’s profile could be great to reduce the initial scope. It implies a fine understanding of what the user likes or dislikes. The data work behind that can prove quite complex, though essential, if the app is to memorize users’ activity and preferences.

🤔What does Kraze bring at this step?

  • ❤️The « map + list + datepicker » combo works really well here. The user gets a relevant overview really easily.
  • ✅Search feature works well, both on events’ title and venues name.
  • ❓As in the first usecase, it’s hard to understand what impact has the favorite « genres » and the party types filters onto the map or list.

#2 – Shortlist my favorite events

Here’s a very specific part of this extended decision process: the user has time to save their favorite picks into a shortlist.
First, in its basic sense, this shortlist just pulls together the events you would like to attend in one single view.
We’ll come back on another opportunities around this shortlist just after.

🤔What does Kraze bring at this step?

  • ✅A fine overview of the events that you « Attend ».
Kraze app: shortlist of events

#3 – Explore similar events

And here comes one other very solid step to be addressed within the decision process: the suggestions based on what you’ve already shortlisted.
For Netflix generation, exploring « similar content » has become a basic gesture.
« If you like this event, you might like that one ». This way, the user widens their opportunities once again. But only with events that should be more relevant.
In this context, each event has a set of attributes: the music genre, the venue, the time frame, the artists or even labels, etc. And what is key here, is to chose wisely the attributes that the algorithm takes into account to make good suggestions.

Kraze app: second usecase, third phase : explore similar events.

🤔What does Kraze bring at this step?

For now, the recommendation part is not implemented in Kraze. It requires a solid and rich data set. And getting this data set ready requires a high pace of content management and tagging. Pretty hard mission for a 2 years old startup, all on its own. So it is highly possible that they work on industrialization of this process without having to do it manually on their own…

#4 – Help me making the best decision

My favorite part!
The users planning something for the weekend come back on this decision process several times in the week. It’s not a one shot.
So this is when our « Shortlist » feature comes back again with an another purpose. Today, « saving » something or saying you’re « interested » (poke Facebook) is like subscribing to an activity feed. So actually, what the user half-thoughtfully expects by doing this, is to be further updated or reminded by their shortlist of events.

Facebook interested into an event

But when and for what should the user be updated?
Here’s a suggestion of classification:

  • Very important information: the venue changes ; an artist cannot perform ; a new artist is booked ; early bird ticket sale is about to end = push notification + email.
  • Other substantial information: a friend attends the event ; hosts post some pictures of the place ; artists post something about their live set = in-app notification + activity feed.

Time is on user’s side to refine their shortlist mindfully. They can dig into music related content, compare DJs, read some comments about the venue or even ask for details to organisation, etc. Thanks to the information they are fed with, they reduce the shortlist little by little.

🤔What does Kraze bring at this step?

Kraze has not yet worked on personalized activity feed.

#5 – Final commitment

To take the final decision, users need to share some links with their friends. They also need to purchase tickets online and to put the event in their agenda. They do any extra gestures that could help them being on top on D-day.

🤔What does Kraze bring at this step?

  • ❤️An exclusive feature can seduce users here: Kraze offers some little contests in partnership with some hosts. Directly in-app, the users can participate and win tickets for free. This is a good engagement feature for someone planning their weekend.
  • ✅Kraze offer a « share event » integration. It triggers native OS sharing options (messenger, whatsapp, email…).
  • ✅You can add the event to your calendar.
  • ✅Link to online ticket purchase is, again, within easy reach.
  • ✅How to get there integration is also again very useful to prepare your journey.

For what usecase is Kraze tailored?

I find the second one very interesting, but also rather difficult to address. It requires a very fine and mature understanding of what tools or information users need, and at what moment. Thus, it also means a higher number a features. And more complex ones.
Irrigating an activity feed with artist’s or venue’s content is a challenge!

But on the other hand, is it worth for Kraze to settle only on the first usecase? Emergency is also interesting to address and requires expertise. But how can you monetize or build a strong community with an app that people use only 5 minutes per week? Hard mission, I think…

☝🏻Bastien, to understand Kraze’s positioning, could you tell a bit about your community? Who they are, and how you’d like it to grow?

« Today we have around 5000 monthly active users (out of 20k). They come from the inner city and are mostly between 18-23. It’s not so much, but it’s only in Paris and we think that we are still in an experimental phase. We are on our way to deliver THE solution that will revamp and boost access to parties.
In the future, we’d be happy to see more occasional partygoers on our app, older ones, who’d use Kraze bang at the moment they need it
« 

A place to seize on the nightlife app market?

Without giving too much details, I’ve compared Kraze to some of their competitors to maybe guess what place could it seize on this nightlife app market:

Selected competitors for competition analysis : Kraze, Resident Advisor, Facebook, ClubGo, Xceed, OhSchonHell, ParisBouge, Fever
  • Resident Advisor: famous historical electronic music content company, that delivers an app focused on underground music events.
  • Facebook: the social network has improved its event suggestions during the past years. How does it rank today?
  • ClubGo: Indian app, positioned on the same night industry.
  • Xceed: also good competitor from Barcelona.
  • OhSchonHell: one of the German competitors that expanded from the web, to the app world.
  • ParisBouge & Fever: I found interesting to go through those general event recommandation apps in Paris. They have the same native audience, but they’re not positioned on the same exclusive music industry. What’s interesting here is the user experience.
Competition mapping on the the usecases.

From where I stand, and when analyzing both usecases for each app, it seems that Kraze holds a great place already, amongst competition.

Traditional and historical app, Resident Advisor, hasn’t evolved so much these last year and suffers a lack of modern and important features. Geolocation, for example, should be a start on every nightlife app today. Still, they have amazing content and good event coverage that help people get good details about events if they need to.

Facebook has a strong set of features. Especially, they know how to entertain a user that has declared being interested by a couple of events.
But the whole thing lacks focus. You cannot feel any community approach where electronic music gathers everyone.

My surprise was Xceed. They have a large set of features. And have started implementing data-related features, in order to personalize the experience according to preferred artists or venues. Their go-to-market strategy is not so clear, but on their website, you can spot this pro interface that seem already quite advanced. So we can guess that hosts already contribute a lot to delivering detailed content to the community. And this way enrich the available data set.

Xceed backoffice for hosts

☝🏻What’s your biggest business challenge today Bastien?

« We must seduce event hosts on our plateform. They’re key to future monetization and community expansion. If this works in Paris, we’d love to expand to other european electronic music capitals : London, Berlin, Amsterdam/Rotterdam, Barcelona…« 

Kraze back office for event hosts

On the users’ side, extending the set of features to get a strong positioning on one of these usecase could be great.
And for the business part, it seems obvious that the extension of the community and product has to go through a professional plateform that delivers good value to event hosts.

These are big and exciting challenges, we can wish them a lot of success and show them great support for the coming versions! 🤙🏼

Bastien et Thung, cofounders of Kraze


What do you think? Rate this article from ❶ to ❺ 😉
(polls)

🎧 Spotify Stations: on top of hands-off music listening

Spotify released its Stations app not so long ago, in ßeta in the US. Here’s a complete review of this very interesting project, before it comes down on our devices!

Full app review of Spotify Stations

Spotify released its Stations app not so long ago, in ßeta in the US. Here’s a complete review of this very interesting project, before it comes down on our devices!

Let’s say it once before examining the whole thing: Stations is a smart, modern and audacious project. Here are the main take aways

  • #TheApp : « Stations » is an independent Spotify app. It focuses on a feature that already exists on Spotify: the stations. When you play a station, the music basically never stops, because a track starts just when the other stops.
    You can have several different stations, for each genre or artist that you like. And each station’s queue of tracks is composed by the app, following your preferences.
  • #UX – Spotify Stations only contains… stations. That’s all. No podcasts, no playlists, no events, no artists or song search. Only stations. As such, the user experience is incredibly powerful while being very simple. In terms of features, the app « only » replaces a classic set of FM stations, but here, these Stations are powered by Spotify’s famous recommendation algorithm. Very audacious bet and proof of confidence of the company in its number one appreciated asset.  
  • #UI – The interface smartly delivers this experience. Turning on the app for the first time is enough to access a station without a single click. Then, the user has barely 4 or 5 moves to make to enrich their stations. No more. And native interface components are modern and very intuitive.
  • #Algorithm – The algorithm training of a station relies on adding artists into it and on liking tracks while they’re played. Sure, it’s satisfying. But in the use it is probably not pushed at its maximum since the app doesn’t cover on its front the complete set of links between artists/tracks/albums/events/playlists/etc. that the original Spotify app contains. Especially, the playlist concept has been completely wiped.
  • #Positioning – There was a place to take on the easy music-listening. So with this app, all of us who need – always or sometimes – to just let it play while we do something else, we will be satisfied. More precisely, « who exactly does Stations target? » I don’t know, but I would be so curious to see the analytics few weeks after the public release and try to draw core personae out of it…
  • #Deepness – Details around a track have been blown away too. Click on a track that’s playing: nothing will happen. You will access no related album, nor artist page or related stations. Experience is really about a « hands-off » mode. Launch the app, it will play the music that you like, that’s all you need to know. Some of us, passionated listeners that are interested into the content that surrounds a track, might be a bit frustrated.

A personal opinion? I spent a few hours letting it play and enriching stations… It’s incredibly powerful and accurate – as expected. And content around stations, if some is meant to be produced/curated/featured, could be key to further success. Let’s dig into it and try to understand Spotify’s project in a wider stand. 

Are you a Spotify user?
(polls)

On-boarding

Next screenshots will be in english 😉 I couldn't retrieve these screens in english...

After first login – with classic credentials or Facebook connect-, you get to this nice three-screens slider. Short and quite usual.
Maybe you recognize the Circular font (Arial family) of Spotify?
There’s some new stuff: Spotify used to on-board their users with another atmosphere, especially with lifestyle pictures in background (see here). This new iconography and flat unified colors in the background might be a first shot for the ßeta version, but I like it this way. It’s rather punchy.
And it actually paves the way for the awe-inspiring simple UI of the rest of the app.

Wooops what happened to second slider's subtitle? Looks like they forgot to center it...

To finish the on-boarding, a little guidance is available in-app, if needed. And that’s it. You’re all set.

I find a bit old-school the « Back » and « Next » text links in bottom corners. Arrows would probably have been enough (see Instagram stories on-boarding).

Single-screen app

« So let’s start with the menu. Wait… Where is it? Where’s the menu? »
No menu my friend. Stations app only has one screen for the main layer. From here, you can only get down in the app, nothing to see on the sides! It’s already a huge gap of simplification compared to the classic Spotify app, that displays a four-pages bottom menu.

The main body of this .gif shows the top-down slide that switches playlists. A very intuitive choice of display and touch for this action. Hence, stations switching is supposed to be the major daily move made by users. Might be, yes. We’ll come back to this question later on.

When the user glances at the top bar, looking for the menu, they integrate that only one action will actually ever be available.

The top bar stages only one main action: « + Add station« . Very interesting choice. In fact, we better know top bars dedicated to title, search and even more frequently, to menu access. So when the user glances at it, they mentally replace the « Looking for other options » expectation with the « There’s only one option » reality.

Still on top bar, we find an easy-to-reach device controller, plus the settings. Again, I think it’s a lucid choice not to have merged them. My guess is that over the last years, smart home devices (Google Home, Alexa,…), connected speakers and wireless headphones/ear pods have proven their daily weight into music listening. Analytic review of this already-existing feature on classic Spotify must have guided this choice.

It’s already playing

Open Stations, it starts playing. Even on first on-boarding.
« But you had no pre-configured stations, had you ? ». In fact, Spotify groomed my favorite artists and my top songs and built a first batch of 4 Stations out of it. Remember 2018 wrapped content? This superbe retention campaign was ment to higher purpose than just engage their users in few clicks on January 2019…
In other words, if you already have a Spotify account with a bit of listening with it, you CAN NOT DISLIKE the first track that Stations plays. You’ve heard it, liked it, shared it, playlisted it so many times that it has to be a match with your expectations, had you any.

I must admit that I found it disappointing not to find any stations based on my dear Spotify Playlists. I mean, they’re juste like pre-built stations, aren’t they? « Maybe I’ll get them later… » I thought.

Enriching a station

Let’s come back on our UX journey.

Since we already have a few stations, it might be interesting to work on one of them.
You can enrich a station with two actions that will train the recommandation algorithm:

  1. Like or dislike the tracks that play
  2. Add artists to the « root » of the station (see .gif)

Artist search is pretty simple, starts with only 1 character, and artists pictures are helpful.

What’s interesting comes next: once you’ve added an artist, other ones related pop immediatelysmartly inserted in-between the one you’ve just added and the last artist you had added before.
Notice that within this discovery phase, it’s the first time that the user sees the recommandation algorithm in action. Everything before was just blind confidence into pre-built stations.
It’s like taking a breath with a behavior that we finally know, after all this newness.

A little down, to me, is this double « save » click that’s needed to complete the update of the station. I missed it once or twice.

Unless you try the app, you will have to trust me on this one: I think that 1 extra gesture is missing.
When I am listening to a station, it can happen that a song that I like pops up. But it actually would fit ANOTHER station – or several others. So I don’t wan’t to dislike the music, because I like it. But I don’t want to like either, because I don’t want the algorithm to integrate that it was a good choice to put this song into this specific station.
In that case, it might be useful to drag and drop the music played into the playlists that I aim.

Creating a station

Back to our « + Add station » button.
A click will take you to a two-tabs screen that will let you start your station with either « Artists », either « Genres & Moods » (see screenshots below).
For each genre/mood, you get sub-genres stations.

First thing about this artist recommendation: I really don’t like Damso‘s music. Nor Maitre Gims‘. I mean, that’s even possible that I disliked them in classic Spotify already.
So let’s cut this review right away, this app totally suc…
No wait. It’s actually interesting. Remember the point about playlist-based stations? Well they’re still missing here.
Why can’t I just turn on a station that’s based on a playlist that I’ve put so much effort into? If I could do so, I might not have to deal with artists that I don’t like.
It’s a real down to me here since I’ve spent literally hours on building them. Naming them. Picturing them. Spotify even had me creating folders, and subfolders of playlists…
Playlists are like babies to Spotify users. And now Stations wants us to forget about them? Tough.

Work, commute, drive = Stations

With an objective Product Management stand, I must say that this app is a remarkable shot towards all people used to listen to music when they work, commute, drive, ride, read, play,… it’s obvious that a big market hole was waiting for this to come. Well, actually Pandora was first on this field but I’m not sure they will last long against Stations. UX is still too complex, UI not as qualitative.

Soundcloud, Deezer, Apple Music… same thing. For sure, they all are great products. But not meant to deal a single feature.

  • Soundcloud doesn’t even suggest a default station that you would be keen to listen. And when you launch one, they prove being quite light on the recommendation part, since they invite you to modify the queue. So it becomes a self-made temporary playlist, not a station.
  • Deezer Flow and Apple Music have way more complex UIs than Stations with a lot of deepness and relations between objects. It’s great but it looses someone that just wants to launch a station right now, and then forget about the app while music continues.

Odds are great for previous generations too. We know that our parents were used to their set of 6 FM stations, but over the last decade, they might have neglected it – except in the car, for sure – because of other more powerful medias. If firsts releases’ data prove it, marketing & communication could hard-push on the 40-50 years old. They don’t have any relevant substitues to their FM stations yet.

Music is dead, long lives Music?

Now, in a more personal stand, I think I am too interested into the music that’s playing to be a core Stations’ user.

On Stations, click on an artist? Nothing happens. Click on a track? Still nothing. Looking for a song? You won’t find any. Want to check if this band released something lately? Trust the algorithm, it may bring it to you one day – or not.
So first, about internal objects’ linking: I am very frustrated to know that I will never get deeper (or sides) than the station level.

Second point, and more important to me, is about all the content that makes the music a living art.

Let’s talk a bit about what could have been added to Stations, so that profound music lovers become core users.

1. Easy integrations 📲

  • Spotify partnered with Genius and their lyrics database: Stations and Genius would be plugged. And stations would show the lyrics to those who want to sing or to learn what’s the music about.
  • Spotify launched Line-in in 2018, so that users can suggest edition on music metadata (release dates, genres,…): Stations and Line-in would be plugged too. And superpowers would be given to those who prove being educated/passionated/experienced users and contributors.

2. Share & collaborate 🤝

Again, easy stuff for Spotify, they already have every thing:

  • Sharing a playlist on classic Spotify is a very common action. Why not replicate it in Stations?
    Instead of URL link or emmebed, which exist today, sharing could be triggered from the app, and appear as push notif, or in-app notification on another device.
  • Exact same thing with multi-user collaboration on a playlist. Spotify already has it. On Stations, users would also collaborate with their friends and family on a common Station.

3. Beloved content

  • Spotify recently acquired Parcast and Gimlet Media, which are two huuuuge Podcast producers/broadcasters. They know how to deliver extraordinary content in a mic. What if they took the place of presenters? We would have so great stories about backstages, studios, events, talks about creativity or technique…
  • As « new generation FM presenters », these teams would also invite artists to perform live. For example, traditional Seattle KEXP 90.3 FM, and US NPR’s Tiny Desk Concerts made there reputation out of it: they know how to push artists beyond their limits and create very emotional atmospheres (here are two examples 1, 2). Stations’ audio content would be sure also available in video format. To be broadcasted on website, social networks, personal channels…
  • Red Bull Academy (RBMA) has done a tremendous work on stimulating music industry. Without a single line of code. They curate lectures of artists, boost partnerships between labels, and broadcast high quality content. And they’re not even native of this industry. In a wider and very exciting version of Stations project, it would be so great to have Spotify fostering artists and communities around music.

All right I think that’s enough reviewing for today. I hope you liked it:)
It was a pleasure to me anyway.
Stay tuned for more or write me at tristan.bochu.pro@gmail.com if you feel like talking about this one or anything else 😉
By the way, here’s the song I had at Stations Opening:

Souls of Mischief – 93 ‘Til Infinity
What do you think? Rate this article from ❶ to ❺ 😉
(polls)