<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Emmanuel Martin Chave on Medium]]></title>
        <description><![CDATA[Stories by Emmanuel Martin Chave on Medium]]></description>
        <link>https://medium.com/@manu.martin.chave?source=rss-6a00f89edec2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*b0r7eeZvgWflsQd4</url>
            <title>Stories by Emmanuel Martin Chave on Medium</title>
            <link>https://medium.com/@manu.martin.chave?source=rss-6a00f89edec2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 11:01:45 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@manu.martin.chave/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Gérer une équipe de Data : Trois leçons de BlaBlaCar]]></title>
            <link>https://medium.com/modern-data-network-mdn/g%C3%A9rer-une-%C3%A9quipe-de-data-trois-le%C3%A7ons-de-blablacar-509395ba1e70?source=rss-6a00f89edec2------2</link>
            <guid isPermaLink="false">https://medium.com/p/509395ba1e70</guid>
            <category><![CDATA[management-and-leadership]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Emmanuel Martin Chave]]></dc:creator>
            <pubDate>Sat, 04 May 2024 08:02:01 GMT</pubDate>
            <atom:updated>2024-05-04T08:02:01.972Z</atom:updated>
            <content:encoded><![CDATA[<h3>Trois conseils inattendus pour gérer son équipe Data.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ME5CMGHEZ_rR9tDzC0Qi4w.jpeg" /></figure><p><em>Basé sur </em><a href="https://podcasts.apple.com/fr/podcast/datagen/id1539383194?i=1000647920440"><em>mon podcast avec Robin Conquet sur DataGen</em></a><em>, je souhaitais partager trois conseils avec mes collègues Data Managers. Premièrement, les responsabilités entre les contributeurs.trices individuels.elles (CI) et les managers. Deuxièmement, la formation au management. Troisièmement, l’adoption d’une vision à long terme pour la construction d’une équipe. Cela vient de mon expérience à la direction des équipes Data de BlaBlaCar pendant 7 ans. Je vous laisse décider comment cela se généralise à d’autres contextes managériaux ! English version available </em><a href="https://medium.com/blablacar/unexpected-tips-for-data-managers-c44a71db6594"><em>here</em></a><em>.</em></p><h3>Jouer tel un orchestre</h3><blockquote>“Devenir manager n’est pas une promotion pour un CI. C’est un travail différent.”</blockquote><p>Les musiciens s’attendent-ils à être promus chefs d’orchestre ? La question n’a pas de sens : la promotion implique un classement. Dans ce cas, il n’y en a pas, les musiciens et les chefs d’orchestre ont des responsabilités différentes. L’harmonie naît lorsque ces éléments se combinent en un tout cohérent et agréable. Une équipe Data se comporte comme un orchestre. Sa performance vient des interactions entre le chef d’orchestre (manager) et les musiciens (CI). Ce sont des métiers différents, il n’y a donc pas de promotion entre eux.</p><h4>Définir les responsabilités distinctes des managers et des CI</h4><blockquote>“Une idée courante — mais fausse — est que les managers et les décideurs sont identiques. La réalité est plus riche. Les équipes prennent beaucoup de décisions. Certaines appartiennent au manager. Par exemple, décider d’une promotion ou d’une augmentation de salaire. La plupart des décisions sont prises par les CI : de la conception technique à la livraison d’un projet en passant par la gestion de projet.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*h2v3jyUSiQN37iJY" /><figcaption>La plupart des décisions au sein d&#39;une équipe sont des décisions des CI. Leurs décisions et les décisions du manager déclenchent les actions de l&#39;équipe. Ces actions conduisent à un besoin de prendre de nouvelles décisions, par une boucle de rétro-action.</figcaption></figure><p>Pour qu’une équipe Data fonctionne bien, il est crucial de reconnaître les responsabilités distinctes des managers et des CI. Cette division claire des responsabilités évite les chevauchements, qui génèrent de la confusion. Elle permet à chaque rôle de se concentrer sur ses fonctions principales, ce qui améliore la performance. Dans les équipes Data de BlaBlaCar, nous avons mis en œuvre les distinctions suivantes :</p><ul><li><strong>Les managers créent de la valeur en se concentrant sur les connexions</strong>. Au sein de l’équipe, il s’agit des connexions entre les personnes ; elles mettent les personnes en relation autour des projets ou des processus. Cela inclut les décisions relatives à l’allocation des membres de l’équipe aux projets, à la gestion de la performance, au développement de carrière et à la dynamique d’équipe. En dehors de l’équipe, les managers entretiennent des liens avec l’écosystème plus large de l’entreprise. Ils partagent de l’information, représentent leurs équipes ou gèrent les dépendances avec d’autres équipes.</li><li><strong>Les CI créent de la valeur en s’appuyant sur leur expertise pour mettre en œuvre des solutions de haute qualité</strong>. Cela nécessite de prendre des décisions concernant la conception et la mise en œuvre de solutions, la réalisation d’analyses et la clarification des besoins des parties prenantes. Les CI sont comme des musiciens : ils pratiquent un art techniquement exigeant. Ils décident comment jouer leur rôle, en collaboration avec le reste de l’orchestre.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*POfv2ym6c6TINoz4" /><figcaption>Les managers créent de la valeur en nourrissant les liens entre les équipes. Un contexte riche conduit à des équipes plus performantes. Elles ont plus d&#39;information, qui leur permet de prendre de meilleures décisions.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Gd_uRS6JWlcFsXXO" /><figcaption>Au sein de l&#39;équipe, les managers se concentrent sur la mise en relation des membres de l’équipe. Les managers créent également de nouvelles connexions - par le biais de processus par exemple - pour faire réussir les CI.</figcaption></figure><p>Ces distinctions ne sont pas exhaustives. Construisez celles qui sont pertinentes pour votre équipe et faites-les évoluer au fil du temps. Cette répartition des responsabilités permet aux managers de fournir un leadership et un soutien efficaces, tandis que les CI peuvent se concentrer sur leur expertise technique et stimuler l&#39;innovation. En définissant et en séparant clairement les responsabilités des managers et des CI, les organisations créent des équipes Data efficaces.</p><h4>Déplacer la prise de décision technique vers les CI</h4><p>Prendre de mauvaises décisions mène à l’échec ; alors, évitez-les ! Les CI possèdent généralement l’expertise technique la plus approfondie, ainsi qu’une expérience pratique. Cela les rend les mieux équipés pour décider des aspects techniques de leur travail. Par exemple, demandez aux <em>Data Scientists</em> et aux <em>Data Engineers</em> quelle architecture d’IA ils recommandent, plutôt que de l’acheter sur étagère, puis leur dire “vous n’avez plus qu’à l’utiliser”. Ne demandez pas aux managers de construire votre plateforme technologique. Demandez-leur plutôt de fixer des objectifs et contraintes que les solutions devront respecter. Ensuite, donnez aux CI l’autonomie nécessaire pour choisir les solutions en fonction de leur contexte, de leur expertise et de ces contraintes. Cela limite les mauvaises décisions et fait gagner du temps.</p><p>Dans les équipes Data de BlaBlaCar, nous avons mis en œuvre ce précepte par le biais de nos forums <em>Data Stack Governance </em>et <em>Chapters</em>. Nous avons copié les forums existants — qui fonctionnent bien — du département <em>Engineering</em> au sens large, dont font partie les équipes de données. Il y a par exemple le <em>BlaBlaCar Architecture Team</em> — nom de code <em>BAT</em> 🦇 — pour examiner les décisions d&#39;architecture, le <em>Backend Guild</em>, la future <em>Software Factory Guild</em>, ...</p><p>Le passage de la prise de décision aux CI leur donne du pouvoir. C&#39;est parce que dans une organisation, la prise de décision donne du pouvoir. L&#39;autonomisation favorise une culture d&#39;appropriation et de responsabilité. En donnant aux CI les moyens de prendre des décisions, les organisations réduisent les goulots d&#39;étranglement, accélèrent la livraison et améliorent la qualité de leurs produits data. Le déplacement des responsabilités en matière de prise de décision technique aux CI aide également à développer leurs compétences en matière de leadership et de résolution de problèmes. Cela les prépare à des postes avec plus grandes responsabilités. En faisant confiance aux CI pour prendre ces décisions, les organisations favorisent une culture d&#39;innovation et d&#39;amélioration continue.</p><h4>Créer des opportunités de croissance pour les CI</h4><p>Pour fidéliser et responsabiliser les CI talentueux, les organisations doivent offrir des opportunités claires et exploitables. Il existe trois leviers puissants pour créer des opportunités pour les CI :</p><ul><li><strong>La mobilité interne permet aux CI d’expérimenter différents rôles et projets.</strong> Ils capitalisent sur leur expérience, tout en découvrant de nouveaux sujets. Rejoindre une nouvelle équipe permet le partager des connaissances ; cela donne aux managers plus de flexibilité en matière d’allocation des ressources, mais aussi pour la gestion de carrière, comme pour les promotions. En offrant ces opportunités, les organisations déclenchent un cercle vertueux : la réduction du nombre de départ signifie un effectif complet, qui est plus percutant, créant de nouvelles opportunités.</li><li><strong>Les projets de <em>Chapter</em> permettent aux CI d&#39;acquérir des responsabilités dans leur domaine de compétences. </strong>Un <em>Chapter</em> est un forum pour les personnes qui travaillent avec la même technologie. Dans les équipes Data de BlaBlaCar, nous avons quatre <em>Chapters</em> : <em>Data Engineering</em>, <em>Data Analytics</em>, <em>Data Science </em>et <em>Data Back-End</em>. C&#39;est un excellent moyen de s&#39;améliorer techniquement et d&#39;avoir un impact transversal. Cela permet aux organisations de créer de nouvelles opportunités pour les CI, comme le rôle de responsable de <em>Chapter</em>. Les responsables de <em>Chapter</em> guident et encadrent d&#39;autres CI dans leur domaine d&#39;expertise, définissent des normes, identifient les compétences et représentent les intérêts du <em>Chapter</em>. En établissant ces rôles, les organisations favorisent un sentiment de communauté, fournissent un mentorat précieux et augmentent la satisfaction au travail et la performance de l&#39;équipe.</li><li><strong>L’échelle de carrière donne aux CI une perspective sur leurs responsabilités futures</strong>. Elle établit les attentes sur la façon de progresser. Dans les équipes data, nous avons constaté un changement de comportement significatif lorsque nous avons commencé à promouvoir les CI à des postes de senior. Cela a créé des modèles concrets à suivre. Un avantage supplémentaire de la construction d’une échelle de carrière : elle rend les évaluations de performance des managers plus objectives.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*oF6Z-ycyog2qhjYJ" /><figcaption>Les CI se développent en acquérant de l&#39;expérience au sein de leur équipe, par le biais de projets de <em>Chapter</em>, et parfois en passant dans une autre équipe. Cette expérience accumulée les fait progresser au sein de l&#39;échelle de carrière de leur rôle.</figcaption></figure><p>En investissant dans la mobilité interne et les responsabilités des <em>Chapters</em>, les organisations créent un environnement de travail dynamique et épanouissant pour les CI. Elles fidélisent les talents précieux et assurent leur succès à long terme.</p><p><em>Attention : les CI ne croient pas qu’aux promesses. </em>Ils veulent de vraies opportunités ; les organisations doivent donc remplir leurs engagements. Mesurez le nombre de mouvements de transferts internes. Rapportez le temps consacré aux sujets du <em>Chapter</em>. Énumérez les décisions que les CI prennent par rapport aux managers. Ce sont des critères mesurant l&#39;impact de votre stratégie de croissance pour les CI. Dernier point, sans doute le plus important : la politique de rémunération doit soutenir cette stratégie. Cela signifie que certains CI gagneront plus que leurs managers. Cela se produit dans nos équipes Data. Ne pas aligner la rémunération et les opportunités réduira vos efforts à neant !</p><p>Quelques chiffres sur les équipes Data de BlaBlaCar.</p><ul><li>Depuis le deuxième trimestre 2023, nous avons eu plus de 10 % des CI qui ont changé d’équipe. Cela exclut les changements de ligne de management durant les réorganisations du département.</li><li>Nous avons 4 rôles de responsables de <em>Chapter</em> depuis 2022, avec 6 CI différents ayant été ou étant dans ce rôle.</li><li>Nous avons créé des postes de <em>Senior</em> et de <em>Staff </em>pour les CI, clarifiant les attentes et créant des modèles.</li><li>Au deuxième trimestre 2024, la répartition des CI dans le département est de : 26 % <em>Junior</em>, 45 % <em>Confirmed</em>, 21 % <em>Senior</em>, 2 % <em>Staff</em>. <em>NB : ces chiffres ne somment pas à 100 %, car le niveau de certains postes ouverts est encore indéterminé.</em></li></ul><h3>Retourner à l’école</h3><p>Les managers font des erreurs et ces erreurs sont coûteuses. C’est parce qu’elles ont un impact sur les employées, et qu’elles ne sont généralement pas réversibles. La formation des managers les aide à éviter ces erreurs. Décider quand commencer la formation au management est difficile : trop tôt, c’est très théorique ; trop tard, et les erreurs s’accumulent. Ma recommandation est de former les managers au bout de 6 mois en poste. Chez BlaBlaCar, tous les managers suivent le même programme de formation. Au-delà de cette formation initiale, je suis favorable à l’organisation de sessions d’amélioration continue. Au sein des équipes Data de BlaBlaCar, nous avons mis en place une session de co-développement d’une heure toutes les 6 semaines. Au cours de cette session, 5 à 6 personnes résolvent ensemble le problème d’un.e participant.e. Nous trouvons <a href="https://docs.google.com/document/u/0/d/1jOR3gICPnnDjNmnqO5er4ejIYf7rsk4VCr5hBURHtuc/edit">ce cadre</a> très efficace, tant en termes d’amélioration continue que du renforcement de l’esprit d’équipe. Les managers des équipes Data font également partie d’un groupe plus large, le <em>Engineering Management</em>, qui se réunit mensuellement.</p><p>La formation est indispensable pour les nouveaux manager. Elle est également précieuse pour les managers en poste, contribuant à la construction d&#39;une culture de management cohérente dans toute l&#39;entreprise. La formation en management a un avantage secondaire : clarifier l’ensemble des compétences attendus des managers. Cela s&#39;avère utile pour le développement de la carrière des CI, en clarifiant les facteurs clés du succès dans la voie managériale. Ces connaissances leur permettent de prendre des décisions plus éclairées lorsqu&#39;ils envisagent de devenir manager. Vous trouverez ci-dessous trois compétences en gestion qui sont souvent négligées :</p><ul><li><em>L’exemplarité, et le calme malgré le stress</em>. Les managers veillent aux intérêts de leur équipe et de l’entreprise, même dans les moments difficiles.</li><li><em>La conscience de soi et la polyvalence dans les interactions</em>. Les managers cultivent la conscience de soi pour compartimenter la forme du contenu. Ils doivent maîtriser l’écoute active et les différents canaux de communication.</li><li><em>Le courage</em>. Les managers doivent avoir le courage de prendre des décisions difficiles et de les mettre en œuvre.</li></ul><p>Sans être manager, il est facile de passer à côté de l&#39;importance de ces compétences — et de la difficulté à les maîtriser. Elles sont comme le service public : vous n’en remarquez l’importance que lorsqu&#39;il est défaillant. C&#39;est l&#39;une des raisons de les mettre en évidence ici.</p><blockquote>Le management est plus qu&#39;un titre. Vous devez apprendre à bien le faire. On ne naît pas manager, on le devient pas à pas.</blockquote><h3>Construire son équipe : louer ou acheter ?</h3><p>Lorsqu’on agrandit — ou réduit — une équipe, les managers doivent anticiper les décisions qu’ils prendront dans deux ou trois ans. Cette perspective garantit que la gestion du personnel inclut les besoins et la croissance futurs, et pas seulement les besoins immédiats. Les décisions relatives à la gestion du personnel sont similaires à l’achat d’une maison : vous ne décidez pas d’acheter une maison avec un horizon des 6 prochains mois. Si vous ne pouvez pas vous projeter au-delà des 6 prochains mois, envisagez de louer…</p><p>Un exercice que je fais systématiquement lorsque nous ouvrons un poste dans l’équipe Data est de projeter à 2 ans la répartition des profils suivant notre échelle de carrière. Cela permet d’identifier où nous aurons des goulots d’étranglement, ou une concurrence pour les promotions. Je reviens ensuite à aujourd’hui. Cet exercice me permet de gérer les opportunités de croissance pour les CI. Il permet d’aborder les conversations difficiles en amont, ce que je trouve utile ; nous pouvons les anticiper. Par exemple, lorsque nous remplaçons des postes de <em>Senior</em> par des postes de <em>Junior</em>. Nous parions sur le fait que d’autres profils <em>Confirmed</em> évolueront vers des postes de <em>Senior</em> dans un horizon de 2 à 3 ans. Ce pari déclenche des discussions sur les responsabilités, le mentorat et la productivité. C’est également lors de ces discussions que nous lançons les mouvements de mobilité interne. En adoptant une approche proactive pour anticiper les défis futurs, les organisations se protègent contre les “ce qu’elles savent ignorer”.</p><blockquote>“Vous prenez des décisions de gestion du personnel radicalement différentes en fonction du contexte de l’équipe. Une perspective de gestion sur 6 mois par rapport à 2 ans engendre des dynamiques et des équipes qui n’ont rien à voir.”</blockquote><p><em>NB : une partie du contexte de gestion d’une équipe est sa taille. Elle limite les solutions : une équipe de 2 personnes a moins d&#39;options qu&#39;un département de 15 personnes. Savoir qu&#39;une équipe de 2 personnes doublera de taille — ou diminuera d’autant — en 12 à 18 mois est cependant un facteur de décision important.</em></p><p>Pour conclure, je vous rappelle les trois éléments qui font la différence pour les équipes Data de BlaBlaCar. Premièrement, notre investissement dans la construction de carrières longues pour les IC. Nous avons clarifié leurs responsabilités et leur avons offert des opportunités de croissance. Deuxièmement, l’investissement dans la formation au management. Troisièmement, l’adoption d’une perspective à long terme pour la gestion du personnel, alignée sur la progression dans l’échelle de carrière de l’entreprise.</p><p>J’espère que ces trois conseils vous seront utiles !</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=509395ba1e70" width="1" height="1" alt=""><hr><p><a href="https://medium.com/modern-data-network-mdn/g%C3%A9rer-une-%C3%A9quipe-de-data-trois-le%C3%A7ons-de-blablacar-509395ba1e70">Gérer une équipe de Data : Trois leçons de BlaBlaCar</a> was originally published in <a href="https://medium.com/modern-data-network-mdn">Modern Data Network (MDN)</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Unexpected Tips for Data Managers]]></title>
            <link>https://medium.com/blablacar/unexpected-tips-for-data-managers-c44a71db6594?source=rss-6a00f89edec2------2</link>
            <guid isPermaLink="false">https://medium.com/p/c44a71db6594</guid>
            <category><![CDATA[management]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Emmanuel Martin Chave]]></dc:creator>
            <pubDate>Thu, 02 May 2024 09:06:12 GMT</pubDate>
            <atom:updated>2024-05-03T12:34:52.836Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*biNf-nCfOajnTqDgSpLRoA.png" /></figure><p><em>Based on </em><a href="https://podcasts.apple.com/fr/podcast/datagen/id1539383194?i=1000647920440"><em>my podcast with Robin Conquet on DataGen</em></a><em>, I reflected on three tips to share with fellow Data Managers. First, responsibilities between Individual Contributors (ICs) and Managers. Second, Management training; and third, taking a long-term view on staffing. This comes from my experience managing BlaBlaCar’s 50-strong Data teams for 7 years. I’ll let you decide how this generalizes to other contexts!</em></p><h3>Playing Like an Orchestra.</h3><blockquote>“Becoming a Manager is not a promotion for an IC. It’s a different job.”</blockquote><p>Do musicians expect to be promoted to conductors? The question makes little sense: promotion implies ranking. In this case, there is none, musicians and conductors have different responsibilities. Harmony arises when these elements combine into a consistent and pleasing whole. A Data team behaves like an orchestra. Its performance comes from the interactions between the conductor (Manager) and the musicians (ICs). They are different jobs, so there’s no promotion between them.</p><h3>Divide Distinct Responsibilities for Managers and ICs.</h3><blockquote>“A common misconception is that Managers and decision-makers are the same. The reality is richer. Teams make a lot of decisions. Some belong to the Manager. For instance, deciding a promotion or a pay rise. Most decisions are ICs’: from technical design to project delivery through project management.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*hnL_CMsKYLtX7N0A" /><figcaption>Most decisions in a team are IC decisions. Both them and Manager’s decisions trigger Team actions. Which should lead to further need to make decisions.</figcaption></figure><p>To ensure a well-functioning Data team, it’s crucial to recognize the distinct responsibilities of Managers and ICs. This clear division of responsibilities prevents overlap, which generates confusion. It allows each role to focus on its core functions, improving delivery. In BlaBlaCar’s Data teams, we’ve implemented the following distinctions:</p><ul><li><strong>Managers create value by focusing on connections.</strong> Inside the team, that’s people management; it connects people together, from projects to processes. These include decisions on staffing, performance management, career development, team dynamics. Outside the team, Managers nurture connections with the broader company ecosystem. They share information, represent their teams, or handle roadmap dependencies.</li><li><strong>ICs create value when leveraging expertise to implement high-quality solutions.</strong> This requires decisions on designing and implementing data solutions, conducting analysis, and clarifying stakeholders’ needs. ICs are like musicians: they perform technically demanding art. They decide how to play their part, in collaboration with the rest of the orchestra.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*3L58UcAdAOt5pY9w" /><figcaption><em>Managers create value by nurturing the connections between teams. Abundant context leads to higher performing teams. They have more information, which enables them to make better decisions.</em></figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*BZjxx6uPXsZpdNe5" /><figcaption><em>Inside the team, Managers focus on connecting people together. They also create new connections — through processes for instance — to make ICs successful.</em></figcaption></figure><p>These aren’t exhaustive lists. Build the one relevant for your team; and evolve it over time. This division of responsibilities allows Managers to provide effective leadership and support, while ICs can focus on their technical expertise and drive innovation. By clearly defining and separating the responsibilities of Managers and ICs, organizations create efficient data teams.</p><h3>Shift Technical Decision-Making to ICs.</h3><p>Making bad decisions leads to failure; so, try to avoid them. ICs typically possess the deepest technical expertise and hands-on experience. That makes them well-equipped to make informed decisions on the technical aspects of their work. For example, ask the Data Scientists and Engineers what their recommended AI architecture is vs. buying it off the shelf and telling them to “just use it”. Don’t ask managers to build your stack. Instead, have them set constraints for solutions. Then, give ICs the autonomy to pick solutions based on their context, their expertise and these constraints. That limits bad decisions, and saves time.</p><p>In BlaBlaCar’s Data teams, we implemented this through our “Data Stack Governance” and “Chapters” forums. We copied existing, well-functioning forums from the broader Engineering department, which Data teams are part of. There are the “BlaBlaCar Architecture Team” — codenamed “BAT” 🦇 — to review architecture decisions, the Backend Guild, the soon-to-be Software Factory Guild, …</p><p>Shifting decision-making to ICs empowers them. That’s because in an organization, decision-making gives power. Empowerment fosters a culture of ownership and accountability. By empowering ICs to make decisions, organizations reduce bottlenecks, speed up delivery, and improve the quality of their data products. Shifting technical decision-making responsibilities to ICs also helps to develop their leadership and problem-solving skills. This prepares them for senior roles. By trusting ICs to make these decisions, organizations foster a culture of innovation and continuous improvement.</p><h3>Create Growth Opportunities for ICs.</h3><p>To retain and empower talented ICs, organizations must provide clear, actionable opportunities. There are three powerful levers to create opportunities for ICs:</p><ul><li><strong>Internal mobility allows ICs to experiment with different roles and projects</strong>. They capitalize on their tenure, while discovering new topics. Moving to a new team helps share knowledge; it gives Managers more staffing flexibility, including for promotions. By providing these opportunities, organizations kickstart a virtuous circle: reduced turnover means a well-rounded workforce, which is more impactful, creating new opportunities.</li><li><strong>Chapter projects empower ICs with project ownership and leadership responsibilities within their skillset</strong>. A Chapter is the forum for people using the same technology. In BlaBlaCar’s Data teams, we have four Chapters: Data Engineering, Data Analytics, Data Science, and Data Back-End. It’s a great avenue to improve technically and deliver impact transversally. This enables organizations to create new opportunities for ICs, like the Chapter Lead role. The Chapter Leads guide and mentor other ICs in their area of expertise, define standards, identify talent, and represent the Chapter’s interests. By establishing these positions, organizations foster a sense of community, provide valuable mentorship, and increase job satisfaction and team performance.</li><li><strong>Career Ladder gives ICs perspective on future responsibilities</strong>. It sets expectations for how to grow. In the Data teams, we saw a significant change in behavior when we started promoting ICs to Senior roles. This created concrete role models to look up to. An added benefit of building a career ladder: it makes Managers’ assessments more objective.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*4CjFlesly14p00lH" /><figcaption>ICs can grow through acquiring experience in their team, through Chapter projects, and sometimes by switching to another team. This accumulated experience move them up in the career ladder.</figcaption></figure><p>By investing in internal mobility and Chapter responsibilities, organizations can create a dynamic and fulfilling work environment for ICs, retain valuable talent, and drive long-term success.</p><p><em>Beware: ICs don’t buy promises.</em> They want real opportunities; organizations must walk the talk. Measure how many internal mobility moves your organization delivers. Report on the time dedicated to Chapter topics. List which decisions ICs are making vs. Managers. These are yardsticks measuring the impact of your growth strategy for ICs. The last and most important point: compensation must support this strategy. This means that some ICs will earn more than their Managers. It happens in our Data teams. Failing to align compensation and opportunities dooms the effort!</p><p>Some facts about Data teams at BlaBlaCar.</p><ul><li>Since Q2 2023, we had ≥ 10% of the ICs doing internal mobilities. That excludes reporting line changes for reorganizations.</li><li>We have 4 Chapter Lead roles since 2022, with 6 different ICs having been or being in this role.</li><li>We created Senior and Staff positions, clarifying expectations and creating role models.</li><li>As of Q2 2024, the seniority distribution of ICs in the department is: 26% Junior, 45% Confirmed, 21% Senior, 2% Staff. <em>NB: this doesn’t add to 100%,</em> as <em>we have open positions whose level is TBC included.</em></li></ul><h3>Going Back to School.</h3><p>Managers make mistakes and these mistakes are costly. That’s because they impact people, and are typically not reversible. Training helps avoid these errors. Deciding when to start management training is difficult: too early, it’s very theoretical; too late, and mistakes pile up. My recommendation is to train Managers after 6 months. At BlaBlaCar, all managers take the same training program. Beyond this initial training, I support organizing recurring sessions for continuous improvement. Within BlaBlaCar’s Data teams, we set up a 1h co-development session every 6 weeks. In this session, 5–6 people collaboratively solve one participant’s problem. We found this <a href="https://docs.google.com/document/u/0/d/1jOR3gICPnnDjNmnqO5er4ejIYf7rsk4VCr5hBURHtuc/edit">framework</a> very efficient, both in terms of continuous improvement and team building. Data Managers are also part of the broader “Engineering Management Monthly Meeting”.</p><p>Training is indispensable for new managers. It’s valuable for tenured managers as well, contributing to building a consistent management culture throughout the company. Management training has a side benefit: clarifying managerial skill sets. This proves useful for ICs’ career development, clarifying the key factors of success in the managerial track. This knowledge equips them to make more informed decisions when considering a move into management. Below are three managerial skills that are often overlooked:</p><ol><li><em>Exemplarity, composure under pressure</em>. Managers keep their team’s and company’s interest in mind, even through challenging times.</li><li><em>Self-awareness and versatility in interactions</em>. Managers cultivate self-awareness to compartmentalize form from content. They should master active listening and various communication channels.</li><li><em>Courage</em>. Managers must have the courage to make tough decisions, and to carry them out.</li></ol><p>From the outside, it’s easy to miss the importance — and difficulty — of these competencies. They are like utilities: you notice them only when they are missing. That’s one reason to highlight them here.</p><blockquote>Management is more than a job title. You need to learn how to do that well. You are not born a manager, you painstakingly become one.</blockquote><h3>Staff Like You Buy a House.</h3><p>When scaling a team, managers must anticipate decisions they’ll make in two or three years. This perspective ensures that staffing includes future needs and growth, not only immediate ones. Staffing decisions are similar to buying a house: you don’t decide to buy a house looking 6-month ahead. If a 6-month time horizon is the maximum you can get, consider renting…</p><p>One exercise I run systematically when we open a position in the Data team is to project our seniority distribution in 2 years. That identifies where we’ll have bottlenecks or competition for promotions I then walk backwards to today. This exercise ensures I manage growth opportunities for ICs. It frontloads difficult conversations, which I find useful; we can anticipate them. An example is when we backfill Senior roles with Junior ones. We bet on other Confirmed profiles growing into Senior roles within a 2–3 year horizon. That bet triggers ownership, mentorship, and deliver discussions. That’s also when we kickstart internal mobility moves. By taking a proactive approach in anticipating future challenges, organizations protect themselves against known unknowns.</p><blockquote>“You make wildly different staffing decisions depending on the team’s context. A 6-month vs. 2-year staffing perspective yields specific dynamics and teams.”</blockquote><p><em>NB: part of the staffing context is the team size. This limits solutions: a 2-people team has less staffing options than a 15-people department. Knowing that a 2-people team will double in size — or shrink — in 12–18 months is an important decision factor though.</em></p><p>To wrap up, I’ll point out three things that made a difference for the Data teams at BlaBlaCar. First, our investment in building meaningful IC careers. We clarified their responsibilities and provided them with growth opportunities. Second, investing in Management training. Third, taking a long-term perspective for staffing, aligned with progression through the company career ladder. <br>I hope you’ll find these three tips useful!</p><blockquote>Why do ICs declare they want a managerial position? Too often, that’s the only way they think they can grow. When you break that mental barrier, you see ICs staying on their track longer.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c44a71db6594" width="1" height="1" alt=""><hr><p><a href="https://medium.com/blablacar/unexpected-tips-for-data-managers-c44a71db6594">Unexpected Tips for Data Managers</a> was originally published in <a href="https://medium.com/blablacar">BlaBlaCar</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Scaling Data Teams: 5 Learnings from BlaBlaCar]]></title>
            <link>https://medium.com/blablacar/scaling-data-teams-5-learnings-from-blablacar-9e00949957f3?source=rss-6a00f89edec2------2</link>
            <guid isPermaLink="false">https://medium.com/p/9e00949957f3</guid>
            <category><![CDATA[data-team]]></category>
            <category><![CDATA[data-mesh]]></category>
            <category><![CDATA[organization]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[management]]></category>
            <dc:creator><![CDATA[Emmanuel Martin Chave]]></dc:creator>
            <pubDate>Fri, 30 Jun 2023 08:53:12 GMT</pubDate>
            <atom:updated>2023-06-30T08:53:12.014Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*35PQZ3yC0Wf0HVd_N9-9GQ.png" /></figure><p>Start simple, be pragmatic, mirror the business, create Chapters and invest in your platform.</p><h3>Now and then: today’s organization and where we come from</h3><p>As of June 2023, BlaBlaCar’s Data department sits within the Engineering organization. Its 45 people are divided into 6 teams: 5 multidisciplinary squads, and one platform-oriented team. The squads are composed of Data Analysts, Data Scientists, Software Engineers and Data Engineers. That makes them autonomous on all data projects.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*c6yOTmqspg1fQckZ4y590g.png" /><figcaption>Multidisciplinary squads: autonomy, together!</figcaption></figure><p>In 2021, the Data department was organized by technological layers: a Data Engineering team, a Data Science team, a Data Analytics one, etc. This organization optimized for skill-based expertise. It also created a layered technical stack; interoperability between layers was low. This showed whenever one project required work across two or three layers.</p><blockquote><em>The stack is the consequence of the organization (Conway’s law). An organization by centers of expertise produces a technologically-layered stack.</em></blockquote><p>This model stopped scaling in 2021: Data teams became a bottleneck. Their delivery speed of large projects slowed; prioritizing work across technological layers was difficult.</p><p>Transverse topics, like data quality, had ambiguous ownership. Integrating data from M&amp;As took years. Our data architecture and organization were rigid; this made it costly to address new use cases, e.g. streaming data for ML applications.</p><p>The need for organizational change was obvious. Data teams needed multidisciplinary squads tied to business domains. Thus, they could operate effectively, autonomously and locally scaling with business needs.</p><h3>Getting inspiration from Engineering</h3><p>The principles of the Data Mesh inspired me to change our paradigm.</p><p>This architecture decouples technological layers from teams’ organization. Each squad becomes responsible for managing operational and analytical data within its business domain(s). To enable this ownership, they must be autonomous; that’s why we have multidisciplinary squads. These squads can scale at different speeds and directions, according to business inputs.</p><p>The paradigm was enticing. But I didn’t know if it would work in practice.</p><p>I needed concrete examples; unfortunately I couldn’t find many Data teams that implemented a Mesh: <a href="https://martinfowler.com/articles/data-monolith-to-mesh.html">Zhamak Dehghani’s foundational article</a> was just 2-years old. So I took best practices from our Product and Engineering colleagues. They experienced similar challenges. Multidisciplinary feature teams are common in Software Engineering. Learning from Product and Engineering teams at BlaBlaCar, I validated the paradigm and got tips on its implementation.</p><h3>5 Learnings from our experience</h3><p>Changing a 40+ people organization doesn’t happen overnight. We made incremental changes to incorporate new knowledge:</p><ul><li>Q3 2021: Diagnosis. Understand the problem that the department faces. With a clear problem statement, build a vision towards the solution.</li><li>Q4 2021: Onboarding managers. Align on the problem statement and the target vision. If they do not <em>own</em> change, they will not lead it effectively.</li><li>Q1 2022: Creation of 2 squads, out of 6 new teams in the target vision.</li><li>Q2 2022: Integrating feedback and creating the remaining 4 teams.</li><li>H2 2022: Test and learn. Share operational insights between squads.</li><li>2023: Focus on the technical migrations to complement the organizational changes.</li></ul><p>Below are our five learnings from this experience.</p><h3>1. Where to start?</h3><p>We picked the easiest areas to make independent. We factored in known internal moves, accelerating some changes. We created last the hard-to-isolate domains, leveraging experience from our previous moves.</p><h3>2. Should we work on all aspects of the Data Mesh at the same time?</h3><p>No. Aim for something pragmatic, not perfect. This change impacts our stack, skills, management, processes, and mindset. It is not necessary to excel in every aspect to achieve a functional implementation.</p><p>At BlaBlaCar, we invested the least in the <em>Data As A Product</em> aspect. For instance, there is no Data Product Owner for each domain at BlaBlaCar. Likewise, squads do not operate 100% of their ingestion pipelines. Some are still centralized. Decentralizing them added non-urgent work for squads, while they already had substantial changes to absorb.</p><h3>3. Are there any changes to be made outside of Data, especially on the Business side?</h3><p>There weren’t for us. That’s because we designed Data teams to mirror the business organization. The Data department is an independent entity reporting to the CTO. By molding our teams on business areas, we maintained close links with stakeholders.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*WGBfec5AfaXpdEZfD4I0wQ.png" /><figcaption>Mirroring business: organize your teams along business lines. Or spend time solving misalignment…</figcaption></figure><h3>4. What about keeping the expertise for each job?</h3><p>BlaBlaCar wanted to maintain common practices among experts in squads. So we created Chapters.<strong> </strong>They are communities by expertise, where peers meet, challenge their technical choices and learn. Chapters embody the notion of federated governance. They guarantee the convergence of practices despite skill distribution among squads.</p><blockquote>Autonomy and boundaries between squads need a counterweight: federated governance and common practices.</blockquote><p>Having Chapter Leads also empowers individual contributors. It prevents the concentration of responsibilities on managers.</p><h3>5. How do we keep the Data machinery well oiled?</h3><p>One transverse team — Data Ops — provides the common infrastructure and services to other squads. For instance, Data Ops builds ingestion patterns consumed by the Data Engineers in the squads. At our current size, this setup works without making Data Ops a bottleneck. As the team grows and matures, some parts start being distributed to squads.</p><p>To avoid creating silos, members from the DataOps team often temporarily join squads. This benefits everyone by: (i) guaranteeing centrally designed tools and patterns meet squads’ needs, and (ii) allowing team members to explore and learn new topics.</p><h3>My advice: clarify which problem you’re trying to solve</h3><p>We succeeded because we knew what problem we wanted to solve. We were no longer efficient at our scale. This problem statement was our beacon when making difficult decisions. My best advice here is to invest time to align on the problem. Don’t implement a solution if you aren’t clear what problem it’s solving!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9e00949957f3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/blablacar/scaling-data-teams-5-learnings-from-blablacar-9e00949957f3">Scaling Data Teams: 5 Learnings from BlaBlaCar</a> was originally published in <a href="https://medium.com/blablacar">BlaBlaCar</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Thoughts on Written Communication.]]></title>
            <link>https://medium.com/@manu.martin.chave/thoughts-on-written-communication-9cd6814c1bc3?source=rss-6a00f89edec2------2</link>
            <guid isPermaLink="false">https://medium.com/p/9cd6814c1bc3</guid>
            <dc:creator><![CDATA[Emmanuel Martin Chave]]></dc:creator>
            <pubDate>Wed, 21 Jun 2023 07:34:54 GMT</pubDate>
            <atom:updated>2023-06-26T13:17:00.081Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vlX_NRWqQ9aHrRQ37TbJ6g.jpeg" /></figure><p>I never read something from M. Leonard. But this quote summarizes efficient writing. It’s simple, memorable and reader-oriented.</p><p>The below sums up my own rule to writing efficiently. I shared this to my teams at <a href="https://medium.com/u/aeddeb1dc8cd?source=post_page-----714cf44738f3--------------------------------">BlaBlaCar</a> in April 2021. It helped people write shorter, crisper documents. That’s always helpful. Efficient writing became business critical in the context of our increasingly asynchronous work.</p><p>After receiving positive feedback on this piece, I’m making it public hoping it helps more people. Thanks in advance for your comments and feedback!</p><p>My goal is to share principles that maximize the signal-to-noise ratio in written communication. It applies to documents, not chats. Following them is not mandatory, but strongly encouraged.</p><p>The strategy is to increase the amount of signal and reduce the amount of noise. Increasing signal starts with asking questions. Why are you writing? To whom? What’s in it for them? The writer must answer these questions before penning a word. Reducing noise means whittling and disambiguating. To crystallize these, here are 3 principles:</p><ol><li>Start with: why, to whom and what’s in it for them. You can’t answer all the questions? Stop: you’d be writing for yourself.</li><li>No sentences longer than what fits in your short-term memory.</li><li>Build pyramids: the first sentence recaps the following ones, which defend the first.</li></ol><p>Here are three tools to estimate the signal-to-noise ratio: <a href="http://gunning-fog-index.com/index.html">Gunning Fog Index</a>, <a href="https://app.readable.com/text/?demo&amp;_ga=2.81538068.268573781.1617894643-1317098466.1617894643">Readable</a> and <a href="https://hemingwayapp.com/">hemingwayapp</a>. Two interesting articles: <a href="https://medium.com/@raphael.moutard/writing-the-amazon-way-1f853cbe6937">writing the amazon way</a>, <a href="https://review.firstround.com/a-founders-guide-to-writing-well">founders guide to writing well</a>. Please share if you have other good ones.</p><p>Here is what I wrote to myself before starting this document.</p><h3>Why am I writing this?</h3><p>Having defaults foster cohesion in a group. Cohesion matters because it sustains groups to tackle complex problems. It also makes them resilient. Cohesion achieves that by improving communication and decision-making.</p><p>I want to lay out strong defaults for my staff to promote its cohesion. Starting by writing principles. It’s the start of a series of food for thought. The next one will be on decision-making.</p><h3>To whom?</h3><p>My manager for feedback and then my staff.</p><h3>What’s in it for them?</h3><p>Probably some frustration/anxiety at first. But that’s the investment to make the group more cohesive. Written communication is essential for remote work. Improving it benefits everyone.</p><p>You wrote a sentence. Great! Read it. Close your eyes. Say it aloud. You can’t? It’s because it’s too long. Remove what’s not essential. And repeat the process.</p><p><em>We are in a very competitive field so we have to always be better than the other advertiser in order to get the cheapest cost of traffic (= having the right ads, the cleanest accounts,…) and the best final performance (= being present on the right keywords at the right time &amp; price).</em> <em>We are in a competitive field. Having the right ads lowers our cost of traffic. Bidding the right keywords improves our conversions. Both are needed to beat competition.</em></p><p>Whittling and disambiguating are difficult. They require replacing adjectives and adverbs by numbers: replace <em>much</em> by <em>5</em>, <em>excellent</em> by <em>4.5 / 5</em>. They kill the weasel and peacock words. E.g. <em>greatly</em>, <em>somewhat</em>, <em>few</em>, <em>low</em>, <em>a lot</em>, etc. They force the use of absolute references: <em>09/04/2021</em> in lieu of <em>tomorrow</em>.</p><p>Keep one idea per paragraph. Start by your conclusion. It saves everyone’s time. Expose the rationale or details after. You need arguments to convince, or details to explain. Conclude by what you started with. You can follow up with another idea in a separate paragraph. That helps you keep one idea per paragraph.</p><p><em>We’re looking for someone who enjoys making data available by building robust data pipelines, doing project-specific engineering, and ensuring its quality and usability. That person is at the intersection of data engineering and data analysis, coupling capabilities in data processing and optimization with a deep understanding of how and for which purposes analysts use Product data.</em> <em>The successful candidate mixes two skills: engineering and product-thinking. Engineering skills ensure a high level of quality. For instance, when building data pipelines. Product-thinking guarantees clear problem statements. This requires understanding stakeholders’ needs. It’s the mix of engineering and product-thinking that creates success.</em></p><p>The worst foes to writing efficiently. They add no signal, and sometimes confuses readers. They are common because they are cheap and look good. Peacock terms promote a subject without providing attribution. Weasel terms tend to promote the thought that something is widely accepted, whether it is or isn’t. That’s why they add noise to your writing. Remove them whenever you can.</p><h3>Examples of Peacock words</h3><p><em>One of the best, a well-known, the indisputable, an iconic, the legendary, immensely, one of the greatest, outstanding, a world-class, most respected, among the most notable.</em></p><h3>Examples of Weasel words</h3><p><em>Many, few, some, lots, little, somewhat, often, lower, higher, clearly, probably.</em> <em>😱 Mark Zuckerberg is one of the most influential people of 2010.</em> <em>🤩 Mark Zuckerberg was named </em><a href="https://content.time.com/time/specials/packages/article/0,28804,2036683_2037183_2037185,00.html"><em>2010 Person of the Year by Time Magazine</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9cd6814c1bc3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>