Have you ever wondered how to choose between Snowflake, BigQuery or Redshift? What’s best between Tableau and Looker? To whom a data team should report? Who to hire first: a Data Engineer or a Data Scientist?
If you’re familiar with these questions, we know how you feel. If these decisions impacted your work, we experienced that too. And if you couldn’t help someone struggling with them, we bet what’s following could prevent that next time.
The Modern Data Network’s objective is to build a community to inform data teams’ decision-making. We’ll achieve it by sharing experience and learning from each other. We believe this will help make better decisions. And ultimately facilitate data teams’ mission to deliver value to their stakeholders. NB: by data teams we mean the whole data family: analysts, engineers, scientists, etc.
Data teams make decisions determining their companies’ ability to leverage data. The data value chain is complex and fragmented. Data teams must build their stack from a cornucopia of tools. It is moreover increasing at a rapid pace. Accessing reliable experience is very slow. All these reasons make their job difficult.
We–Edouard, Louise, Manu and Nolwenn–all struggled to find such a community when we needed it. Thus, we decided to build the environment to foster one. This is the Modern Data Network initiative. 100 data leaders have joined it in two months. They build content for one another and will share it on this Medium account, through meet-ups and via Slack. Join us!
Why is it so difficult to make decisions for data teams?
At the beginning there was no data. And then, in between two fundraises, the C*O created the data team. They hire one promising young leader, set unrealistic ambitious objectives, provide the embryonic team–or not–with extravagant budget (according to the CFO) for staffing and tooling, cook the whole thing under a lot of pressure, and patiently wait a couple of weeks for this team to produce the data fuel doubling the company’s growth rate.
That’s caricatural, but some of it does ring a bell, doesn’t it?
Data teams struggle making decisions on their strategy for three reasons. We think that the last one is a pain point we can solve:
- They are consequential, expensive and hard to reverse;
- The data ecosystem is fragmented into hundreds of tools and evolving daily;
- They don’t have fast access to draw on reliable experience.
Here is a concrete decision example: selecting a dashboarding tool for the company. Pick one that’s inadequate and few stakeholders will use it. That typically triggers a flood of ad-hoc requests swamping the data team. After realising you got it wrong, you relent, and upgrade to a better one. You now face the daunting triad of migrating reports, re-training stakeholders, re-writing data pipelines. This is a consequential, expensive and hard to reverse decision.
There are tons of articles picturing the data ecosystem, here is one if you’re curious. We won’t go into the fragmentation of data ecosystem today and the pace at which it evolves (but that’s a great idea for another day). Both make it hard to make decisions. Let’s move on to #3 then.
Our own experience taught us the perils and anguish of making decisions without the ability to draw on internal expertise. “Sure”, as Edouard recalls, “there are tons of Meet-ups, Medium articles or Slack groups, but they focus on specific technical problems, that never exactly matches yours”. He found no single place aggregating knowledge about how to organize your team, examples of data strategies, exhaustive stack benchmarks, etc.
Louise spent hours scouting for external expertise to learn from to prepare for her new role. “I met with 15 people–who helped me a lot–but this doesn’t scale for routine decision-making. It’s too slow”, she said. Nolwenn concurs: “It’s too bad to be on your own. Thanks to others’ experience, you avoid reinventing the wheel and make 10x faster decisions, with more confidence”. The information we need exists, but it’s too slow to get it compared with the pace of decision-making.
As a result, we end up relying on information from partial vendors–they react very quickly–, or unrealistic examples from the GAFAs–which are available online. But that’s not reliable information. “When I am discussing with vendors about bits of my teams’ stack”, says Manu, “it’s hard to go beyond the sales pitch. Understanding how users adopt a tool is as important as the features it offers”.
The reliable experience we want to leverage exists: it’s there, in every decision-maker head. Slow access makes using it prohibitive. With the MDN, the community aggregates its collective experience and access it instantly.
How do we propose to do it?
We believe that community-powered content perfectly solves the pain point above–fast access to reliable experience–because: it scales, adapts to the ebb and flow of the data ecosystem, avoid bias through self-moderation, and represents the plurality of opinions indispensable to make the complex decisions data teams face.
Communities thrive when their members adhere to common rules. Here are the principles the Modern Data Network abides by:
- Tolerance. There are no stupid questions, only dumb listeners. Being argumentative is ok, being disparaging is not.
- Equality. There is no hierarchy in this group; you can hung up your fancy “VP” title at the door before coming in. NB: yep, like all stories, it had to begun with an initial cast. But Louise, Edouard, Manu and Nolwenn are here to provide the infrastructure and to coordinate the logistics, nothing more.
- Proactive help. The community thrives when its members proactively help each other out. Feedback is a gift, please be magnanimous in offering it and everyone will have plenty of it!
- No self-promotion. While vendors are welcome to contribute to the pool of knowledge, this group is not a marketing channel to engage with prospective clients. Likewise, this is not about grandiose storytelling of imaginary successes, we aim for true and honest content.
What’s in it for you?
We’re happy that you’ve read to this point, congrats and thanks for bearing with us! Now comes the treat and the specifics of what the Modern Data Networks puts in place to help you make decisions about all things data-related. Content created by the community is shared through three channels: our Medium blog, meet-ups and a Slack workspace.
Here on Medium, you’ll find articles published by the community on topics they’d like to share about broadly. Expect readings like: how to grow a data team from scratch, benchmarks about tools, thoughts about organizing teams, etc.
Meet-ups are dedicated to specific topics, to promote live interactions. The first one took place on Wednesday, March 24th. 5 speakers shared how tracking is implemented in their teams, and discuss the merits and drawbacks of their tracking stack. Meet-ups are public and you can follow the agenda here (yeah, there are no events planned currently 😅 and yes, we will work on a better system).
The Slack workspace is an invitation-only channel where senior data leaders share their experience and current challenges on managing data teams. This workspace discussions feed the articles published on this Medium blog. It ensures the Modern Data Network community benefits from high-quality content written by professionals having experienced what they broadcast. It also triggers workshops. They are invitation-only events dedicated to making decisions, focusing on Q&A. A workshop will happen in April, comparing two visualisation tools in vivo: Looker & Tableau.
True to our principle on proactive help: please share back with us how useful this content is, as well as topics you’re interested in reading about. Send us an email to firstname.lastname@example.org, we’re looking forward to it!
Thanks for reading this far and see you soon in the MDN community!
The MDN founders: Edouard Flouriot — Aircall, Emmanuel Martin-Chave — BlaBlaCar, Louise Columelli — AssoConnect, Nolwenn Belliard — Tiller.