Build knowledge pipelines, not data pipelines.
Knowledge management is critical to analytics teams success.
👋 Robert here! This week we’ve got a great guest post from Matt Blasa on knowledge management.
Matt runs a data consultancy called Aspire360, where he provides support for data & analytics initiatives (you can get more info by contacting him at info@aspire360.com). You can also follow him on LinkedIn, Twitter, or
, where he regularly shares insights around data work. And as always, if you enjoyed this post, I’m sure he’d appreciate a follow/like/share.Finally, the themes here around knowledge management are precisely what we’re solving with Hyperquery. You can sign up today for an account at hyperquery.ai, or read more about our knowledge management tips in our documentation. 🚀
We’ve all felt the high of cracking a data puzzle — writing a bit of elegant code that resolves disparate data points into a titillating pattern. It’s why most of us get into data in the first place: to chase that high. But as I’ve gained more experience not just doing individual work, but managing teams, I’ve felt a different kind of breakage, after the work is done: great insights will often get lost.
It’s ironic, really. We analytics folks have access to a once unimaginable wealth of information and tools, yet the data that matters — the insights — can become buried or siloed. We care so much about building systems that efficiently disseminate data, but we rarely consider how knowledge is handled.
In particular, I believe there are three primary difficulties I’ve found analytics teams run into with regard to knowledge management:
Delivery: Turning knowledge into action
Curation: Retaining knowledge
Incorporating (1) and (2) into team workflows
Let’s talk about knowledge systems.
Turning knowledge into action
The objective of analytics organizations is clear: streamline decision-making. And knowledge management isn’t just about post-hoc insight management — it’s the infrastructure that enables the insights generated by a data team to be leveraged.
And the first step to accomplishing this is to make insights not just available, but actionable — crafting the knowledge in a way that it is obviously useful; taking raw insights and molding them into strategic plans with laser focus. Robert’s written a bit on this, so I’ll share a few of his recommendations here: recommendations should be given; work should be done with the objective in mind (ask why!); don’t waste your time sharing details to show off your pedantry, but align your work to your stakeholder’s decision space.
One key hindrance to this, however, is poor signal-to-noise ratio in modern organizations: drowning in a sea of data isn't an exception; it's the norm. Good knowledge management should act as the sieve, separating signal from noise. Implement a standard process by which work is shared and stored, so stakeholders know what to pay attention to. One clear, articulate insight is far more valuable than 100 unmaintained dashboards.
In essence, knowledge management is the bedrock on which analytics teams build their value. It's not just about collecting insights; it's about delivering them in the most effective, timely manner to those who will act upon them. It's the difference between being data-rich and insight-driven.
Knowledge retention: "tacit" vs "explicit" knowledge
Picture this: you're deep in an analytics dashboard and a subtle blip catches your eye. Not an algorithm, not a set rule, but your gut tells you something’s off. You share that hunch, and your team is able to dodge a big bullet. It’s happened to me many times. It's a moment that highlights the duality of knowledge in analytics. We often think of knowledge as synonymous with documentation, but this is only one side of the coin: knowledge can be either explicit, as with documentation, or tacit.
Explicit Knowledge: It's all the things you can write down: guidelines, code comments, SQL queries, you name it. These are the basics, and yes, you absolutely need them. They ensure consistency and get newcomers up to speed. To manage this, use a well-organized internal wiki. Make sure it's searchable and make sure people actually use it. Got a new way of doing things? Update the wiki. It's that simple.
Tacit Knowledge: This is the intuition you've honed over years, the 'read-between-the-lines' skills that you can't fully articulate. It's what makes a good analyst invaluable. But how do you share something so intangible? Open dialogue, mentorship, even casual water-cooler talks can transfer tacit knowledge. Consider implementing lunch and learns or regular debrief sessions to share what worked and what didn't in recent analytics projects.
Build systems that promote retention of both types of knowledge. A robust knowledge management platform can help for explicit knowledge, while group sessions can ensure tacit knowledge is transferred.
What does this look like, when done well?
I’ve found that thriving and vibrant analytics teams often are the way they are as a result of efficient knowledge management. And the effects are far-reaching: great knowledge management doesn't just improve decision-making. It’s an insight force multiplier that expedites the analytics projects and fosters better team operations.
Think about a time when you were racing against a deadline. Your analysis was almost complete, but you were missing a vital piece of information that a colleague had previously worked on. Because of robust knowledge management, you could swiftly locate this critical data, incorporate it into your project, and meet the deadline. Without such a system in place, you'd be rummaging through emails, notes, or databases, wasting precious time and potentially compromising the project's integrity.
The central tenet here is efficiency—getting the right information to the right people at the right time. This is a key fundamental I learned as a leader: speed of knowledge retrieval can change how data teams operate during times of extreme pressure, when ad hoc data retrieval is typically the default mode of transaction with stakeholders.
Moreover, another side effect of this operational efficiency is an increase in job satisfaction. When team members can easily find what they need, they're more likely to feel competent and valued, which contributes to better morale and productivity. Thus, effective knowledge management not only enhances the technical aspects of analytics work but also the human elements of team collaboration and job satisfaction.
With an organized approach to knowledge management, decision-making becomes a streamlined process. Team members can rely on a centralized repository for insights, which ensures not only that the decisions are sound but also that they are made promptly. Effective tooling is very useful in this situation (Hyperquery is great for storing insights, for instance). Tools need to not only be able to manage the knowledge, but let different users explore the knowledge context.
Final Thoughts
Good knowledge management aims to preserve tribal knowledge, and helps analytics teams deliver a regular flow of successes. I’ve discovered, time and time again, that it gives the punch needed to help analytics teams and data strategy succeed. Now I’m not saying this is going to be overnight success — it’s not. But laying the foundations of knowledge management now is critical. And time and time again, I’ve seen it build a culture of transformation.
To overlook the importance of knowledge management is to jeopardize the team's efficiency and the quality of its work. Given how central analytics has become in shaping organizational strategies and operations, such negligence is untenable.