Image by creativeart on Freepik

Data teams are increasingly working like software engineering teams, embracing engineering and development tools to manage their work. These range from version control systems like Github, to adopting agile practices such as Kanban and Scrum, and include ceremonies like daily standup, sprint commitments and sprint demos. Purpose-built solutions (like dbt for data modeling, testing and integration) have come to market, supporting the software engineering mindset. These solutions power large, distributed data teams to do their best work.

But when it comes to collaboration between data teams and the rest of the business, there is still a lot of room for innovation.

Even the most forward-thinking data-driven organizations still rely on standard collaboration tools and practices (e.g. Slack, email or regularly scheduled meetings) to manage communication between their data teams and business stakeholders. After all, why not? Shouldn’t the data team and its workflows resemble other functions in the organization? This argument, and behavior, works when the interactions are relatively generic in nature. But in situations where team dynamics are more complex (and data is more central to every important conversation and decision), this reliance on generic solutions is insufficient.

As data becomes more central to business operations, data team members often need to wear multiple hats.  In some cases, they need to function as product managers by understanding business users' needs, so they can evolve the data platform.  In other instances, they are required to handle ad hoc requests in a support capacity.  In yet other situations, they need to onboard new users and help them to engage with the data assets available to them.

Generic collaboration tooling and traditional approaches to managing work quickly break down in these scenarios. Product teams and support teams have purpose-built tooling to manage their work.  Don’t data teams also need a solution to best manage stakeholder requests? Or tools for managing their support documentation, or training end users? The best data teams often find themselves struggling with this part of their workflow, and end up adopting solutions built for others (in this instance, product and support teams).

Since most data work and interactions are internal, it can be tough for teams to find the right way to work with business stakeholders without creating confusion and encountering the awkwardness.

Reducing Information Asymmetries

If you investigate the collaboration problems between data teams and others, you are bound to find information asymmetries between builders and consumers of data assets.  On the one hand, you have data builders with deep knowledge about the underlying data, how to manipulate and analyze it, and how to contextualize it within a larger body of data assets. On the other hand, you have data consumers, who are typically domain experts with rich knowledge about the business itself, which can be critical to providing broader context, understanding the data, and evolving the data platform.

Take Jane, for example. She just joined a Fortune 500 company as a sales manager, managing a distributed team of 15 salespeople spread out across the southeast. On the second day of her new job, she is forwarded an email from a colleague with several links to various resources: a spreadsheet with pipeline information, various reports in Salesforce, and a handful of dashboards about individual performance in the company BI solution. After spending a few minutes looking at the data, she realizes that she has no idea what she is really looking at, and what it means. She sends a message to her sales ops manager asking for help, who loops in their partner on the data team who built most of those resources. The data analyst reads the email, sighs, and then spends the next hour writing out a reply. They create a ticket on their JIRA board to “re-evaluate documentation.”

The root cause behind these kinds of data collaboration issues are information asymmetries between builders and consumers, which leave everyone frustrated and unhappy.

Tragically, the folks who are most often impacted by these dynamics are junior employees or middle management on the front lines, because they typically have less power in the organization and the least context for understanding the decisions being made around the data. Without intensive training, these employees are vulnerable to types of communication problems that result from information asymmetries. They are also prone to fall victim to “squeaky wheel syndrome,” where the executives and senior leadership team members’ voices are naturally heard the loudest by data teams (and therefore their requests and needs are prioritized over those of others.)

4 Proactive Tips for Better Collaboration

In order to get a better return on investment from the massive investments made in data tooling and teams, we need to attack these information asymmetries at the heart of our problems. Getting to zero is perhaps an aspirational goal, but data teams should continually strive to close this gap through practices, partnerships and tooling. Doing so will remove friction, increase transparency and trust, and allow everyone to get more out of the company’s data offerings.

Here are 4 proactive tips for data leaders who want to reduce information asymmetries and achieve better collaboration in their organizations:

1. Realign organizational and team structures with the needs of the business. This includes not only reporting models, but also data team roles and functions. We are already starting to see more job postings for roles like “data product manager” or “data scrum master.” These new functions will help data teams manage collaboration challenges which, at the end of the day, are usually about people and processes versus underlying technology problems.
2. Consider investing in a matrixed model where members of your team – or in some cases entire pods – are aligned to specific business units. This will allow alignment of longer term data initiatives to immediate business needs, foster knowledge sharing, as well as closer, collaborative relationships between analysts and those who they support day to day.
3. Start small, and build on your success as you go.  The power of first impressions cannot be overestimated. Initial perceptions of the data team are incredibly important to how their work will be received, so be thoughtful about how that goes with key team members up-front. Focus by building strong relationships with 1-2 key champions in the organization who can help spread the word about how amazing you are. Expand from there.
4. Be mindful of which collaboration tools can be leveraged across the lifecycle of your data initiatives and data products. For example, think about how you want to rally your people, process and systems for each of the below categories. Often what will work for one category will fail miserably in others:
   • Collaboration within the data team
   • Generic collaboration with other employees outside of your team
   • Ad hoc questions, or new feature requests
   • Ongoing support for data products
   • Scoping of new data initiatives or data products
   • Evolving your data offering based on what is valuable to the business

Conclusion

Innovative data teams are already migrating to software engineering best practices and that trend is likely to continue in the coming years.  As you look at investing in data infrastructure to support future growth, think about tools that support business partner collaboration.

 
 
Nicholas Freund is a seasoned SaaS industry executive with over a decade of experience leading startups focused on product-led growth. As Founder and CEO of Workstream.io, Nick spearheads a seed-stage technology startup that helps data teams manage critical data assets. Prior to Workstream, Nick served as VP of Operations for BetterCloud, an independent software vendor that offers the leading SaaS Operations Management solution. Previously, Nick held senior finance positions at Tesla, while earning his MBA at Harvard.