Actionable summary of this article if you have only a few seconds to spare
- Operations and data teams frequently interact with each other, but their collaboration could be often more productive.
- In our 200+ interviews with executives and managers in these teams, we heard many pain points from both sides.
- These difficulties are rooted in three major structural differences between how operations and data teams operate: their distinct nature of work, day-to-day rhythm and incentives.
- Both sides need to take steps to achieve successful collaboration. Solutions among others include self-service analytics, guided processes for collaboration or nested data analysts in business teams.
Thanks to digitalisation companies collect more and more data about their operations. Business teams gladly utilise this information in their day-to-day work. Especially operations teams like to rely on data as they have to optimise cost and customer experience on a daily basis. Given the data-driven nature of this field, operations and data teams frequently interact with each other. After conducting 200+ interviews with executives and managers from such teams, we found that collaboration between them is often far from flawless.
Pain points that we see on the surface
The most frequent difficulties that we heard from operations teams were the following:
- They do not have access to the relevant data.
- When they request data from the data team, they do not get it in a timely manner.
- The received output is often not what they would have expected, therefore additional follow-ups are required.
- When they start working with the data, they figure out that the definition does not exactly match with what they had in mind.
- As such, they cannot react to urgent business issues quickly enough.
When it comes to data teams, the most common challenges mentioned were the next ones:
- They don't receive clear requests from the operations teams, therefore they frequently need to go back to clarify the asks.
- They don't receive the necessary business context that would be required to prepare the relevant datasets.
- They are often contacted by simple modification requests, because of which they feel as a support function.
- Many of the reports or dashboards are never used by anyone.
- As such, they can never catch up with their tasks and satisfy the business needs, and at the same time, they never have capacity for important projects.
Root causes behind symptoms
Hearing these struggles from meeting to meeting we asked ourselves the question: what can be the underlying factors that drive these pain points? We believe all of the above mentioned symptoms are rooted in three major structural differences between operations and data teams.
First of all, we have to acknowledge that the two fields simply have disperse nature of work. To illustrate the contrast, think about the qualities these professionals need to successfully deliver their job. While data analysts need time to think through complex data structures and formulas to come up with the right solution, operations managers need to deliver quick answers to urgent operational issues. Data teams cannot afford not delivering perfect outputs, because then the wrong conclusions may be taken, while operations teams often have to deliver 80/20 solutions quickly that are still better than not reacting to acute problems.
Secondly, as a result of the first distinction, these teams work based on a different rhythm in the day-to-day. Data teams often have quarterly OKRs that are split into (bi-)weekly sprints, while operations priorities frequently change on a daily basis. Because of the fast paced and experimental nature of running operations, new data points are needed constantly that data teams struggle to deliver at the right speed.
Last but not least, these teams may also receive different objectives from the management. Data teams most commonly report to the CTO and solve for security, stability & accuracy. On the other hand, operations teams usually sit under the COO or other business-driven CxO and therefore need to optimise for speed and cost.
In order to improve the collaboration between data and operations teams, we believe companies need to respond to the underlying, structural discrepancies, instead of tackling the symptoms. We also reckon that both sides need to take steps to achieve productive teamwork for which we recommend companies take the following actions:
- Self-service data consumption. Operations teams should be empowered to self-serve their analytics needs as much as possible. They should be able to build dashboards and reports for themselves, and set up monitors and alerts flexibly to respond to their actual needs. Data teams should focus on making higher quality and amount of data available in these self-service systems, instead of responding to repetitive support requests.
- Guided process for collaboration. Even if companies have built self-service analytics systems, operations teams still need to interact with data teams to request new data tables or data views. We suggest that this process is guided with forms or other wizard flows to ensure that both operations and data teams think about all the details that are required to start the data generation process. With that teams can minimise the back and forth and avoid the most common issues such as misalignment on data granularity, format or definition.
- Involvement of data teams in the day-to-day business. Even the most well-designed processes cannot replace the impact of continuously involving and informing the data teams of the actual business priorities. This doesn't only give them the necessary business context to prepare the relevant datasets, but also increases their motivation to put themselves into the operations teams shoes. In more extreme cases it may even make sense to move the analytics team under the COO to avoid misalignments and conflicts of interests.
- Nested data analysts. More and more companies ensure involvement and alignment by dedicating data analysts to specific business functions, including to operations. In such cases the data analysts do not only have a better chance of understanding the respective business area, but that also provides a dedicated capacity to that function, which in result accelerates the speed of response.
- Upgrade operations teams with SQL skills. Operations managers are usually number savvy and analytical by nature, therefore it can also be a great idea to offer them development opportunities to learn basic SQL skills. That empowers them to process simple data manipulations so that they wouldn't be dependent of data teams in all the cases.
If you struggle with similar pain points - be it either on the operations or data side - please get in touch with us via firstname.lastname@example.org or by requesting a demo, we would like to learn more! We can help you!