Asking the right Questions
Working with Data Science teams requires a different mental model for business leaders. The tested language of Mission, Objectives and Outcomes, familiar to strategic planning as we know it , will not work for the data age. Days of discussing the plan and associated goals, tasks, measures and KPIs for hours with your team, will no longer bear fruit.
To find new sources of growth and innovation, leaders need to orient from "What do we need to DO" towards "What do we need to KNOW". The strategic intent or ambition must be clear for all to see, but act simply as a handrail for exploration.
A great example of this was from Carlos Ghosn at Nissan - he simply articluated Renew Nissan as the guide for what was significant strategic renewal in the company. In the data age, this will act as hand rail for discussion around multiple business challenges, but most importantly guide questions to data.
So, instead of populating that goal spread sheet or software tool, organisations need to build the skill in leaders to ask questions, in a simple and consistent way, so that all members of the team can explore, experiment and discover new opportunities through data.
There are some challenges that teams face here, and following our experience we'd like to share some thoughts to help you along the way.
Build situational understanding
Business performance is central to this. Clearly articulate what data will help the team increase understanding around our customers, clients or people.
EXAMPLE: What do we want to understand and how soon?
Then, clarify what each team member is trying to understand. For example, if we want to know what the future customer pain points are in a market through a marketing lens, we may also want to know how this could affect human capital needs from an HR perspective. Surfacing these data interdependencies is where the huge value lies for teams, breaking data silos from the very start.
EXAMPLE: What could affect the outcome of my data? Who needs to know this?
Develop a simple questioning loop for teams, and a simplified and consistent language for the organsiation, so that data scientists can quickly align to the questions and unkowns business leaders have, helping them to act in a more prescriptive way.
In other words, help the data science teams ensure that the right data, gets to the right team at the right time. Business leaders will thank data teams for this, as insights get pushed towards them to stay ahead of change and find growth opportunities at speed.
As teams start to adopt this way of thinking, the muscle will get stronger and become a ritual that teams perform daily, so leaders can jump into a data science project, get up to speed with whats happening and run to their next meeting.
Start with the business
Too often, data scientists start with the data, building models and insights that will hopefully and seamlessly align to the desired business outcome. Unfortunately, this rarely happens. The business team sets the direction by clearly articulating the business context.
Data Science is certainly exploratory, but we can start to build trust with leaders by given the journey a business handrail. When struggling with this, try asking some generic thought provoking questions against your strategy:
EXAMPLE: What do we think we know currently?
Clarify what is the customer group, market, product or service that you’re referring to. Understanding this context will help data science teams operate with clearly defined boundaries. Here, fostering autonomy and alignment in the data collection effort will unlock the real skill in data science team members. Let them go and exlpore within the business' strategic boundaries.
Authority in the team
To get the best results here we need to shift from HiPPOs (highest paid persons oppinion) to the TEAM. The phrase of we don’t know what we don’t know is relevant here. Anchoring experience of senior leaders is important, but a fresh pair of eyes looking at a problem can add huge value. Pull-in leaders from outside of your team or who are further from the detail and can challenge assumptions and competeing hypotheses as you build questions.
By clearly defining the context of the data question against the business performance objective at the start, leaders will be in a strong position to surface ROI opportunities later down the line. This is a huge win for senior analytics teams as they start to measure the quality of the business interaction that will act as a base line for deeper ROI efforts.
Finally, adopting a consistent approach to asking data a question and driving this end-to-end in an organisation, will unlock this great opportunity. It will break silos and enable data talent to share data across teams and diffuse learning throughout the ecosystem.
This is a thinking skill challenge. Not a technical one.