Despite the possible first impression of a superficial observer, car racing is a team sport and data analytics plays a pivotal role in keeping the driver on the winning path. Read on to learn how data-driven companies can overtake the competition.
Lewis Hamilton won his seventh Formula One drivers' title in 2020. He has been described as the best driver of his generation, and one of the greatest Formula One drivers ever. While watching the racing-show, we focus on the symbiosis of the car and the driver. He has a seemingly infinite tolerance for pressure and natural bravery to identify the limits of his car. We hold our breath in excitement when Hamilton executes precision manoeuvres to overtake the opponent.
His eyes are laser-focused on the track and the next turn to take, dosing the right pressure on the break and the right amount of fuel in the engine. Everything else is taken care of by the engineering team who follows every aspect of the car using the telemetry system. The telemetry system monitors over 120 different sensors scattered throughout the car, that generate data about every aspect of the race at a rate of over 10 MB per second
The data is then beamed to the garage via antennae in the front of the car, and from there it can be sent back to a team’s headquarters for further analysis.
Engineers continuously monitor the behaviour of the suspension, the wear of the tires, and the overall efficiency of the engine. The team assesses data points, makes predictions, and guides the pilot to apply the right settings before every turn and sprint. The data can also be shown on a driver’s steering wheel, helping that driver make decisions of their own.
Just like the Formula 1 driver, forward-looking executives also need real-time data about the trajectory of their business, so that they can focus most of their attention on the strategic angle, while autonomous teams are taking care of the thousands of operational aspects of the business. Leaders don't just look at one dashboard behind their wheel. They understand, that's too little and often too late.
Today, way too many companies collect way too little data, way too late to effectively steer the organization. But, what is enough data? When is it delivered fast enough? - you could ask with slight wonderment.
Imagine, observing the Profit and Loss (P&L) sheet at the end of the quarter and spotting a declining trend around a critical product or service. You decide to look at the bottom of this story promptly: you ask the sales teams to reach out to churned customers. Two weeks later you receive a report: the new features released at the end of last year in a hurry to cover some regulatory changes, caused annoyances among the users due to stability and usability issues. Somehow those improvement-requests never got prioritized by the product team. You call them and make sure that the necessary changes are prioritized on the roadmap. Now you just need to wait a few more weeks, until they release the new version and everything is expected to be back to normal. Victory!
But wait a moment! - wasn't there a better way? Couldn't you avoid the loss of business throughout these several months? Wasn't this information available at the customer support department earlier? - you're thinking out loud in the solitude of your office after the euphoria settles. You ask Customer Support to look into the matter and you learn - indeed - there were many complaints around the usability issue. Customers were perseveringly asking for changes months before they started to churn. Customer Support needed to read into several support cases to provide this insight because currently they only track response times, resolution times, and customer satisfaction around the support calls. Nothing about the satisfaction with the products. That was not a metric they thought to track.
So, what's next? How can you assure this situation won't repeat itself? And how could you detect earlier - like way earlier - if you need to intervene?
You ask the analytics team to build a dashboard with a customer satisfaction score for each product line. They run sentiment analysis based on the transcription of customer calls. It is made relatively easy, using today's embeddable cloud components, that extract relevant data points from text or audio on your behalf. The analytics team goes one step beyond and proposes to set some alerts in case these metrics fell below certain thresholds. Victory! - for the second time in the past few days. From now on, the product team will receive an alert every time customers are signaling dissatisfaction with one or the other product features. They will have the chance to fix the issue before they lead to loss of business and become visible in the P&L sheet.
You smile, - just like the Formula 1 driver - you can focus on the track and the next turn to take. You now have the ease of mind, knowing that the product team is continuously monitoring and fine-tuning the engine of your business and will autonomously take corrective actions whenever needed.
What we just described here, is called real-time, actionable insights at the point of impact. Let's spend a moment and unpack that sentence and build a deeper understanding on the impact:
it is actionable because the product team knows what to do when the sentiment drops below a certain threshold;
it is insightful because goes beyond the data-point (the complaint) or the information (the calculated sentiment), and compares aggregated metrics to a baseline, showing the data in its context, as a trend;
and finally, this insight is transported from its origin (the customer support team) to the point of impact (the product team), where it leads to course correction.
If this data would remain at the customer support level it wouldn't be all that useful to the organization as a whole. And finally, this report now is updated continuously and automatically, not just on the quarterly level, when it would be too little, too late.
Aggregated dashboards helped business leaders for decades to set or correct the course of business and make sure they are on track. 30 years ago it was OK to drive the business largely based on wisdom, gut feeling, and financial reporting. But as much the race car industry changed its face and pace, businesses increased its complexity and level of sophistication.
Just like Formula One is becoming more of a team sport, more and more businesses are delegating the decision-making power to the cross-functional, self-sufficient, autonomous teams. Senior leadership is there for alignment on the vision and expected outcomes, while teams build their web of metrics as a guiding star through the sea of decisions.
That is why democratization of data is the most frequent, high-impact initiative of most successful digital-native companies. Democratization means that data collected from various sources is made accessible to everyone while maintaining the data privacy requirements. This, of course, requires a certain level of organizational maturity and many companies struggle to establish the desired baseline.
It’s worth the effort though: insights-driven companies are predicted to grow at an average of more than 30% every year. According to Forrester Research, such companies will be growing 8-10 times faster than their non-insights-driven rivals through 2021“.
How does an Insight Driven company look like?
Data democratization - the initiative to provide access to everything to everybody - is just one side of the coin. Employees should be able to extract actionable insight and be incentivized to use that data as a basis for their decisions. They should go one step beyond anecdotes and impressions when they seek to improve business models and find new opportunities.
When you measure the maturity of the organisation, the first question you can raise is: how many of your team members are accessing last-minute business information before making plans for the day?
Or in other words: how many of them make data-informed business decisions? Be it the account manager who has to decide which customers to prioritize that day, or the marketing manager who will have to choose the top 3 channels to spend a big part of their budget, or the inventory manager who has to order merchandise, or the category manager who has to set a price. Informed decisions require new, underpinning insights every day. And this leads us to the second quality criteria: what percentage of decisions at every level of the organization are based on last-minute insights?
The following flywheel provides a glimpse on how companies can embark on a journey of continuous incremental growth by undergoing a data driven transformation.
Implementing a data-driven decision-making culture not only helps you to delegate high-velocity decisions to those who have most of the context, but better prepares the organisation to drive through deep uncertainty in times of crisis and maintain an appetite for experimentation in times of abundance.
This article was first published on Linkedin and it part of a series covering various aspects of scaling with data products.