<img height="1" width="1" style="display:none;" alt="" src="https://dc.ads.linkedin.com/collect/?pid=306561&amp;fmt=gif">

How to Challenge Business Assumptions with Data

Before data and analytics leaders can use data to change the way their organizations are run, they must first overcome these three common obstacles

Written by Solomon Radley

How to Challenge Business Assumptions with Data

Written by Solomon Radley on Nov 7, 2019, 2:55:56 PM

CDAO UK 2020

The notion that businesses should be making data-driven decisions may seem obvious to a CDAO. But the fact is, people in other business functions don’t always see things that way.

“I still see a lot of business assumptions in our daily processes,” says Ricardo Rodrigues, Global Pricing and Operations Manager at Vauxhall. “Someone yesterday was talking to me about ‘gut feeling’. It still happens today. And honestly, I’m starting to have a hard time to understand that.”

There are three ways attempting to use data to inform business decisions can go awry. The data science or analysis may give the wrong answer. The business may ignore the work’s implications. Or, it may take so long to approve a fresh course of action that the original insight is no longer useful.

To effectively challenge outdated business assumptions with data, data and analytics leaders need clear processes in place that empower them to avoid all three of these potential pitfalls.

The Step Between Data Collection and Analysis

Businesses are now waking up to the fact that data can help them discover better ways of working. But many executives are unclear about what ‘doing data’ really entails.

“What seems to be happening a lot of times is that companies are moving from collecting the data to basically putting the data in one place (or not) and analyzing the data,” says Tomas Sanchez, Chief Data Architect at the ONS. “So, they move from data collection to data analysis.”

The problem with this approach is that it skips a crucial step. In order to use a complex data ecosystem efficiently, organizations need an underlying infrastructure that’s common to all their data systems.

“There is a Forbes article basically saying that 60% of a data scientist’s time is used for cleaning and transforming data into something they can use,” Sanchez says. “This is a consequence of not having this middle layer.”

“There needs to be a paradigm shift of a thinking shift in these organizations,” he concludes. “They can’t really do data seriously unless they have some sort of data management in place.”

Data Programs Need Executive Sponsorship

It’s no using data to find opportunities for an organization to run more efficiently if no one acts on those insights – and that’s where ‘executive sponsorship’ comes into play.

Rodrigues illustrates this idea with a story about one of his early successes at Vauxhall. Historically, the business had assumed that cheap stock moved fastest. But his team’s analysis showed there was more demand for higher priced vehicles than previously thought.

“We took this to our fulfilment department to show them,” he recalls. “They didn’t accept our conclusions at all. Lucky for me, my director is really keen on analytics and he’s very respected.”

With his director’s backing, Rodrigues ran a test with the company’s next batch of orders to verify his theory. Half of the stock was ordered in the conventional way. But the rest was chosen using statistical analysis to determine which vehicle specs to order in each region.

“Guess what? All the vehicles we specified did move faster!” – Ricardo Rodrigues, Global Pricing and Operations Manager, Vauxhall

“The vehicles we specified did move faster,” he says. “Not only that, but since we uplifted the price because we knew there was an opportunity for a higher price, we generated more profits. If you want a number, we’re talking about several dozens of millions.”

This example shows how much easier it is for data leaders to drive change when they have buy-in from the top.

It also demonstrates the importance of prioritizing initiatives that will have a measurable impact on the company bottom line. Early successes like these are crucial for galvanizing support for data-driven decision making at all levels of an organization.

Becoming Data-Driven Requires Decentralization

It’s one thing to make a single data-driven change with help from the CEO. Creating an organization that instinctively uses data when making decisions over the long-term is quite another.

Data leaders can’t count on their bosses to intervene every single time they need to enforce their initiatives. Instead, they must push for structural changes to individual business units that will empower them to make their own data-driven decisions.

“The thing that decision making really hangs on in terms of value is the lead time,” explains Marcus Bitterlich, Director Business Intelligence at technology manufacturer Rohde & Schwarz.

“Escalating decisions can easily decrease the quality of these decisions, due to a lack of context,” he adds. “The guys at the top are not exposed as closely to the market forces as the guys on the street.”

The data profession’s shift towards agile ways of working provides one possible solution to this challenge. Creating product-focused teams that can steer themselves and are accountable for their actions will enable staff to take ownership of their actions.

“Provide them with decision rights and coach them towards this ‘adult’ way of behaving,” Bitterlich recommends. “The more levels of authority to you need to climb up to get somebody to make a decision, this increases lead time and therefore decreases value.”

“I cannot say simply from day one, ‘Employee A, now you’ve got to decide which components to buy’,” he continues. “You need to actually redistribute decision rights to teams.”

“The thing is, at large organizations, still employees are
treated like children” – Marcus Bitterlich, Director Business Intelligence, Rohde & Schwarz

It takes guts to make this kind of change. But it can be done. Once employees have bought into the idea of being data-driven and have the tools to do it for themselves, all they need is the freedom and confidence to go out and make those decisions.

“Once you establish psychological safety for the team, so that it’s actually not blamed after taking a decision but just held accountable,” Bitterlich concludes. “Then, you can easily decentralize decision making.”


This article is an extract from Transformational Data Strategy UK & Europe. For more exclusive insights from the UK and Europe's top data leaders, click the image below now and download the full report.

Transformational Data Strategy UK & Europe

 

Related posts