Overcoming Barriers to Data-Driven Transformation
Data-driven transformation is increasingly a question of life or death across many industries, with the trend accelerated by the wide-ranging impact of the pandemic. At a recent digital roundtable discussion hosted by CDOTrends and supported by Snowflake, executives and data science leaders came together to discuss the data-driven future and share their strategies for successful data transformation.
The changing face of data
Common challenges highlighted by participants include accessing data from disparate legacy systems, and technical considerations such as how to properly clean up incoming data and ensure that its quality is up to scratch. Then there is the issue of implementing proper data governance, no small feat in the face of evolving regulations and a fluid cybersecurity landscape.
To effectively utilize data, there is a new way of operating and a shift of mindset that is needed, says the Director of Digital Transformation at a telecommunications firm. This might entail nudging people who had always done things in a certain way towards a new data-centric approach, he explained.
Security is one area that her organization pays a lot of attention to, says the VP of Global Practice lead for supply chain and analytics at a global logistics firm. “We put a lot of energy and resources to ensure the security,” she said, noting that her CEO had publicly stated that cybersecurity is his number one priority.
Though inevitable given the rising cybersecurity incidents globally, it is another consideration that the team must learn to work with. She said: “From a data science standpoint, our data scientists love the use of data. I think that is sometimes a little bit of a balance between how we ensure the security of our systems and data versus the agility of how our data scientists can access and use the right tools and information to explore new ideas.”
Start small, think big
Of course, the deep impact of the ongoing pandemic means that models and predictions based on certain historical data are less useful than they should be. According to the VP of the global logistics firm, this has culminated in more than one conversation where data scientists must determine if findings based on “pre-COVID” data points are relevant in a “post-COVID” world.
On the other hand, there is no question that the pandemic has increased interest in analytics and data science. The number of requests coming in has certainly gone up, noted a Regional Head of Data Analytics: “And even as we continue to reach out, I feel there's more who are willing to listen to us than before COVID. It's unfortunate that it had to be a pandemic that opened up a few of these opportunities, but I suspect it's the same in other industries.”
One strategy highlighted by more than one participant revolves around starting small by focusing on low-hanging fruits to attract the attention of stakeholders and executives. This could be as simple as starting with a business analytics software or implementing a rudimentary dashboard to better evaluate the opportunities revealed by the data.
“If I go directly to deep learning algorithms in conversations with executives… I lose them”, observed the Regional Head of Data Analytics. “There are very quick wins that we can get. And that sets the right agenda moving forward for driving the agenda towards data-driven decision making or data-driven transformation.”
Momentum for bigger things
Ultimately, long-term success can only come from a sustained, concerted effort by the data science team and leaders. As noted by the Head of Commercial IT and Digital at a pharmaceutical firm, the correlation between data insights and funding isn’t always evident: “It takes time to build the data-centric capabilities in the company – they don't just appear out of nowhere. [You need] the funding to find the right people, to hire people for the future.”
Separately, the Director of the telecommunications firm noted that his organization has already utilized data to generate more insights and leads, as well as helping to reduce cost through better productivity. And by leveraging machine learning (ML), his team had developed models to quickly process trouble tickets and reduce the workload by 50%.
For the VP of the global logistics firm, success comes with both a bottom-up and top-down approach. “We have small [teams] adopting a bottoms-up approach all over the world that mostly look at operational opportunities. At a global level, we have also created a central competency center of a group of 40 to 50 data scientists to work on Big Data and data lake initiatives to more thoroughly [evaluate the data],” she said.
“In Asia, we created a central team that looks at more customer-centric initiatives to use as a hook for the business. Customers appreciate the value that we offer. This customer-centric approach, looking at anything that create value for our customers has helped us create that momentum for bigger things.”
Image credit: iStockphoto/AndreyPopov
