Previously, I wrote about my transformative experience in building up the global data business for SAP, sharing my personal views on how to bring a startup to laser-focused execution and relentlessly drive customer success.
Data and technology have helped businesses evolve so rapidly that in just the past year, innovative technologies and capabilities such as data science, artificial intelligence, data analytics, and Big Data processing have allowed businesses to benefit from data in ways unimaginable before.
New, innovative technologies have allowed businesses to benefit from data in ways unimaginable before
“Go break through with your data” is my advice to every business. Why? Because data powers competitiveness and growth across industries and business processes. Transactional data augmented with additional data sources generate what I call “new intelligence.”
Take the retail industry for example. Retailers today can connect customers’ e-commerce records with social media activities and hyperlocal data to get the pulse of the neighborhood. Couple this with in-store, point-of-sale data as well as behavioral data from loyalty programs, we end up with an unbeatable, personalized consumer experience.
Now, if you infused the newly gained consumer insights into intelligent assortment and supply chain processes, you would have a smart supply chain that constantly ensures the proper merchandise is in stock. Such a closed-loop value chain, which is driven by edge intelligence on consumer behavior and trend patterns, redefines the rulebook.
The Why: Love the Problem, Not the Technology
The possibilities of technology are incredibly exciting; however, technology is always a means, never an end destination. Lead with strategic direction and be clear on what you are targeting with data-driven innovations.
Ask the following questions: What is the purpose of your digital transformation? Which growth opportunities can be captured through data and will lead your business into the future? Are you looking to gain competitive differentiation or pioneer digital data-driven services to establish a new source of revenue? From supply chain to manufacturing to serving your customer, how much friction is due to dispersed systems that aren’t harmonized and semantically integrated?
The How: Team and Customer Focus
The main challenge for data monetization is building a solid foundation for running a new, scalable and profitable data business. You overcome it by focusing on creating horizontally connected teams and putting customers first. Install small teams at the edge of the organization with a clear vision and strategy that aim to drive rapid business outcomes. Ensure that they’re horizontally connected to include a mix of talented outsiders and forward-thinking corporate veterans, who help keep the bonds with the core business and break down silos between core and edge teams.
In addition, ensure that customer experience is at the forefront of your business. The customer experience must be simple and frictionless despite the underlying complexity, where technology is invisible. We have developed a customer-centric approach using design thinking, a practice that focuses on feedback loops and iterative prototypes.
The What: Execution Strategy
Data-driven innovation does not happen by accident, it requires strategic focus and operational perseverance. Developing data applications is about building smart applications that focus on solving specific business problems, using new data insights to make predictions and prescribe and execute actions. Here are my four suggestions for building data-driven products.
- Working with data – fairness matters: Reliable and high-quality data is a must-have. When enriched with other data sources and processed using with machine learning algorithms, the value of data increases immensely. Go for the infinite possibilities but temper with a dose of reality. Why? Not all data is created equal and being aware of bias in the data is crucial for understanding the extent of a model’s accuracy and avoiding shortcomings on algorithms and biased models.
- Think transparency, no black box: Ensure model transparency and that algorithms remain accurate over time, meaning the reasoning behind a decision or recommendation can be explained at any time.
- Build in circular learning loops: Data-driven applications operate on diverse data sources, combining analytical and operational data to predict and prescribe what to do next by providing transparency and relevant insights in the form of industry benchmarks and peer-to-peer comparisons. Simulation capabilities and recommended actions based on algorithms guide the user in the decision-making process. Incorporate user-generated behavioral data and improve recommendations by correlating interactions back to the source of data to close the loop.
- Combine data and domain expertise: This is crucial because data-driven applications solve new problems in new ways by connecting the right data in a meaningful and contextualized way. The creation of value becomes much more problem-solving focused, providing solution for industries and for horizontal business functions like human resources, sales or finance. Use your domain expertise to find what problem to solve and let the data lead to new insights and signals. The best algorithms are the ones based on contextualized domain knowledge on processes in combination with data and behavior.
Get your data to drive breakthrough business outcomes. Stay tuned for my next piece on discovering how data transforms the bimodal world of IT.
Helen Arnold is president of the SAP Data Network.
Follow Helen on Twitter: @arnold_ih, #SAPDataNetwork.