On Big-Data, AI and Digital-TransformationBIG Data

Mohak Shah, Ph.D., VP - AI and Machine Learning, LG ElectronicsMohak Shah, Ph.D., VP – AI and Machine Learning, LG Electronics

Over the past decade, advances in AI and digital technologies combined with big data have fundamentally changed the industry and societal landscape. These technologies have also spawned off entire new industries and introduced novel business models. They have also significantly sped-up the product and offering lifecycle thereby reducing barrier to entries in various domains. Further, commoditization of the building blocks is now allowing organizations and startups to rapidly put the needed software and hardware infrastructure in place to get the ideas off the ground. A swath of new players as well as transformation of existing players in various industries – be it advertising, shopping, ridesharing, autonomous driving, hardware, consumer electronics or lifestyle – is a testimony to this phenomenon. These technologies are exerting a significant pressure for the businesses to move up the value-chain by being more customer-centric and supporting services. The emphasis is on enabling seamlessness and uninterrupted execution of customers’ respective functions both in B2B and B2C contexts. This is nothing short of a new industrial revolution.

​ The core design principles can be a good basis to understand and establish the success and longevity of the initiatives  

The resulting market-pressures are forcing legacy companies to transform themselves to stay competitive. These companies are prioritizing transformative initiatives and targeting relevant goals – from optimizing customer experience, increasing efficiencies and productivity, and devising novel revenue opportunities and business models.

However, these organizations are still struggling to successfully and reproducibly translate the scale and extent of opportunities into tangible outcomes. A recent McKinsey study in fact found that most digital transformation efforts have failed. Analyzing these failures have identified some core gaps – from broader economic forces or market requirements all the way to challenges in chasing the moving target of a new promising business model. While these are pertinent, I believe there are additional important reasons why transformation efforts fail internally – treating transformation efforts as solely technological efforts disregarding the people and cultural aspects; resistance from legacy forces to new ideas and competence; lack of proper incentivization and benefit structures; and morphing of initiatives to align with traditional business.

Organizations must realize the importance of the aspects of digital transformation other than the technology. They need to identify the right transformation objectives keeping in mind both the growth goals and the evolving market landscape. Hence, a business vision is crucial in anchoring the transformation efforts. This also makes sure that the projects are not one-off efforts but contribute to over-arching coherent strategy arising out of this sound vision.

On the technology front, the transformation efforts need to be supported by a realistic data-strategy that focuses not just on internal efforts around data organization and management, but also covers the broader aspects of the ownership, acquisition, governance and persistence of data. The same goes with the analytics strategy that should focus on building an underlying basis that is not dependent on a single set of algorithmic choices but is adaptable as the technology evolves.

The core design principles can be a good basis to understand and establish the success and longevity of the initiatives. One of the typically under-appreciated aspects is the necessity of a repeatable development capability. A core differences between software businesses and legacy organizations is the establishment of an underlying software development platform that allows for rapid development, scaling, integration and adoption of the software products seamlessly. Note that these platforms are not homogeneous. The choice of the type of platform(s) depends on the transformation objectives.

Organizations need to pay special attention to enablement – on technological as well as organizational fronts. This goes beyond the software or data and analytics platform to organizational readiness to enable adoption and integration. Appropriate incentivization for different teams to contribute to the overall effort must also be devised. Traditional performance metrics for sub-organizations that are crucial to data acquisition and integration efforts of big-data and AI projects are often misaligned with the transformation objectives. This results in a reluctance (or inability) from these arms to appropriately invest in labor, software or digitization technologies to support the transformation effort. This misalignment of incentives results in risks for any well-intentioned transformation effort and can contribute to overall failure. Further, accounting for company culture and impact on people are extremely necessary when change management is put in place that aligns with these initiatives.

In conclusion, Big data and AI present significant growth opportunities for the industry. However, these can’t be realized unless the organizations are willing to evolve, adapt and be agile. They need to significant flatten the hierarchies, adopt a fast-cycle business model and build a solid technological, structural, organizational and most importantly diverse foundation – i.e., a long-term outlook. Maintaining a laissez-faire approach is a recipe for failure down the line.

social experiment by Livio Acerbo #greengroundit #thisisnotapost #thisisart