Companies are forging ahead with digital transformation at an unprecedented rate, making it a top priority of enterprise businesses. According to a recent survey by IDC, spending on digital transformation practices, products and organizations will continue at a solid pace despite the challenges presented by the COVID-19 pandemic.
In a survey of Fortune 500 CIOs regarding 2021 budget priorities, more than 77% of CIOs recognized digital transformation as their top budget priority going forward, with more than 65% anticipating increases in ROI of 10% to 20%.
Companies have huge expectations for digital transformation, despite research that shows up to 70% of all digital transformation initiatives do not reach their goals. What causes digital transformation efforts to fail? Digital transformation is complex and challenging for organizations of all sizes at the technology, culture and corporate governance levels.
Industry information for this eWEEK Data Points article is provided by Ryohei Fujimaki, Ph.D., founder and CEO of dotData, a data analytics software maker. Here are five common mistakes that leaders can avoid to beat the odds.
Common Mistake No. 1: Not having the right digital-savvy leader in charge
Having the right people in critical roles, including senior leaders of the organization, increases the chance of transformation. A key trend is the growing importance of the Chief Data Officer (CDO) role, a person who can become the change agent responsible for data-driven transformation. Recent research from Deloitte across 20 diverse companies that recently underwent a digital transformation found that, regardless of industry, organizations that appointed empowered CDOs—backing them with strong mandates and executive support—exhibited significant operational improvement.
Common Mistake No. 2: Data infrastructure, data maturity and misalignment between strategy and technology
Business dynamics are constantly in a state of flux, driven by market forces such as regulations, supplier, customer or competitive pressure. But when business strategy changes, so must your organization’s technology strategy. IT budgets and projects should be aligned with the overall business strategy to support new initiatives, and this is especially true of digital transformation initiatives.
For a successful digital transformation effort, the right data infrastructure needs to be in place. For AI and ML initiatives supporting digital transformation, IT needs to evolve and have a heterogeneous data processing architecture (GPUs, ASICS, core processors) in place to support AI solutions. Equally important is legacy IT infrastructure and how to integrate it into the new digital architecture. Too many organizations collect data only to ignore it in the end or analyze only a fraction of the data. The problem may not be in the algorithms or the AI platform. The fundamental problem is often the lack of a scalable data infrastructure architected for end-to-end data flow, the organization’s data maturity and the inability to make data accessible when needed.
Common Mistake No. 3: Pursuing the wrong projects
Often, disagreement among stakeholders about digital transformation projects and priorities can lead to selecting the wrong projects to pursue. If the business is new to AI and ML, starting with bold, ambitious projects may not be ideal. One of the biggest challenges of data science projects is the upfront effort required despite a lack of visibility into the value. The traditional data science process takes months to complete until the outcome can be evaluated. Enterprises may end up spending three to six months collecting data only to realize that they don’t have the right data for the use case.
The long turnaround time and substantial efforts associated with this approach often result in project failure after months of investment. It is far more prudent to select low-hanging fruits, small projects that can create data-savvy teams and make them ready for AI. This approach will instill confidence and encourage others to follow.
Common Mistake No. 4: Building the entire tech stack inside
The classic buy-versus-build problem is common among medium and large enterprises. Often companies believe that their requirements are so unique and specific that they are best off building their entire technology stack in-house–not taking into consideration the costs, resources and timelines. Another option is to use automated machine learning (AutoML) tools and platforms designed to integrate seamlessly into AI and ML processes. There are many benefits to using AutoML tools. AutoML platforms can make AI more accessible to everyone by automating complex manual data science processes, enabling companies with small data science teams to maximize AI efforts.
There are a wide range of platforms with various capabilities, but next-generation platforms–aka AutoML 2.0–provide the most comprehensive support for AL and ML initiatives. AutoML 2.0 platforms can support data transformation efforts by automating the full data science process–from data preparation, feature engineering to building and deploying models in production. With a few exceptions, an AI automation platform will almost always be faster, efficient and deliver a better ROI for analytics initiatives.
Common Mistake No. 5: Underestimating cultural challenges
Many transformation experts have stated that digital transformation is less about technology and more about the people and culture is at its center. For digital transformation to succeed, digital transformation leaders need to be intimately familiar with the company culture, beliefs and how to connect the shift to the core purpose. The transformation has to align with the culture, address roadblocks and eliminate challenges. Without the support and buy-in from the trenches and various organization levels, the change will not succeed. Clear goals, constant communication from the digital champion and the executive leadership team can help.
If you have a suggestion for an eWEEK Data Points article, email [email protected].