COGNITIVE WORLD

View Original

Winning with AI – Overcoming Cultural Challenges

Depositphotos, enhanced by CogWorld

A recent survey on Big Data and Artificial Intelligence (AI) reported that cultural challenges, not technological ones, were the biggest hurdle to overcome around Big Data and AI initiatives. According to this 2021 survey, the vast majority of respondents — 92% of mainstream companies — continued to struggle more with cultural challenges than with technological ones.

This survey was conducted by New Vantage Partners and while their surveys have a narrow reach, involving only 85 Fortune 1000 and industry-leading firms in 2021 – the results merit attention due to the high quality of respondents. In 2021, 76% of respondents held the role of Chief Data Officer or Chief Analytics Officer at blue-chip, household name companies in financial services, life sciences, healthcare, and retailing. 

The challenges with organizational alignment, business processes, change management, communication, and people skill sets in implementing AI is clear, yet how to overcome them is less obvious. This article sheds some light on what companies might do to leap beyond pilots and small proof of concept endeavors to significant deployments of AI.

Organizational Alignment

Organizational alignment is elusive in the absence of a clear purpose and a shared understanding of strategy. The sad fact is that most companies do not integrate AI into their overall business strategy. There’s a mindset problem. Instead of asking how AI can enable our strategic objectives, executives ask the far less useful question of where is there a use case to apply AI? This is often due to powerful departmental silos that stand in the way of big picture thinking. 

To overcome organizational alignment challenges firms need to begin by looking at their business from the outside-in or the customers’ point of view. Then the performance-oriented firm will focus on building a shared understanding around its purpose and a few crystal-clear strategic objectives which AI can enable. It doesn’t much matter what strategy formulation framework is deployed, although Kaplan and Norton’s strategy map approach has the significant benefit of explicitly recognizing the importance and interdependence of customer experience and end-to-end processes. What’s important is to reimagine how customers are served and how products/services are developed, sold, made, and delivered. A simple extrapolation of last year’s plan will not suffice. 

Achieving alignment is difficult when strategic objectives are defined by departmental activity or by technology. Only when there is clarity on a bold strategy can leadership intelligently ask the right questions such that AI and other digital technologies enable performance.

Business Processes

Value is created via a firm’s large, cross-functional business processes. Not by its departments. 

A firm has very little chance of reaping sustainable benefit from AI if it does not examine how end-to-end processes create value for its customers. Sadly, as one executive recently described, large companies continue to be tempted to centralize and standardize by vertical silos and, as a result, sub-optimize the horizontal, processes via local optimization and the sub-optimization of end-to-end, value-added processes. Departmental silos stifle collaboration and perpetuate data silos. 

Instead of modeling small processes inside departmental boundaries, successful firms will focus on the critical few value-creating end-to-end processes where AI can enable improved performance. That means defining and assessing the performance of the critical few processes involved in serving customers and developing, selling, making, and delivering products and/or services.

Governance matters. Forming cross-functional teams to define, improve and manage the set of end-to-end business processes is a good first step. However, often it’s not enough as cross-functional teams typically do not have sufficient authority or budget to make the right things happen. Cross-functional executive steering teams are needed to overcome obstacles and provide the right governance. 

Change Management 

Many companies struggle in realizing AI’s full potential to transform the way business is done due to poor change management. They typically apply AI to a long list of discrete uses, often in the context of department priorities. This approach makes it difficult to scale. For optimal results, change management should be built into AI deployment. From the get-go, not as an afterthought.

When change management is built into AI deployment, C-level executives state what will be done, why it’s important, and ask the entire organization to pitch in and help. They do this early and several times using various media. Then change management specialists are enrolled as part of the AI deployment team to conduct stakeholder analysis, develop communication plans and work closely with project managers.

By building in change management to the deployment of AI into an entire end-to-end process, organizations can realize major improvements in performance that isolated applications to one-off use cases simply cannot match.   

Skill Sets

In many companies, the availability of technical talent is one of the biggest bottlenecks for scaling AI. Becoming adept at deploying AI takes capacity and skill. Although AI adoption has grown exponentially, many companies are finding that there is a shortage of high-quality resources with AI knowledge. Acquiring and retaining skilled talent is a critical success factor in AI deployment. That means finding and keeping people who are skilled in data science, and user experience, such as data scientists, and engineers proficient in data, machine learning, and cloud. While even the best talent does not guarantee success, a lack of it almost guarantees failure. 

To overcome the challenges of finding and keeping people with the right skills first align and integrate business strategy with AI and machine learning efforts. This will create the right environment and change how people feel and work. Then, focus on having a full portfolio of capabilities and skills in AI, whether that’s achieved by building the talent teams in-house or adding the capacity from external sources. That involves assembling the right team of top-notch technology, data, and process people — along with a strong focus on change management. With the right people, organizations can concentrate on the end-to-end processes and governance that will fuel the AI-driven digital transformations.

Summary

Artificial intelligence is at the heart of digital transformation. To deploy AI at scale it’s essential to overcome challenges with organizational alignment, business processes, change management, communication, and people skillsets. These are not new challenges. The lack of organizational alignment and failure to pay attention to people plagued reengineering in the 1990s. Back then and even to this day executive mindsets needs to change in two ways. First, by looking at what matters to customers as well as what matters to the company. Next, by focusing more on value-creating end-to-end processes than on department responsibility. 

While companies do struggle more with cultural challenges than with technological ones, technology matters enormously. Simple automation is not enough. Digitization is not digital transformation. A model or algorithm which is not deployed — no matter how technically brilliant it might be — will not create value. The best results will be realized by those companies that integrate various technologies in reimaging how customers are served and value is created. Results matter.


Find Andrew Spanyi on LinkedIn. Check out his website.