Creating a data-driven business at McKinsey & Co.

Mike Wright, Global CIO, is turning knowledge into value at the global management consulting firm.

Mike Wright, Global CIO, McKinsey & Co.
McKinsey & Co.

When your greatest asset is the knowledge of your people, how do you keep track of who knows what? Mike Wright, CIO of McKinsey & Company since 2013, uses a combination of artificial intelligence (AI), a simplified architecture, and a user-centric design to turn data into client-relevant value at one of the world’s leading global consulting firms.

I recently spoke with Wright about how he is creating a data-driven organization and, just as importantly, how he is changing the culture of the IT organization to support it. What follows is an edited transcript of that conversation.

Martha Heller: What does “digital” mean to McKinsey?

Mike Wright: We look at “digital” primarily from a business lens – i.e. as a new capability to allow us to do what we’ve been doing for decades: helping our clients make substantial lasting improvements while attracting and retaining outstanding people. While “digital” does not change our two-part mission, it does give us the opportunity to reimagine how we deliver it. For instance, we can eliminate and automate processes and we can use data to personalize the experience for our clients and our people.

How are you using data to improve the customer and employee experience?

We aspire to bring the best of McKinsey to every client. To do that, we need to take advantage of the knowledge that exists across our 17,000 consultants in 66 countries. Historically, we’ve had a central knowledge repository. Now we are asking: how do we expand that repository to include a much greater variety of formats, like video and audio files, code snippets, and data sets, with better curation and search? How do we allow our clients and employees to access validated external resources in addition to our internal knowledge?  As a large firm, we run the risk of losing our grip on that knowledge. Digital helps us counter these scale and geographic distance challenges, and ensure we remain personal to our people and relevant to our clients.

What is the technology you are using to provide that data?

The data is held in relational and graph databases with very good tagging and semantic search capabilities. We also use AI and machine learning techniques to bolster our cognitive capabilities. But it is not the technology that improves the client and employee experience, it’s the data. When you don’t get the answers you need, it is because we have not properly codified or tagged the data and we have not enabled the right search and retrieval capabilities.

How are you overcoming that data challenge?

We know that there are four keys to creating a data-driven organization:

  1. Hire the right people – namely, those who understand the value and relevance that data brings to our clients.
  2. Understand what data is most important to our customer experience.
  3. Use artificial intelligence, machine learning and other tools to augment the power of human curation to make it a really great experience. At McKinsey, we are encouraging people to take ownership of their data, so that the knowledge stays as close as possible to the actual expertise, but we are not there yet. The more we can pre-populate that information from other sources, the easier we can make it for people to refine it.
  4. Provide the right architecture to support a data-driven organization. We have moved away from the “high-rise” architecture where everything is in a massive ERP system, in favor of a “low-rise” architecture with APIs and add-ins to connect different data sources and third-party apps for specific capabilities.

How are you changing the culture in IT to support a data-driven organization?

As a leadership team, we want IT people to adopt user-centric design. We want them to ask “would I enjoy using this application? How can I help people adopt it?” Historically, technology groups have thought of their remit as deploying technology – i.e., If you deliver a system that runs well at scale, then you’ve done a good job. Our perspective is that our job is not done until we’ve got the right people in the right numbers adopting the new capability. That means a whole new mindset for people and it also means hiring the best talent to role model and reinforce those mindsets.

What advice do you have for CIOs who are driving change?

  1. “Lead from the side” by encouraging others to step forward. We’ve developed a digitization leadership group where cross-functional digital leaders learn from each other and share experiences. It is convened by our chief digital officer and members of IT are involved, but it is a business transformation group, not a technology group.
  2. Don’t wait. As Arnold Glasgow says, “The trouble with the future is that it usually arrives before we’re ready for it.” Technology leaders tend to like the precision of the ones and zeros, but we all need to get comfortable with rough edges and work-in-progress. For example, we recently upgraded our expenses application to provide better mobile capabilities for processing receipts. But it’s complex because expense rules vary by country and regions. With our focus on improving the employee experience, we wanted to roll out quickly, which meant not waiting to integrate all the existing capabilities (such as a fraud detection tool). Rather than wait, we agreed with our Finance function to build in functionality gradually. That is to say, we reduced the functionality (to start with) in favor of the user experience.
  3. Always simplify your architecture. It is hard to get funding for pure technology simplification that doesn’t have a direct business impact. Our approach is to have one eye on simplification while delivering new tools. We’ve identified 190 different legacy data stores and applications that need to be retired. So, we set a target of three years, with the goal of retiring a good percentage of them every year. For example, when we implemented a new staffing application, that was the time to retire five disparate tools related to staffing in one form or another. But our IT leadership team did not decide which to retire; we left those decisions to the relevant product managers who can gauge the business value much better. We just keep the score – i.e., How well are we tracking against the decommission list?
  4. Ruthlessly prioritize. Demand will always outstrip your capacity, so it is critical you make sure your teams are working on the most important outcomes. Involving your stakeholders in those tradeoff decisions also helps everyone feel joint accountability for the priorities and respective timelines.
  5. Be transparent. For many organizations, IT is a black box, which can lead to a lot of questions about what you are doing and why. You are constantly educating and re-educating your key stakeholders. If you can increase your transparency in terms of backlogs, priorities, OKRs, and financials, then your stakeholders can see that you are minding the store responsibly. That builds trust and collaboration. Most interestingly, this transparency also helps guide teams with day-to-day decisions that advance the goals of the customers, stakeholders, the function, and ultimately, the firm.
  6. Take symbolic actions. Four years ago, we renamed IT as the “Technology & Digital” group. Does the actual name really matter? Not really, but it symbolizes that digital does matter. That kind of symbolic action is important to supporting your message of change.

Copyright © 2020 IDG Communications, Inc.

7 secrets of successful remote IT teams