How McKinsey is Dealing with the Machine Learning Challenge
All industries are facing a great challenge regarding how to take advantage of all the data they have available to steer their business. How can strategy consulting firms, known for their "generalist approach", help its clients in a topic that requires such specific knowledge?
In 2015, McKinsey & Company, a leading global management consulting firm, with nearly one century of existence, bought the start-up QuantumBlack (QB) – a small company founded in 2009 with the goal to reimagine how organizations coulduse data analytics and machine learning to outperform their rivals.1
The deal is one of several acquisitions that the consulting company has made recently in order to develop the capabilities needed to help its clients do business in an increasingly data driven environment.These include acquisitions of design firm LUNAR, aerospace and defense analysis firm VisualDoD, and retail analytics firm 4tree. By acting through acquisition, McKinsey gets access to a talent and knowledge pool very difficult to create organically in a large organization, and by keeping them relatively “independent” inside the structure of the Firm, it enables the start-up environment to continue to exist and keep bringing sustainable innovation to its clients.
The reason McKinsey moved into the “tech arena” is that data is proving to be an increasingly valuable resource. QuantumBlackuses data science and machine learning techniques to identify solutions for real world problems– which is exactly what McKinsey’s clients need. From detecting fraudulent financial transactions to using performance data to predicting injuries for professional athletes, all the knowledge and capabilities QB developed during its early stages of existence can be used to enhance performance in well stablished, traditional companies too.
And the reality is that even with the increasing amount of data available in virtually every sector of the economy, combined with the several academic advances showing what is possible to accomplish with machine learning, companies are still having a difficult time harvesting all the potential from these technologies. As stated in a recent HBR article, “The gap for most companies isn’t that machine learning doesn’t work, but that they struggle to actually use it.”2
That is when McKinsey can add value: the role consulting can play in these early stages of the Machine Learning era is also related to help its clients to understand the potential and the several uses this technology can have. To enable this, project teams are designed as interdisciplinary teams, consisting of strategists – the regular “generalist consultants” – data scientists, engineers, and designers.The generalist consultants often work as “translators” between the data scientists and the clients, making sure the business issue being tackled, as well as the technical constraints of the technologies used, are aligned between all parties involved.
According to Nicolaus Henke, head of McKinsey Analytics,“Data analytics is one of the most disruptive and potentially transformative developments in management. QuantumBlack, through its Nerve platform, combines data from disparate sources to produce meaningful data around human endeavor—something that previously has not been reliably measured. This is why we’re so excited about the combination of our organizations—together we can provide our clients with a completely new means of making decisions and converting knowledge into action.”3
In a study for a global pharmaceutical client looking to use data to reduce time-to-completion in the clinical trial process, QuantumBlack used it proprietary tool NerveTMto combine data from 350 clinical trials, with 150,000 patients across 30,000 sites (which translated to 250 million rows of data), and by linking these complex data sets they were able to provide their client an integrated view of their data and, resulting in 15% reduction in time per clinical trial and 11% reduction in cost.
The impact machine learning can bring is indisputable. Moving forward, one option for McKinsey would be partnering up with universities to stay at the frontier for the development of new technologies inside the data analytics environment. Of course, these new developments will not be business-ready in the short-term, however getting involved in the early stages of development will help McKinsey understand the trends in the data analytics space and be recognized as an authority in this topic.
Also, internally, as projects become more and more dependable on data analytics and its complex approaches, McKinsey also needs to make sure the capabilities needed for these types of project are also spread across the organization between all consultants.
Following the trends on innovation, how can McKinsey and other consulting firms build the evolving capabilities needed to support their clients in this new business environment? What are the biggest competitors the consulting firms will face entering the AI and other high-tech driven market? (723 words)
1QuantumBlack company site – January 2018 [URL], accessed November 2018
2B. Schreck, M. Kanter, K. Veeramachaneni, S. Vohra, R. Prasad, “Getting Value from Machine Learning isn’t about fancier algorithms – It’s about making it easier to use”,Harvard Business Review (March 2018)
3 “Accelerating with QuantumBlack”, New at McKinsey Blog– December 2015 [URL], accessed November 2018
Student comments on How McKinsey is Dealing with the Machine Learning Challenge
I appreciate the perspective to focus on the value-add McKinsey can bring by connecting companies/startups specialized on Machine Learning with the needs and wants of their clients. From my perspective, McKinsey will and should not strive to be the most innovative company in the space of AI or Machine Learning – as it simply does not reflect their core competencies. However, their close work with clients across a broad range of industries enables them to better understand where and how to apply this technology which will be key in the future.
I have learned so much about consulting firms since I started HBS. What impresses me most is their ability to use what they learn from their clients to make themselves better firms. The above article is a perfect representation of that.
Helping companies better use technology is a huge area of need. this is especially true as older companies fight to stay relevant and at the forefront of the industry. Mckinsey already has very established relationships with these companies already, so this acquisition is a natural fit for them.
I’m curious to learn why AI/ machine learning “does not reflect [McKinsey’s] core competencies” as mentioned in the previous comment.
I think McKinsey’s acquisition of QuantumBlack to develop their in-house analytics function is a very strategic move, and in the long-term it’ll reap market share and add to McKinsey’s competitive advantage. I’ve seen McKinsey Analytics present some of their case studies, and the predictive ability and granular analysis they are able to perform for any industry is truly revolutionizing.
In addition, I found the infographic made by McKinsey Global Institute very helpful in informing which industries are at most risk/ safest in the face of the coming wave of AI replacing human jobs. (Link to Infographic: https://www.mckinsey.com/~/media/mckinsey/business%20functions/mckinsey%20digital/our%20insights/where%20machines%20could%20replace%20humans%20and%20where%20they%20cant/sector-automation.ashx)
I find the article very interesting considering the challenges and technical expertise that machine learning can face when implemented in traditional organizations. I am aware of the strong focus that McKinsey has on this arena, having friends and family that have joined the company not as traditional Consultants but as “Data Translators”, as the McKinsey names this pretty new position. The quote that “The gap for most companies isn’t that machine learning doesn’t work, but that they struggle to actually use it.” is not only very illustrative but also very applicable to all kind of companies across industries and geographies, and reinforces the idea that machine learning is a tool for organizations and employees to make better decisions, which will be more or less valuable depending on the people’s level of understanding of the system, the data, and its potential. At this still early stage, positioning as consulting company that can thrive on data analytics and machine learning is without doubts a significant edge and hedge for McKinsey in the future, and a way to innovate in the status quo of the professional services industry.
Thank you, NCB (“National Commercial Bank a.k.a. AlAhli Bank”?) for your post. Your post was a joy to read as it brings up a theme that I, Bruce Willis, have heard repeatedly (yes, I engage in thoughtful conversations, not just “Die Hard” movies). The theme is: Everyone speaks about data, but how to use it? Especially for old industries such as yourself in an emerging market (do confirm if you are NCB). Helping companies such as yourself make use of data will help emerging countries improve productivity and advance in their pursuit of becoming a developed country.
Reading your post gave me further food for thought. First, will McKinsey stifle Quantum Black’s growth? Quantum Black appeared to be a threat to all consulting firms, and now being owned by one of them raises the question if McKinsey will let Quantum Black grow unhindered or if in protecting their current business model, McKinsey will try to integrate them limiting Quantum Black’s full potential. I understand the benefits you describe but being in a still nascent field (as machine learning) I would be afraid (and I’m normally not afraid) that the parent company, won’t be nimble enough to adapt to new fast moving requirements.