Data & Digital Transformation
The ability to use data to anticipate the future is critical in today’s fast moving economy, with data being created at an ever increasing rate, how are some companies able to separate the signal from the noise? It starts with the principles of data, to transform data into actionable insights. Your mind is probably racing with ways to apply analytics at your firm, but to fully embrace and leverage analytics, your team needs to move beyond one off projects and establish a sustainable practice that’s ingrained in everyone’s day today.
Take Airbnb for the example, the company fostered a data driven culture by creating an in-house data university. 30 data science courses that are open to all employees Amazon, Zillow and Google have similar data obsessed cultures. If your company aspires to be like these analytical leaders, there’s a five part framework you can implement.
I found in working with companies that they often would say, OK, I buy that analytics can be useful. How do I get better at it? How do I improve my analytical capabilities? What do I need to do so when I was thinking about how to help companies improve their analytical capabilities, I wrestled with what’s the best way to represent it and I came up with this acronym Delta D for data. E for enterprise orientation or strategy, L for leadership, T for targets and A for analysts.
First D for data. It bears repeating great analytical companies relentlessly gather and leverage high quality data. It’s what distinguishes the leaders from the pack. Data is the not so secret sauce of analytics and I find it’s one of the greatest differentiators between the companies. Really good at analytics and those that struggle to find an exclusive line of sight into your industry. Get creative and look for underutilized or hard to capture data.
Take the Boston Red Sox. Who leveraged underutilized data to gain a competitive advantage, the Red Sox. My favorite baseball team who decided they wanted to learn more about who would make for a good player. So they went to NCAA headquarters to look at paper records and classify. The attributes of every rookie in professional baseball and then figure out which players ended up performing at a very high level so they could then know what attributes they should be looking for in up drafting players nobody else was willing to go to the trouble of looking through these paper records, recording the information this unique angle into scouting talent paid off. The Red Sox won four championships between 2004 and 2018.
Next in the Delta model is E for Enterprise moving analytics from sporadic and siloed activity to a focused and networked capability. When I first started working with companies on analytics, there were all these little pockets. Maybe there’s a little bit of stuff going on in the market, research department or a little bit of analytics in the quality department, but nobody really talked to each other about it. They didn’t share approaches, there was no central group at also an enterprise orientation to analytics says. As an enterprise, we’ve decided this is an important business capability, strong enterprise strategies until several components, including setting and analytics strategy and road map managing a unified data and analytics platform and improving data literacy across the company. Key to a successful enterprise strategy across the board. Essentially managed analytical resources. You don’t want every group having a separate approach to analytics. When I started working with departments that were really good at this, they had a centralized group. They manage data at an enterprise level. They prioritize their applications, their targets at an enterprise level. So they’ve just realized this is something important that we do as an organization and political leaders separate themselves from the pack. By going beyond one hour projects and taking an enterprise wide approach to analytics, take Capital One. Capital One was one of the companies that I analyzed very closely and you know they even had a cultural slogan for analytics. They called it. Information based strategy IBS and that worked well for many years ,basically you talked to anybody in the organization about IBS and they said yeah, that’s the core of what we do. Analytics is based into all of our decisions and actions at a time when the banking industry was defined by one size fits all products, Capital One went against the grain, leveraging IBS to create a broad portfolio of tailored products and personalized marketing offers. This strategy helped Capital One win over fragmented markets. Including auto lending and retail banking, the firm success demonstrates another capability among best in class analytics companies.
Leadership: leadership is a really important thing in analytics. I found a number of analytics leaders over the years that were steadfast in their appreciation for how important this was, Capital One CEO Rich Fairbank exemplifies such commitment to analytics. Fairbank co-founded the company on the basis of the analytics he planned to use in 2002. Fairbank hired the world’s first chief data officer and over his tenure has fostered a culture of data driven. Product development. These are people who not only are committed for the rest of the organization to do it, but they set an example in terms of their own decision making. They push back on their employees and even other managers when they don’t use analytics to make decisions. When they start telling anecdotes about customers rather than. Using data to describe what’s going on, they sometimes get personally engaged peers, even CEOs helped design the data architecture for the organization. They just get highly engaged in the process of analytics and touted to the organization as a way to compete effectively, however, you don’t have to be in the C-Suite to influence your firm’s progress with analytics managers and individual contributors play a critical role in an area where many firms struggle deploying models that integrate with everyday operations the ultimate. Goal of any analytics oriented project is doing something differently, so it’s important to deploy models because nothing changes in the business unless you do that. You don’t get any more revenue or profit, you don’t get better at a particular process unless the model is actually being used on a daily basis. As per a survey, 87% of models are never deployed beyond the pilot lack of funds, aren’t the main excuse, especially since open source software and cheap cloud computing make analytics nearly free. The largest barrier to deploying models into everyday use is changing human behavior. They’ve set up problems are, I think, often the most difficult to solve, which is they involve changing people’s behavior. Change management in general, persuading people to adopt a new system, even when supported by rigorous data analysis, is difficult. It’s persuading a senior executive a look this is really going to improve your business, and it’s worth shaking everything up in your operations to implement it. It’s persuading the organization. Look. I’ve got a great model. I want you to inject this into the CRM system so we do a much better job of prioritizing leads. It’s persuading salespeople, look, you’ll sell more if you use this lead scoring system, then you would if you just kind of relied on your intuition. It’s not just leaders that need to be influenced. Models often impact workflows several steps into implementation. If impacted, employees don’t buy into adopting a model on the job regardless of their position, then the models value is void.
Take UPS drivers, one of the key stakeholders in their organizational effort. One of the biggest analytics projects ever was at UPS. UPS decided that if they used an algorithm to tell drivers where to go next, they could save a lot of money in fuel. They could save a lot of driver time, make them much more efficient. But you know the drivers, they’re pretty attached to their daily routes and thought they knew better to ensure success. The firm invested heavily in change management. They spent several $100 million in deployment, so just a huge amount of change management work. They were surprised how much it was necessary. The whole process took over a decade to totally roll this out.
Change management efforts can also benefit from aligning a project to the 4th Delta capability targets or having clear business priorities that can benefit from analytics. We can’t be analytical about everything at once, we have to make decisions about what’s really critical to your business strategy.
Where do you need to get better at analytics first, a strong target takes one aspect of your business and lays out a lofty but feasible road map for how to get there. So. Target needs to be ambitious in terms of you know, how it will change and improve a business, but it also needs to be approachable. So maybe you say for my customers, for example, I want to be the best in my industry at certain aspects of Customer relationship management. I want to have the best targeted offers. I want to have the best marketing lift from my marketing campaigns. I want to have the best customer data in my industry that’s nicely integrated and of high quality and so on. Take DPS Bank in Singapore. Which was named the world’s best bank in 2020. Piyush Gupta utilized internal scorecards to set and track targets for its digital transformation initiatives. The scorecard listed the firm’s top priorities of the weight of each priorities importance and what metrics indicate success. The approach has been credited for the banks dramatic improvement and customer service.
Finally, analysts no analytical progress is possible without talented data professionals to attract and retain data, scientists look to provide challenging problems, support continuous learning, encourage analytical decision making, and offer competitive compensation. Take. Goldman Sachs to stay competitive with other banks, the company recently raised its junior analyst salary by nearly 30%, from 85,000 to $110,000 a year in order to better attract top talent. The financial services company also offers challenging work and opportunities to learn data science. Concepts and programming languages. If you can’t compete for Premier data talent or have trouble finding these professionals in your area, begin developing a pipeline with your local universities. MIT, a great institution, have an analytics program now that also has, and internship companies supply problems to solve, so don’t just take the graduates after they come out. Work alongside with professors, work with students in those programs to identify the ones that you want. So firms looking to evolve analytics into a competitive advantage. There are a few key takeaways. First, use the delta model to score your data and analytics competencies. No matter what your role is, in my view, it’s helpful at almost every level of the organization, even if you’re not an enterprise analytics person, you can say, OK, I run the marketing. Function. We’re going to do some analytics in marketing. What do I need to be successful at? It turns out you need the same five general factors and you can look at your own progress even as a marketer, you could even apply it to your own individual analytical capabilities as you make decisions. Second, maturing as an analytical company involves more than collecting data and designing processes. Apply empathy to all employees who may be impacted by a deployed model. In the case of UPS, this meant riding along with drivers in the early stages of the process, they talked to drivers they had the algorithm developers actually, you know, go out on the routes. So they understood a little more about what was going on. I think it’s quite fair to say that empathy is a personality attribute that helps a lot in analytics.
Capital One is the newest kid on the block of top American banks, its peers, JP Morgan Chase, Bank of America, Citigroup and Wells Fargo. Have all been around for more than a century, despite its youth, Capital One is the 6th largest credit card issuer in EU S It’s returned on shares grew 156% in the past year alone, the highest by far of all major banks. How did Capital One grow to compete with such established Titans? Capital One success is a direct result of its early adoption of data analytics. The firm doesn’t just position itself as another financial services firm, but as a data driven technology firm that happens to be in financial services in doing so. Capital One is a pioneering example of embracing Delta across their organization. First, data Capital One has access to vast amounts of organized data. Since the 1990s, Capital One is mine customer data and credit ratings to feed its proprietary relational database management systems that track. 10s of millions of customers, Capital One also makes strategic acquisitions to acquire new data. For example, Capital One acquired data aggregator firm Bundle Corp in 2012, bundle had access to over 20 million Visa and MasterCard branded cards and public data from EU S government Capital One. Gives us this data to enhance its understanding of customers. Capital One is also known for gathering data from experiments when it needs to. It started with very ambitious direct mail experiments analyzing every aspect of an offer, including the color of the envelope. Now it uses the same experimental approach to evaluate. Email, web and mobile offers second, enterprise Capital One is committed to taking an enterprise approach to managing systems data and people. One example is Capital One On premise data centers to the public cloud and its leaders said that because it didn’t have to manage all of these on premise systems, it could devote a lot more attention to analytics and AI. This all increase the ability to handle data at scale, providing greater access to even larger and more varied datasets across the organization. Thousands of analysts access to platform and conduct millions of queries simultaneously without any loss of performance. Employee training further ensures that capital one’s Open Access platform is accessible to employees. This also helped Capital One achieve companywide buy in for its cloud first policy training wasn’t limited to engineers. Either business executives and non technical stakeholders were encouraged to participate as well. Third, leadership Capital One was founded on the belief that the banking industry would be revolutionized by information and technology. The founders who took over Signet bank’s credit card business in 1988 believed that. Building a data driven culture would help its teams offer customized solutions to customers. This led to the creation of Capital One information based strategy, or IBS, which is still part of Capital One’s DNA. Today. Rich Fairbank, the founding and current CEO, continues to champion data by appointing the business world. First known chief data officer, or CDO, in 2002. Since this pioneering move, the number of CEOs has skyrocketed. In a survey from 2012, only 12% of Fortune 1000 companies had a CDO. By 2018, almost 68% of firms reported. Having 1/4 targets, Capital One analytics support the firms information based strategy at the firm’s founding. The primary focus of analytics was to provide targeted customers with custom product offerings. For instance, Capital One introduced the innovative Balance Transfer credit card. This incentivize customers who had money on higher interest cards. To transfer to Capital One with a lower introductory rate, more recently, analytics and AI have been embedded within the products themselves. For example, capital one’s intelligent assistant, Eno helps customers with tasks such as managing fraud, alerts and viewing account balances he know is pre trained using machine learning. And continues to learn overtime as every interaction adds to its language database. This ties into Capital One future vision for banking. The industry will be driven by real time automated, intelligent and proactive interactions with customers.The company is able to retain its employees by investing in their growth Capital One developed an analyst development program that offers new graduates hired as business analysts and data analysts. A two year rotation across various departments. Program associates receive hands on learning to become innovative problem solvers using data. Although other top financial firms hire similar quantities of analysts, Capital One stands out because it hires at least twice the number of engineering tech talent. This supercharges capital one’s analytics capabilities, as ITN business team collaborate closely to foster real time problem solving for companies interested in applying the Delta model. Capital One, I think offers a few key learnings. First, communicate effective data analytics strategies as part of a broader company wide vision Capital One positioned its information based strategy as a North star for all company initiatives. Anybody you talked to in the company understands what is meant by. IBS. Secondly, technology talent is key to creating a culture of data collaboration Capital One not only recruits and hires the right talent, but has programs to develop and upskill employees. Analytics approaches as new technologies and techniques become available. For example, Capital One shifted from a premise based technology environment to a cloud first environment. Now to an AI first approach, AI and machine learning are now tools used to power virtually every facet of their business.