Approach Your Data with a Product Mindset
When it comes to leveraging data to improve their day-to-day operations, organizations need to look beyond building a team of skilled data professionals and arming it with the latest analytics tools and technologies. A well-equipped analytics team is a necessary condition of using data to drive business value, but it’s not a sufficient one. To more actively contribute to the achievement of critical business outcomes, analytics teams should start viewing their random forests and neural networks through a product development lens. By bringing a product development approach to the planning and execution of analytics, organizations can scale isolated successes into the kind of sustained, organization-wide, data-driven decision-making that serves as the foundation of genuine digital transformation.
According to an International Data Corporation (IDC) report, at least half of global GDP will be digitized by 2021. Unfortunately, as the report explains, “While most organizations are attempting [digital transformation], only a small percentage are getting it right.”
In most cases, the tech isn’t the problem. Many organizations’ failure to adapt to a highly digitized business landscape stems from their inability to convince employees to embrace data-driven decision-making. A NewVantage Partners survey found that 48% of executives cite “people challenges” as the foremost barrier to being more data-driven as an organization, compared to just 19% who cite “technology.” As such, while a well-equipped analytics team is a necessary condition of using data to drive business value, organizations need to look beyond building a team of skilled data professionals with the latest analytics tools and technologies if they want to leverage data to improve their day-to-day operations.
To more actively contribute to the achievement of critical business outcomes, analytics teams should start viewing their random forests and neural networks through a product development lens. By bringing a product development approach to the planning and execution of analytics, organizations can scale isolated successes into the kind of sustained, organization-wide, data-driven decision-making that serves as the foundation of genuine digital transformation.
Adoption is the key metric of product development, and as such, identifying — and addressing — end users’ unmet needs is a central imperative of nearly every product development methodology. Acting on this imperative requires analytics professionals to take a step outside their comfort zones and assume more collaborative roles than they are accustomed to.
Instead of independently developing a new method of performance measurement or a new algorithm for automated customer segmentation, analytics professionals should start by asking their colleagues in each department, “What key business questions are you trying to answer?” and, “What outcomes are you trying to achieve?” A marketing team might be trying to solve for something different than an IT team, which might be trying to solve for something different than an executive team. Each of these teams is only going to adopt an analytics solution if it solves for its own unique problems, so an analytics team must conduct extensive stakeholder interviews to develop a clear understanding of the diversity of unmet needs across its organization.
Only after it has gathered all these unmet needs should an analytics team determine which datasets and analytical techniques it will use to develop effective point solutions. That said, agility and iterability are also central imperatives of product development, and analytics teams should treat them as such — nothing about data necessitates a plodding, epochal approach.
As an example, imagine you’re a data professional at a major retailer that’s trying to navigate the disruption caused by the COVID-19 pandemic. If you ask your Director of e-Commerce, “What key business questions are you trying to answer?” their answer is unlikely to involve insufficient traffic — the tragic circumstances of the pandemic have forced an inordinate number of shoppers online. They might, however, note that many circumstantial online shoppers have a low propensity for engaging with information-capture mechanisms like newsletters or “secret discount offers.” As such, your directive might be to analyze these shoppers’ online behaviors to figure out the best way for the e-commerce team to capture some information that could be used to increase the shoppers’ customer lifetime value.
At the same time, your EVP of Store Sales is likely to offer a dramatically different answer to, “What key business questions are you trying to answer?” With most of their stores temporarily shuttered, their most pressing problems might be figuring out how to adjust their inventory management in a way that supports the sudden shift to e-commerce-only operations and how to minimize friction for traditionally brick-and-mortar shoppers as they transition to online shopping. These problems require entirely different analytics solutions than the Director of e-Commerce’s problem.
Their differences notwithstanding, each of these analytics solutions must be prototyped and iterated on quickly to ensure the ongoing viability of your business. This is where applying the kind of agile product development methodology commonly used in software development becomes invaluable. In a high-pressure, high-volatility business situation like the one precipitated by COVID-19, you need to deliver a new information-capture mechanism to your Director of e-Commerce and new inventory management guidelines to your EVP of Store Sales in no more than a week or two. Your best bet is to develop some sort of analytics solutions to put into play immediately, observe how they’re received by both internal and external stakeholders, and iterate as necessary.
The example above illustrates how a retailer might bring a product development approach to the creation of analytics solutions, but it doesn’t cover the creation of full-fledged analytics products. Your new information-capture mechanism and inventory management guidelines might prove to be tremendously effective, but they’re only point solutions — solutions to highly specific business problems.
Productization involves abstracting the underlying principles of successful point solutions until they can be used to solve an array of similar, but distinct, business problems. Here’s how this might play out in the context of our example:
As these kinds of discoveries accumulate, full-fledged analytics products begin to take shape — products that can do much of the legwork of dismantling organizational silos. Discoveries in the first vein might lead to a product that can be utilized to drive customer engagement in a wide variety of circumstances. Discoveries in the second vein might lead to a product that can be utilized for resource optimization across your company’s departments — marketing, sales, technology, etc. These products’ settings and source data will vary depending on the context, but they’re fundamentally consistent (though not necessarily entirely static) packages of analytics tools and techniques.
Developing a robust portfolio of analytics products is an essential component of moving toward sustained, organization-wide, data-driven decision-making. But, as alluded to above, fostering the adoption of these products is equally essential — it’s difficult to characterize a product no one uses as a success.
Adoption hinges not only on the extent to which analytics products are calibrated to address unmet needs, but on the way the products are packaged and presented. It’s unusual to think about analytics as anything other than massive spreadsheets and incomprehensible algorithms, but true productization involves taking the target end users into account.
A brand manager at your retail company would struggle to identify — let alone use — an unnamed suite of abstracted analytics point solutions, but they could easily use your company’s prepackaged, internally branded Customer Engagement Platform. Simply put, easily identifiable analytics products that offer out-of-the-box functionality in a variety of business contexts are the key to driving data-driven decision-making across an organization.
Ultimately, digital transformation isn’t a matter of purchasing a 1,000-user license for a popular business intelligence or data visualization tool and encouraging employees to use it, nor is it a matter of giving an analytics team free rein to do whatever it wishes in a silo. It’s a matter of identifying each team’s unmet needs, developing analytics point solutions that address these needs, abstracting these solutions into analytics products that can be used to address additional needs down the line, and packaging these products in a way that drives adoption by stakeholders across the organization.
This product-oriented approach to analytics breaks down team silos through the Agile development process, promotes internal technological innovation, and maximizes the number of employees who actively use data to inform their day-to-day work. It could be deployed just as easily by a pharmacy retailer attempting to accommodate the skyrocketing demand driven by the COVID-19 pandemic as by an apparel retailer attempting to adapt to a sudden pivot to online-only operations. It could be used to see an organization through a time of crisis just as easily as it could be used to optimize operations in a time of calm. In short, irrespective of the contextual particulars, it lays the groundwork for organizations of all kinds to leverage data to achieve critical business outcomes.
Jedd Davis is the chief product officer at Publicis Health.
Dave Nussbaum is the chief data officer at Publicis Health.
Kevin Troyanos is head of analytics at Publicis Health.
Approach Your Data with a Product Mindset
Research & References of Approach Your Data with a Product Mindset|A&C Accounting And Tax Services
Source
0 Comments