Data Visualization to Data Imagination
Recently, while scanning through my Instagram timeline on a chilly morning waiting for the barista to serve my morning dose of freshly brewed coffee, I bumped into a visually powerful ad from L'Oreal that was artistic and making a statement to men in the power circles of large companies to hire more women that's worth an investment. While the ad campaign is hard-hitting with a great copy, what intrigued me was the simple yet innovative way of representing data using cosmetics to demonstrate the power and benefits of hiring women leaders.
“Visualization gives you answers to questions you didn’t know you had.” – Ben Schneiderman
While this ad kept ringing in my mind at the start of the weekend, as a data enthusiast and product manager it got me fascinated towards applying the same thought and belief towards data visualization. While data-driven reporting which was earlier confined to technology teams, business focused teams were either completely dependent on IT or had their reporting force dishing out reports using spreadsheets and charts on presentations showcasing results of business KPIs. While those stop gap solutions kind of solved the expectations using data, the question is - did that report give you the signal vs noise that is in large historical data sets? Well, not many realized this until the power of big data started to unleash its potential a few years ago paving the journey of building dashboards not only for the business user, but even to customers who engage in business. These could be small and nimble dashboards on spends, transactions and promotions that can be viewed on a mobile app, to the heavy duty enterprise dashboard that will put even a power company's alert line of "High Voltage - Keep Away" running for cover.
While the journey for data visualization is still evolving, what can we learn from the mind numbing ad from L'Oreal? Data visualization should not be considered as a dashboard which has a bunch of Histograms, Combo Charts, Stacked Charts, Bubble Charts or Heat Maps. While these representations are an absolute necessity but does it engage with the persona, will it enable the persona to make a decision, or should it also predict the potential risk that can be mitigated to avoid business losses? These are questions that need to be answered while the urge for coffee started to build up to satisfy my caffeine hungry soul.
As product leaders, we should be presenting data that is larger than life, data visualization that is memorable for users. As the saying goes - "A picture is worth a thousand words", visuals that are intuitive can and more strongly be tied to memory. Data can visually help users to commit and make decision making easier, it also enables users to pay specific attention to not only pattern based reporting but also predictive information in a hyper competitive world. Additionally, in today's day and age companies portraying business in a fast paced environment, the short attention span of users hardly pick up cues from a not so visually appealing dashboard that aligns to their products, services or customers. Dashboards should and must tell you a story, dashboards that is a canvas that you can use to not only dish out a report but add business value.
There are several data visualization tools from open-sourced or those which is as expensive and complicated that makes you feel that it's beyond your pay grade to even understand the benefits of such a tool. But all have a business value and competitive stronghold that separates them from each other, there is immense time, patience, research and investments that have gone into those tools over a long period of time. But it is yet to answer one question - Does the data visualization tool connect me to my business or organization, well some say yes and some say "maybe" while most of them say i really don't know with a notion of reports looking like a massive spreadsheet waiting to explode.
Organizations should start thinking about data visualization that has a combination of art and science where even a 10 year old kid on the block would understand. It should start with the process of understanding the end user, identify teams that would benefit from the visualization of data and what drives them using it. Simple charts that don't intimidate users with large data sets, models that provide visual validity and meaning to their business. Representations that showcase patterns, changes, trends and enable the user to verify a hypothesis. Amalgamate such data representation to the core brand and products of the organization, this interjecting process can potentially be the next big wave where user experience moguls and the army of graphic designers with their design thinking concepts and ammunition tools can help in laying the foundation of what and how data should be presented.
When personas, user groups, and users across sales, marketing, operations, digital transformation experts, product analysts, and most importantly the corner office have been identified for visual representation, the next step is to test and iterate to gain impressions and opinions on what they have never seen in their business domain. Applying research conducted by large companies and foundations will help in influencing iterative decision-making address human perception and psychology to identify what data should stay or stand out. While such exercises is a long drawn process; it's worth the effort in understanding the fundamentals and lay the road to great dashboards that not only will be a visual treat and avoid stressful faces when creating reports for specific business cases.
While most checkboxes are ticked off understanding the annals of a users mind and requirements, the designers and available tools can help in preparing dashboards that not only help in showcasing organizational KPI statuses, it should also invoke the most granular detail making it a storytelling episode on any form factor of choice. Some of the dashboards can be line charts with smooth edges that illustrates continuous data with the brand identity for weekly and monthly sales for a retail setup using a time series model and predictive trends based on various variables from a historical standpoint.
Another aspect is histograms that can be represented as pills of a pharma company showing the specific distribution of products that are plotted against quantitative data over a period of time with typical metro line maps showing the status of their supply chain from manufacturing units, shipping lines, road transportation and dealer bubbles on stocks and sales. There is so much creative energy that can be used and applied while offsetting traditional ways of plotting data. The humble pie chart that looks like a coin where extrapolated data shows the financial condition of a borrower's assets and liabilities with operating cash flow enabling an underwriter to make a credit risk assessment and arrive at a potential amount for lending. And then we have the evergreen category managers who would like to assess FMCG products using a Nightingale chart that looks more like a flower than a chart but driving the point towards sales and adoption with each unfolding petal.
Product strategists who have the flair for data and a design bent of mind should be not just developing data-heavy dashboards but drive design excellence on providing contextual and fluid data that is appealing with a clear sense of what is most important for the user, the logical data flow with creative thinking on what could potentially be the next steps that can be envisaged in a data-driven organization.
As I begin to sign off, the curious case of the barista walking up to me with a smile - Another cup of coffee? while I politely decline to come back another day to delve on making the bold and beautiful dashboard into a lethal and trustworthy product enabling organizations to be competitive using data science models and machine intelligence.
