摘要：
Businesses today have lots of data, modern data warehousing, AI (Artificial Intelligence) tools, and nice visualization platforms. Still, users across small and large enterprises globally are frustrated by their inability to quickly get answers to their questions from their data.
We now have data everywhere and are drowning in a sea of data. We have transformed from a data-scarce to a data-overwhelmed society, struggling with information overload. But users simply want answers from their data. Marketers, for example, want to know when a competitor’s product takes the spotlight from their brand. Doctors want to give pointed advice to their patients on what to do based on their medical records. There are many examples, but each user’s need is the same: to get answers to their questions.
The obvious question is: how did we get here?
Over the last few decades, we have continually digitized across a growing number of industries. Internet and mobile phones have enabled easy publishing and we now have data across text, images, sound, and videos stored in accessible cloud servers. We have warehouses like Azure and Redshift, and analytics like Big Query. These are exciting times for us data professionals. Unfortunately, I fear we are getting caught up in the tools hype. We are continuing to push more and more numbers, charts, and tables to our users. That is precisely the problem. We are not offering directions on what to do. Per Dell’s study, “Data overload has become a significant barrier to transformation”.
Data users don’t ask whether we use the latest AI or for every data-related detail possible. When we share large volumes of data, they push back, stating, “I don’t know what to do with all of this. Can you provide simple, actionable tips and not overwhelm me?” Users want to save themselves from data overload.
In retail, brand managers want to understand whether and how shoppers see their products. If a brand isn’t visible on Black Friday, that’s a massive problem. If a brand’s promotions, prices, or shipping isn’t as compelling as those offered by competitors, shoppers won’t buy. You can share prices, reviews, visibility, and more with a brand manager. However, the details alone aren’t valuable. You must really tell the brand manager when a competitor takes his/her spot or offers a better deal.
In healthcare, electronic medical records store patients’ medical conditions, medicine taken, lab results, x-ray, and more. However, the physician must know which patient to remind about scheduling a mammogram, colonoscopy, or other wellness procedure. The doctor must know which patient is likely to get sick and needs intervention. If a home monitoring device detects that a sick patient falls, the device must notify 911 to get them to the hospital and notify their doctor. All these values must be provided from the data.
In accounting, managers want to understand and have insight into financial ratios. Imagine a business owner not quickly knowing that cash is running low due to a bad accounts receivable collection. A customer service representative would need to inform a customer they will not get service unless payment is received.
Our industry is flooded with articles on data tsunami, overload, and fatigue. Users want to understand what happened, why, and what to do next. Gone are the days when you could tell your users that you will get back with answers tomorrow. Stale data is not helpful. Users demand answers now.
The right question is – how do we as data professionals address these user needs? I believe we must do the following:
1. Deeply understand the needs of data users
Creating the right answer that the user finds impactful takes many conversations and product iterations. We must directly ask users what they find valuable. Usage data is a good proxy to understand customer adoption. Understanding whether customers are correlating ROI with the presented insights is key.
2. Engage an interdisciplinary team to create insights
We want the undercover key performance indicators (KPIs) that allow users to spot opportunities to grow revenue and reduce cost. To remain unbiased from a data-heavy mindset, we need interdisciplinary team members including; colleagues from data science, customer success, content marketing, and design – to speak customer needs.
3. Combine macro trends with notable events
We must provide high-level trends with pointed trends in interactive visuals. Current tools like Tableau and Google Data Studios visualize the stories in the data. To gain user trust in our data, we must be transparent on how we generated our analysis.
4. Allow data users to access insights on their tool of choice
User preferences have evolved from “I will log into your software to understand what I need to know” to “alert me on my phone to tell me what I must do only when something important happens.” Today, data users ask devices like Alexa and Google Nest to remind and inform them of the news, time, weather, and more. This means that businesses must push information to a data user’s application of choice such as, podcasts, digital assistants, YouTube videos, email, and so on. Wherever our users are, that is where our solutions must be.
Big data, cloud computing, and analytics advances offer us tremendous power. To share these benefits with our users, we must provide impactful, simple, and easy-to-use insights quickly and cost-effectively using a data users’ tool of choice. Only this will provide the value today’s data users are asking for.
留下你的回應
以訪客張貼回應