UX vs UI, they are used interchangeably, which creates problems for product managers and business owners. Find out the difference between the two.
Why You Need UX Based on Data Science to Guarantee Product Success
August 12, 2020
Created in April 2019, Tesla uploaded a 2-minute video on YouTube to showcase the full self-driving mode. All models have a 360-degree camera, 160 meters forward-facing radar, and 12 ultrasonic sensors. Not to mention the touch screen display and over-the-air software updates. If anything goes wrong with the car, the company will check, and if it turns out to be a software problem, Tesla will update its software. Tesla's approach to the customer experience is very different from traditional automotive companies that need to recall their cars in order to fix the issues.
The example above is about data science. To be precise, data science is a combination of statistics, computer science, and more. It is the process to mine data, analyze it, model it, and present the results. Most Google searches ask about how to become a data scientist. Still, this particular article will talk more about how to utilize data science for your business of products and services.
Managing data science
Lecturers at John Hopkins University created a book titled "Executive Data Science." In the module, they mentioned that we could call a data science experiment is successful when:
- It creates new knowledge
- Decision or policies are made based on the outcome of the experiment
- It results in creating a report, presentation, or app with an impact
- Or, when the conclusion is that the data can't answer the outstanding question
Data science begins with a question built in mind, followed by data collection, formulation of the data model, and finally, communication of the results.
Building a data science team needs integration in the company: each individual can work separately but can be working together as a team on a big project. They also need to communicate with other teams in the company, such as the marketing team or the product team, who can be the users of their projects.
How big companies use data science in their businesses
Facebook has a research division with data scientists. They investigate a lot of things, including how people react with Facebook Lite. One exciting research was about social comparison on Facebook. A finding states the fact that when you ask people to think when they feel worse about themselves, 1 in 5 could recall a time they felt worse after seeing a post. And yet, they didn't know you can unfollow or snooze someone to hide their jobs.
In Netflix TechBlog on Medium, they mentioned that one of the problems they are facing is to determine which movies are trending based on languages. If the content is more prevalent in Language A than Language B, then Netflix will sequence their efforts accordingly. The way data science helped their decision making is how they can turn historical viewing trends consumed across a variety of languages into insights.
Users of data science and AI
Both customers and employees can benefit from data science and AI. In many cases, customers utilize an app from a company to get updates on their order deliveries. From the customer's point of view, the steps of communication involve browsing products, purchase, confirmation of purchase, delivery notice, tracking update, and arrival notification. However, at the back-end, we see that the task requires multi-party communication. As the customer fills in personal information, his or her profile is stored at a Customer Relationship Management program. After that, the order information is sent to the warehouse, where it confirms availability. This information is passed to the delivery team with trucks ready for shipment. Note that if the product is unavailable, then the warehouse will contact suppliers to prepare the stock.
Now, imagine if there's a delay in the delivery of a product or service due to unforeseeable circumstances (e.g., winter storm). Airlines could potentially delay or even cancel flights. At this time, customers would ask for a new ticket with complimentary lodging and meal coupons. Then the customer will get an update from the company with the coupons. From the employees' perspective, the back-end took care of multiple things at once. First, CRM pulls up a specific customer profile. There, the employee can know whether the customer is a member or not. The employee then arranges for new tickets and coupons to give matching the profiles. He or she updates the information through the airline's app or possibly to the customer service desk.
Data science is robust because it creates an integration of multiple data sources.
Launch, Mature, Growth: Your current stage of business matters
In Executive Data Science, the authors mentioned how startups should first focus on infrastructure: making sure that your data house is in order. This makes sense because when you're in the early stage, you are collecting data from customers. For example, you can get their personal information, such as email addresses and home addresses. You can also start collecting behavioral data that will accumulate over time as they use your product or service. Managing your data house means you consider how the computer will store inputs, how the computer will process data, and how the computer will show the results to users. Depending on the user type (is it sales? Is it the product?), the related program will show filter parameters differently.
As you mature, you now see the potential to enhance your product and service offerings by using existing customer data. To do this, your data science team will have to do experiments. For example, they can use machine learning to predict future consumer behavior based on past purchases. The data scientists will write an algorithm that will be implemented in the whole system. As a manager, you would want the result in the form of crucial visualization: you should know what works and what doesn't work.
In the Growth stage, your company is then looking for opportunities beyond your current revenue channels. What this means is that you might be looking at growth in specific demographics of your users and decide to expand your business to cater to their needs. An example would be the Indonesian unicorn startup "GoJek" that initially was a third-party platform connecting motorcycle riders with customers. They grow their business to deliver food "GoFood," and even more, they integrate a mobile payment system called "GoPay." In terms of data science, you will then need a team that can experiment with many different trial projects. And in doing so, you will need a dedicated data science team that overviews the results and successes.
Data science during a downturn
McKinsey wrote an article about leveraging analytics in times of downturn. They encourage leaders to strategize using AI in the company, move the AI topics to priority, reskill workers, hire when others don't, validate data and models, and establish methodologies or protocols. The key here is to view AI as an investment in the longer term.
Keep in mind that technology is becoming cheaper, accessible, and more advanced. When you consider investing in technology, you can look at it as a long-term investment. You can use Cloud services, such as Amazon AWS, Google Cloud, or IBM Cloud Pak & Data.
The key is the willingness to learn
Tesla didn't start with self-driving mode right away. Their first goal was to defy the myth that electric vehicles are slow and ugly, and they proved it with a prototype: Roadster. It was in 2012 when Model S was launched, and only then, they started to develop autopilot features. Tesla consistently upgrades its hardware and software to improve its service quality.
In a sense, any business needs to adopt data science and AI slowly but surely. You can start by focusing on doing small things: automating emails, utilizing Google business hours, or personalizing your customer's experience.
At Designial, we offer Digital Transformation as a service to upgrade your business. Our approach to achieving business goals involves researching, analyzing, and documenting current processes. Then, we offer digitized touchpoints for employees and customers. Together with clients, we develop a digital strategy to come up with a conceptual solution given the budget, timeline, and technological constraints. In our case study with a company with more than 1,000 stores across 49 states that generates annual sales of roughly $19 billion, we increase efficiency by six times.
How to build UX Design to cater to customers' needs
Now, the question is how to apply data science in building business apps? Often the case that we forget to address the critical issues during the development process.
To begin with, we need to acknowledge the purpose of integrating technology into your business. Ultimately, the goal is to grow the business. You can achieve that either by increasing revenue or decreasing costs. Apps can be useful tools to do both.
In building a UX Strategy, it is essential to know your customer's pain points. They are the points when customers find difficulties getting what they want. For example, an e-commerce customer has to input his personal information to purchase a product repeatedly. Although he can still buy the product, the task feels like an extra effort he needs to do. A quick fix will be to add a system that can recognize a returning customer. Amazon offers a '1-click purchase' that fastens a customer's transaction process by seconds. Your business can provide the same feature to streamline the customer's journey.
We often think that apps are 'personal' and so, we focus on how the app can help one person. But in reality, there are millions of people using apps. In that sense, when you improve and upgrade your business apps, it would help many users at once. It is all about global-scale effects.
The other attempt you can do to increase your revenue is to increase retention. We can look at this task through emotional design. Apps should be designed to create emotional connections, which is critical to retain customers. You can use UX research to get to their mind better, building a customer base that fits the customer persona. A good example would be Instagram and Facebook. They use extensive data science to advertise businesses through personalized feeds.
Meanwhile, an app that reduces cost is an employee-facing one. One way is to increase efficiency, and an example is an app for warehouse management. Employees should be able to navigate the app with ease: knowing the amount of stock currently available, reporting damage or false shipment address, and tracking delivery to retail stores. For example, a customer comes up to a furniture store manager, asking if there's remaining stock of a new black office chair. The store manager would pull up his apps and check immediately. If the app is reliable, the manager can mention how many are in the storage. If none, he can suggest the next delivery date. If it's still the next week, the manager could refer to the nearest store chain with a stock.
The questions you want to ask and answer through the brainstorming process are this: What are the challenges the customers and employees face right now? How can apps and data science help solve the problems?
Building a data science team requires proper management, protocols, and methodologies, clearly defined questions to answer, and definition of success. Your company can benefit from data science, starting from a small-scale experiment: automation in operations, such as email marketing and Chatbots. As you build data infrastructure in your growing company, you can experiment through different projects to create a better customer experience.