"OK, Google, play my Friday Starts Now playlist."
"OK. Playing your Spotify playlist called Friday Starts Now."
Such is the advertisement for Google Home. Amazon, Google, and Apple are competing for Smart Home technology. All three created assistants that enable actions, such as playing music, turning on the TV, etc., through a voice command. As Home Depot said, a smart home can save you time, energy, and money by automating performances based on your lifestyle.
Some of us might question this technology. What is automation? Why do we need that? What if we do things like before: manually turn light switches on or adjust the room's air conditioner? How exactly can automating things help us?
Dr. Stephen Galsworthy once spoke at Spark + AI Summit 2020 as a representative of Quby, a European smart home energy company. He mentioned how more relationships between customers and utility, would increase the probability of energy saving. In the first stage, we have the traditional billing system where we pay after using energy at home. Next, we can get insights from past historical data. We can learn how we can save energy and money. Taking a step further, we can monitor the energy used at home. This idea plays a critical part in our energy usage because it increases our consciousness in spending energy. Finally, we can have insight from appliance diagnostics. In this case, Quby offered an app where users can look into the appliance's energy consumption and compare it to the industry standards.
Smart home technology illustrates how computers can use data and produce important customer insights. A good insight not only makes customers better informed but also helps them in decision-making.
As mentioned, computers can perform automation: a task where repetitive tasks are routine. Think for a second about your budget and expenses. For some people, it is hard to see how they go over budget every month. They try to save paper receipts and the data in a journal book. They wanted to record expenses and compare them to the account. But still, it's hard to perform the task daily.
Mint is an app that allows users to plan and track budgets. With the user's permission, it connects to bank accounts and automatically identifies expenses to different categories (for example, food and beverage). It alerts users when they go over budget for a certain category.
Let's now switch to the computer's perspective for a second. How could computers possibly do repetitive tasks so well? The answer lies behind algorithms. Algorithms involve a sequence of instructions to solve a problem. A useful algorithm first gathers data. How it works depends on how you define the problem. Let say the task is to categorize a transaction. The algorithm then asks for each transaction, does it fit the category "food and beverages"? If not, does it fit the category "utility"? And so on. Such is the repetition a computer goes through with algorithms.
Another example would be smart home technology. The algorithm compiles data based on your lifestyle. Let's say we have two algorithms: one for the assistant and the other for the lights. You enter input through a voice command to switch the lights off. Through algorithms, the assistant asks itself: does the input fit task "TV"? If not, does it fit the task "lights"? When it matches the task "lights," it sends an input to the second algorithm that is related to the lights. In this case, the light collects the input and asks: Does it ask "turn on lights"? Or does it ask "turn off lights"? It then turns off the lights as an output.
There are many languages to communicate with computers. Historically, there was FORTRAN, Pascal, C, Python, Java, Ruby, etc. It goes back to the 19th century when Ada Lovelace and Charles Babbage invented the difference engine to perform calculations. Since then, programming languages have evolved throughout the centuries. While it used to focus on the computer's perspective, modern languages take the user's perspective instead.
We use computers not only by ourselves (for example, in calculating numbers in Microsoft Excel or typing documents on Microsoft Word) but also with other computers (for example, sending emails). A computer communicates with other computers through what is called TCP/IP. TCP/IP is a protocol agreed-upon set of rules. It breaks each message into packets to faster the process. Once received, the packets the other computer reassembles them.
The same logic works for MP3s. Back in the old days, one Compact Disc has 32 MB of storage. Meanwhile, MP3 compresses the song to 3 MB so that it's easier to download and it's easier to store.
David Alayon, Co-founder of Innuba, wrote about the many algorithms that play essential roles in today's world. Some of them are d by Google, Facebook, online dating websites, Amazon, and MP3 players. First, we have Google's PageRank, which classifies and ranks search results online. Second, we have Facebook's EdgeRank that updates Facebook news. There is also an algorithm used in online dating OkCupid and eHarmony. Not to mention the "you can also enjoy" feature on Amazon pages.
Fast Fourier Transform (FFT), Link, Data Compression, Dijkstra, RSA, Proportional Integral Derivative, and Sorting have frequently used algorithms.
There are many challenges today with algorithms, and from biases to privacy issues, algorithms have become our concerns. We see fake news around us, and Facebook tried to leave the job of sorting out real vs. fake news to algorithms. Unfortunately, that didn't work. Not to mention Microsoft's "Tay" chatbot on Twitter, which soon became racist and sexist with its machine learning algorithms. Many people also wonder how wise it is to allow cookies for the internet to trace our digital footprints.
Nielsen Norman Group discusses personalization in customers' experience with machine learning algorithms. Some algorithms we are familiar with would recommend movies, artists, books, Choice of advertisement display, Deals, personalized offers, and One-click access. Based on NN/g's analysis, Netflix did the best job in providing transparency: users understand how their actions would influence the recommendation algorithm ("Because you watched Movie X" or "Because you added Movie Y to your list"). Meanwhile, Facebook and Instagram don't display much information about the Feed, and users assume it is because of their preferences settings. NN/g criticized that the outputs given have issues, mainly because of other people's actions. Outputs on Facebook, Instagram, and Netflix are shown based on the 'relevancy metric.' Yet, the order of items is not predictable, the outputs are sometimes not within the user's interest, and some movies/news are left out.
Some advice on the future of algorithms is to humanize them. We need to increase algorithmic literacy and transparency. In a way, it's like coffee fair trade campaigns where we know how the price of beans helps plantation workers. Likewise, we need to know how data is collected and how the computer's decision-making works. Regarding privacy, we might consider adding 'noise' to our existing data. A result is a different number than the precise data, but it protects the customer's privacy. Additional advice would focus on the ease of control of algorithm outputs, particularly recommendation algorithms.
There are many ways you can grow your business using the right algorithms. Following the Netflix example, we can try to personalize our customer's journey. The idea is to give a unique experience to each customer. As suggested by N&N/g, we must be mindful of disclosing how the algorithm works. By explaining, "Based on your previous search," etc. customer needs a sense of control in determining their decision. Thus, they need to be aware of how personalization is developed.
If you Google some businesses in your area, you can look at their locations, opening hours, and popular times. The latest is a new feature from Google's Geo-tracking. It means that Google utilizes Google Maps to get live data and tracks how many people visit the store. Then Google algorithms visualize the statistics to help you decide when is the best time to visit: during a busy time (for example, if you prefer a crowded bar) or a less busy time (for example, when you go to Starbucks to order a coffee). A similar solution can be useful for your business and customers.
Google also showcases how businesses improve by working together with them. For example, Tajawal, a Middle-Eastern company, uses Google Ads to bring more revenue coming in. Another one is Nescafe which personalizes its product: Dolce Gusto. On the other hand, Apple mentioned how different businesses thrive with their products. AXA Finance, for example, utilizes iPad Pros and works with IBM to create an app to illustrate financing to their customers. IBM also helped healthcare by utilizing Artificial Intelligence in identifying breast cancers. You can use existing tools to leverage your business, such as Google Analytics, Google Ads, and Apple products. But not only that, you can build your algorithms to ease your business processes.
An excellent example of both customer-facing and employee-facing apps would be Uber and Lyft. You can see how they provide real-time information about drivers around customers. When customers order a car, they can look at the map with real-time location, updated remaining time to the location, and even a button to report a driver. It is useful to think comprehensively about algorithms: to provide both tools for customers and employees to increase efficiency.
Algorithms work for your benefit in the Launch, Mature, and Growth stages. In the Launch state, you can focus on sorting algorithms to create relevant searches and listing customers registered to your services. Another algorithm e is Dijkstra, which considers the fastest route in different scenarios. Speaking of tools, you might want to consider Google Ads to let your customers find your business.
Meanwhile, when you are in the Mature stage, you want to give better services to the customers. You can personalize their experience. Another way is to create helpful apps for them to use. For example, banks today allow check deposits through picture uploads taken from the camera from the customer's mobile phone. You can also upgrade the app to show nearby store locations. Let's say you just came out from a doctor's visit. After the visit, you next need to buy prescribed medications. You can create an app using Augmented Reality to show the nearest drugstore that sells the prescribed medicines.
In the Growth stage, perhaps you want to explore different fields using new algorithms. The idea is to expand your business to a larger scale and make the algorithms work for you. At this stage, you already have historical data on your customers. The next thing you want to do is to target a demographic for your next experiment.
We can look at the Tiktok company under ByteDance. It is currently on the rise as it targets the younger generation: Gen Z. However, other generations started to adopt it, especially with the Stay-at-Home situation. Likewise, maybe you want to expand your business to an e-commerce store. You realize that most electronic shoppers are men, and most fashion shoppers are women. You can study their behavior better while tweaking a few things in the shopping experience.
It is essential to keep growing as a company. Do you remember how Nokia lost to Blackberry? which then lost to Apple's iPhone, which now competes with Samsung. Nokia was leading the market and forgot to innovate when Blackberry came in. On the other hand, Blackberry didn't want to invest in touch screens when the iPhone came in. Even nowadays, tech giants compete with each other through innovations that don't limit them from copying one another. Facebook owns Instagram, which adopted Snapchat's feature and introduced it as InstaStory. Facebook also upgraded its video option on Facebook messenger to become a video conferencing tool, like Zoom and Google Hangouts. If you don't grow, competitors can copy your features and beat you in the market.
A quick tip would be to have a Growth Lab, Innovation Team, or Experimentation Division. Their job would be to hypothesize, experiment and deploy tests to innovate your business. The team can have dedicated app developers, dedicated UX designers, dedicated marketing strategists, and dedicated operations. It is crucial to build the right infrastructure for your innovation and growth.
We first need to know that today's technologies are getting cheaper and faster. When we are f economic downturns, we can still invest in technology to better our business performance. You can rent computers to make algorithms work for you.
Think about implementing algorithms on different platforms. You can look at your product's user experience, design strategy, prototyping, and other processes. You can also hire consultants to work together with you.
Computers work best on repetitive tasks. They do calculations fast and accurately. Therefore, you can utilize algorithms to streamline your business and focus on what matters. For example, Link algorithms help with your SEO and how customers find your business. When that process is automated, you can focus on content marketing: publishing high-quality blog posts for your customers.
Algorithms build the world we live in today. We can make algorithms work for us to better our customer's experience. Using existing tools, such as Google Analytics, Google Ads, and more is best. In your business. But more importantly, you can also build your algorithms. You can personalize the customer's experience while interacting with your website and apps.
At Designial, we provide different services, such as Digital Transformation and User research and testing. Talk to us.