Table of Contents

    There are some insights that can only come from the data you produce or collect in-house. Historical sales figures, for example. Foot traffic through your store. Sign-ups and downloads of your app. 

    But no matter how complete that data, no matter how intelligent your prediction models, you don’t exist in a vacuum. You can’t collect every single relevant piece of information yourself. You need to know what’s happening outside your organization if you want to compete and survive. 

    In other words, you need to think about how to incorporate external data to improve your processes, reduce risk, and make smarter decisions. This is particularly true for smaller organizations, who may struggle to generate or access large enough data pools of their own to really understand where their business or the market is headed.

    Let’s take a look at some common alternative data use cases and applications whereby SMBs use external data to cut costs, drive up revenues, or access new markets. Case Study

    Alternative Data Use Cases

    Machine learning in financial services

    External, alternative data sources are extremely useful to the financial sector, especially when it comes to assessing a loan applicant’s suitability and reliability for credit, calculating the risk of lending, anti-money-laundering efforts, and detecting suspicious behavior. 

    That’s partly because alternative data provides applicants who have no credit history with a means of generating a digital footprint. Around one in 10 adults in the U.S. have no rating with the main credit bureaus, meaning they can struggle to get credit despite doing nothing wrong. Allowing these people to share their payment histories (typically their utility bills and phone/internet payments) helps reliable people that fully deserve credit to get it. 

    This is great news for smaller financial services companies or any SMBs that need to run credit checks on new customers before they do business with them. You can guide these customers towards the best products and services for them without worrying that your time will be wasted if they simply don’t have enough information out there to pass a conventional check.

    Machine learning in manufacturing

    With so many moving parts to worry about — both figuratively and literally — it’s little wonder that manufacturing processes and companies stand to benefit so much from data-driven machine learning strategies and alternative data use cases. 

    According to McKinsey, process manufacturers who use heavy-duty assets and equipment are finding ways to reduce energy consumption, boost throughput, increase yield rates, boost hourly profits, and automate certain tasks to allow systems to run on autopilot. This comes from monitoring and syncing streams from sensors and performance data, but also external feeds especially supply chain partners.

    Collecting external data also makes it possible to generate insights on who buys your products and when. You can also analyze customer feedback on products that have already been released into the world, as well as perceived gaps in the market. That includes invaluable information such as social media posts and online reviews, which would have taken far too long to scour in the past. With the right technology, though, you can glean insights swiftly, informing the next round of product development.

    AI in retail

    Consumer brands and retail outlets have been some of the most enthusiastic converts to AI and ML innovations and have the alternative data use cases to prove it. These technologies are used to anticipate demand, tailor promotions, understand consumer behavior, drive sales, and segment campaigns with impressive precision. 

    These approaches aren’t limited to behind-the-scenes, they also take place in-store. For example, at Walmart, AI robots scan shelves looking for items that are missing or need to be restocked, as well as price tags to be changed. Sephora employees now wield handheld scanners that use machine learning algorithms to match each customer’s skin tone with a huge, nuanced library of makeup colors. Activities like that boost operational efficiency, improve customer service and strengthen customer loyalty.

    That said, when it comes to performing truly effective retail data analysis, you may not be able to get all the information you need from within the company, or your own stores. For example, those Walmart robots can alert you when you’re already running low on umbrellas on the shelves, but they can’t predict when you might have a sudden rush on these that would decimate your stock. By incorporating external weather data, however, you would be able to predict that — allowing you to optimize hyper-local stock reordering and replenishment.

    Product personalization

    This is an increasingly popular way to enhance user experience — and with good reason. When you open up Spotify or Netflix, you get an extensive and often eerily accurate set of recommendations for songs or shows you might like, all driven by machine learning algorithms. 

    In Spotify’s case, the app even compiles multiple sets of recommended playlists daily, based around trends in your listening activity. Analyzing consumer and behavioral data to these ends is a highly effective way to keep users and customers engaged. 

    But Spotify doesn’t rely solely on internally-generated data to figure out which bands to recommend to you. It also crawls through millions of items of text on the internet, including blog posts and other web pages, looking for references to bands, artists, and particular songs. 

    Using natural language processing (NLP) and other forms of machine learning, Spotify notes instances of bands being referenced alongside each other and tries to figure out what people are saying about them. This information informs its recommendation algorithms, helping users discover new artists and keeping them hooked on the site.

    You may be thinking: this is all very well for a huge company like Spotify, but how can a smaller business compete? 

    In fact, that’s one of the great things about using an augmented data discovery tool for external data. You don’t have to rely on your internal capacity for data collection and processing to tap into these insights. You just need to find a tool that can provide data that will help you better understand your customers. You don’t have to do all of this yourself.

    Final thoughts: choosing the right tech

    As we’ve seen, the businesses that take the leap and innovate with the most illuminating data streams reap rewards in every industry. The key is to seek out the information that best answers your questions and solves your problems, wherever that might be. There is no reason to limit yourself to what you can collect and analyze yourself. 

    At the same time, you’re only as effective as the technology you have at your disposal. In a small or medium-sized business, you are highly unlikely to have the resources you need to pursue strategic external data projects yourself, entirely from scratch. Focus on finding the ideal platform that will automate as much of this as possible. One that will handle issues like consistency, scalability, and old or out-of-order data so you don’t have to. 

    That way, you can spend your time figuring out how to make the most of the data, turning your insights into meaningful action. That’s how you’ll really boost your bottom line. Case Study