How Alternative Data Can Help Predict Market Behavior

Corporates Are Leveraging on External Data to Gain a Digital and Competitive Edge

Over the years, companies have increasingly relied on integrating alternative data sets to add real-world context to internal business decisions. Companies can more accurately identify their competitors and predict their market behavior. They can also extract deeper consumer insights to better understand their users’ shopping behavior and to see whether to develop their products further or expand their sales into new markets.

92% of data analytics professionals said their companies needed to increase the use of external data sources*.

Consider Foursquare, the New York firm which popularized the concept of real-time location-sharing and checking-in. Foursquare famously predicted that Chipotle same-store sales would fall 29 per cent after the Mexican chain was hit with E. coli outbreaks, based upon check-ins on their application. The actual decline ended up being a spot-on 30 percent. Given the high degree of accuracy, it is not surprising that Foursquare announced Foursquare Analytics, a foot traffic dashboard for brands and retailers in 2017. 

Alternative data comes in a variety of sources and can be structured or unstructured. Web traffic or logistics data, social media data, satellite data, email and credit card data are just a few sources of alternative data used to supplement traditional data sources when drawing insights for predictive analytics.

According to Mckinsey, retailers exploiting data analytics at scale across their organizations could increase their operating margins by more than 60 percent.** In this instance, email transaction data is particularly useful as it exhibits a high correlation to actual sales, and is therefore accurate for tracking and predicting retail sales.

Alternative data often becomes the most useful when they are processed with data science techniques and data scientists need to determine which data have the potential to be useful. However, incorporating alternative data into business decision making is not without its challenges. Incomplete datasets, poor quality, limited volume, unverified data and a lack of historical data for back-testing are all headwinds faced by companies looking to leverage on alternative data to garner more insights.

Most companies already have plenty of customer data but don’t tie it together to create a richer picture of their consumers. The real value of alternative data only arises when it is combined with ‘traditional’ data and then coupled with scalable analytics prowess. As observed by McKinsey, the most fruitful insights comes from combining transaction data (such as purchase amounts over time), browsing data (including mobile), and customer service data (such as returns by region). Yet, up until recently, combining and coupling has been a complex and time-consuming process.

That said, the improvement in artificial intelligence over the years provides the ideal opportunity for data analysts to gather essential information from several different sources and allows an integration with a cohesive application.

As far as the credibility of the alternative dataset is concerned, email receipts are pretty much accurate in tracking and predicting retail revenues. 

Measurable AI is the market leader when it comes to alternative data such as transactional email receipts for emerging markets across South East Asia, Middle East and Latin America. Email transaction data accurately tracks retail sales and companies with a data analytics team are exploiting this alternative data source to gain deeper insights into their user base and use it to predict purchase behaviour amongst different demographics. 

Take food delivery companies for example, one of Measurable AI’s main client segments. From a business intelligence angle, food delivery companies are using transactional email data to better understand what products users are buying most, what hours are the most popular, and the average basket size. From email receipts, they can also gain insight as to how much of an impact does offering discounts or free delivery have on the frequency and volume of orders. From this, they can better predict the behaviour of their user base and also see what percentage of their user base also order from their competitors, along with new investment opportunities for expansion. 

Having such level of granularity in data provided in e-receipts can also help corporations make more informed decisions related to predictive discounting, inventory management and personalized campaigns.

“What we learn from the survey data vs. the consumer panel data are sometimes very different. Consumers may say they love something when asked; but in reality their purchasing behavior indicated by their transactional emails reveals that they actually don’t shop for that item a lot, or probably never. Recognizing this discrepancy we decided to rely on actual transactional email data. This is way more credible to understanding consumers spending patterns better: what they actually love buying, how much they spend, and who else are they buying from.”

Data Engineer, International E-Commerce Company

From the global pandemic to retail apocalypses to looming economic downturns, there are a plethora of factors that affect corporations and retailers alike. This is why simply guessing or hoping for the best is not a viable solution for smart corporate strategy and can result in costly mistakes. Instead, more data-driven corporates are choosing to leverage on alternative datasets and engage technologies like artificial intelligence to ensure that their business decisions are smartly backed by real data and are therefore more accurate.

Measurable AI is the leader with the largest consumer panel for transactional email receipt data for emerging markets. We are here to help corporations make smarter business decisions on the back of our unique dataset. Contact us at [email protected] for more information or to request a trial of our data.

*Forrester, “Digital is driving the next generation of data marketplaces,” December 2017. View in article.

** McKinsey and Company. Big Data, Analytics, and the Future of Marketing and Sales, March 2015.