For many years, marketers, consultants and ecommerce specialists have used web analytics to make business decisions. A particular product page received 10% more pageviews month-on-month, therefore an investment should be made in pushing that product. A recent email campaign had a higher than average open rate, so produce more of those type of emails.
But is this base of data enough? We’ve noticed that there’s both an increased demand, and an increased opportunity, for data science across all of our key industries. Web analytics was once enough to help improve performance, but we’ve recognised a need to expand the scope of data being analysed and take a more scientific look at client data. Using data science we can address gaps in performance, predict trends and future outcomes and ultimately help clients stay ahead of their competitors.
Better data, better answers
Let’s say you’ve noticed a sudden and unexpected drop in web traffic from a given source. Traditional web analytics still has a place here, and likely provides useful information such as pages/products seeing the highest drop in traffic or how traffic behaved following entrance to the site. But what other sources can we look at to understand why traffic has decreased? For a fully-rounded view of what’s impacting your website, and your business as a whole, we need to look at a wider range of data sources. The answers could come from anything from a general downturn in the market or sentiment changes in social media, to adverse weather conditions or changes in competitor pricing. Not only that, but traditional web analytics usually covers more descriptive analysis rather than prescriptive or predictive. How can we use pattern recognition and data modelling to predict downturns in traffic or certain behaviours in future? In other words rather than reporting on what has happened, can we execute tactics based on what is likely to happen?
Data science has been adopted by many companies around the world, and for good reason: it helps them to make better business decisions. From making inventory decision around expiring stock to considering new logistics plans, advanced analytics can help you.
Who’s doing it well?
Uber Eats uses advanced analytics techniques to estimate delivery times for customer orders, a multi-stage process, which requires several data sources to make a reliable estimate. Uber Eat’s data scientists use gradient-boosted decision tree regression models to predict delivery times, taking historical and real-time data into account. Combining several data points allows the app to estimate delivery times far more accurately than relying on just one data source.
Starbucks is another big-name company using data science to improve the customer experience. The coffee giant’s Digital Flywheel Program uses AI technology and data from its Starbucks Rewards members’ accounts to make food and drink recommendations to both new and old customers based on factors including order history, current weather conditions, day of the week and time of the day.
What’s more, when expanding to sell products in supermarkets, Starbucks used data not only from their own app and in-store orders, but also from market intelligence about how people consume tea and coffee at home. By combining their own data about customer orders with wider industry data, Starbucks have been able to make smart decisions about how to expand to a new market.
Unlock more insights with data science
Whatever vertical you operate in, advanced analytics allows previously unseen patterns to emerge from your data and from this your business can project future trends and customer behaviours. It can help you to future-proof your strategy and move ahead with more confidence.
By layering traditional web analytics with a more advanced statistical approach, we can better analyse behaviour from different perspectives. We no longer want to only provide our clients with standard descriptive analysis, e.g. highest revenue-driving products, or cart abandonment rates . We want to be able to identify every potential metric which contributed to a particular product driving the highest sales, and determine to what extent each factor contributed to that outcome. Our data work helps clients improve commercial performance and minimise risk, by predicting and planning for highs and lows.
At its core, data science and advanced analytics allows us to gain a richer understanding of their business, be it supply chain, distribution, payment systems or sales. Armed with that information, we can help our clients to do what they do, better.
Get in touch today to find out how we can help you do more with your data.
