According to a report by the Columbia Business School, while 100 percent of CMOs surveyed believe that successful brands use customer data to drive marketing decisions, 39 percent of marketers report that they cannot turn their data into actionable insights, and 36 percent ofmarketers report that they have “lots of customer data”, but “don’t know what to do with it.”
Diving into the Big Data pool can be daunting: you may have heard that implementations can take years and require major IT investments and hard-to-find Data Scientists. Here are some ways you can get started now with the data you have:
- Product Recommendations: Everyone is familiar with Amazon’s “people who bought this also bought…” recommendations, but do they really work? Yes! In a recent study, web pages with product recommendations were shown to increase revenues by 50% over the same pages without product recommendations. Use your order data to determine these same “people who bought this product also bought…” connections between your products, and then merchandise related products with every product in your inventory. If you don’t have any related products for some of your products, use best sellers in the same product category.
- Personalization: Merchandising products that are relevant to your customer is a great way to create a personalized experience. Use the data you have about your customers to determine which of your products might be the most relevant to them. A good place to start is to merchandise a variety of the related products determined in #1 for any products customers have expressed an interest in, such as viewing a product detail page, liking a product, adding a product to a wishlist or cart, or purchasing a product. Use these personalized recommendations on your site when you recognize a customer and/or in your communications with your customers. You can also use best selling items in product categories that your customer has expressed an interest in if you don’t have many related products. Mix it up – show them different recommendations each time to keep them interested.
- Proactively Deal With Potential Problems: Identify “negative outcomes” that end up reducing the profitability of a customer or order, such as returns, refunds, lost by carrier claims, complaints, and fraud. For each negative outcome, examine all the data you have for the orders or customers that had that outcome, with the goal of finding common elements between them. This can be done programmatically using similarity analytics, but if your data isn’t huge, you can still accomplish a lot with the best pattern detection tool even invented – your brain! Once you have identified which factors are correlated with these negative outcomes, then use proactive outreach for orders or customers with similar characteristics.
- Get More Great Customers: We’re going to do the same exercise as in #3, but this time for positive outcomes, such as newsletter signups, click-throughs from your marketing emails, repeat purchases, high value customers, and highly satisfied customers. Once you have identified the factors that are correlated with these positive outcomes, you can reach out to new customers with similar characteristics with offers that encourage them to take the next step in deepening their loyalty and engagement. Identify your VIPs before they become VIPs, and treat them accordingly!
When you are able to identify potential positive or negative outcomes in real-time, and automate the appropriate responses, such as queuing potentially fraudulent orders for review before credit card processing, or sending a particularly attractive offer to potential new VIPs, you’re getting real business value.
The human brain is great at pattern matching, but can easily get overwhelmed looking for patterns in mountains of data. When you’ve got millions of products, orders, customers, or data points, it’s time to leverage purpose-built tools to accomplish these goals. If you’re in this position, a big data analytics or BI project will likely have a clear and positive ROI, and it is worth your time to investigate which tools and what approach will fit in best with your culture and systems.