Why You Need To Start Using Predictive Analytics ASAP
Predictive analytics. What a dry, geeky term. Yet it is igniting a true revolution in consumer marketing. You’ve personally already experienced it firsthand when you’re shopping on Amazon for a printer cartridge, and it suggests you might also need a ream of paper. Or if you use a travel app like TripAdvisor to book that dream trip to Hawaii, and you may be told you’d like Bora Bora, too. Or, enjoy listening to Beyonce on Last.fm? You could be offered a recording of a poetry reading by Warson Shire. All these things are done using predictive analytics.
The reason for predictive analytics: Personalization drives sales. Recent research shows that if you push special offers based on your customers’ personal preferences, customer engagement increases fourfold. If you use predictive analytics to send the offer at exactly the right time — at the point in time when your customer is considering a purchase, for example — you could achieve sevenfold increase in customer retention.
These are compelling numbers, which give you a taste of what you could do for your business. In this blog, we’ll explain what predictive analytics is, how it works, and why you should deploy it ASAP.
Predictive analytics: a primer
Predictive analytics is a branch of analytics where known past events are used to make predictions about unknown future events. Whereas using traditional analytics helps you understand what happened yesterday, predictive analytics tells you what will probably happen tomorrow.
In business, this involves finding patterns in operational and transactional data — last month’s overall sales, how many of each dress SKU you sold last week, how many people abandoned their digital shopping carts — to help you anticipate who will buy what, and when. It does this down to the individual consumer level using techniques such as data mining, statistics, modeling, machine learning, and artificial intelligence (AI).
According to the Aberdeen Group, companies that use predictive analytics are twice as likely to identify high-value customers, and to market the right offers to them. They’re also more likely to see patterns that would otherwise escape them. For example, when Arby’s tracked sales at individual stores around the country and analyzed what factors led to more sales, it turned out that renovated stores sold the most food. Arby’s then used that knowledge to accelerate store remodels fivefold over the next 12 months.
Not only does the personalization that comes from predictive analytics please customers — and encourage them to buy more — it can save you considerable capital. After all, if you can accurately forecast what your customers are likely to want next month, you won’t have excess inventory of one product and go out of stock of another — both extremely costly mistakes to make.
The data’s the thing
Once you start using predictive analytics, there’s virtually no limit to what you can find out — and improve — about your customer relationships, and your business.
The customer data needed for successful predictive analytics is readily available if you’re an online retailer. After all, you know how much time customers spend inside your store, down to the second. You can collect detailed information on what customers search for and browse through, not just what they ultimately purchase. By doing what-if analyses, you can play with prices — and learn over time which customers respond to financial incentives, and which are more brand loyal.
Brick-and-mortar retailers have it a little harder, but they’re beginning to be cleverer about collecting data — using smartphone app data to log when customers enter their stores, what kind of products they look at, and whether they actually purchase anything. When this kind of data is fed into a machine-learning predictive analytics model, retailers get rich insights that allow for much smarter pricing and merchandising decisions that would be made using traditional demographic-based customer segmentation.
A new technology that is helping traditional retailers is RAIN RFID. Retailers are swiftly moving from barcodes to this new way of tracking products and gathering data.
But the issue isn’t just about collecting data. You also need to use it. If you’re like other small and medium-sized businesses (SMBs), your customer data is fragmented among your point-of-sale (POS) systems, customer relationship management (CRM) software, enterprise resources management (ERP) solutions, customer service emails, and social media. Getting all this data into one place so you can analyze it is a challenge.
Another issue facing SMB retailers is that few — if any — have the skills in-house to attempt predictive analytics on their own. Even the largest companies are having trouble finding data scientists.
Happily, vendors known for selling predictive analytics tools to enterprises are beginning to offer cloud-based versions for SMB customers:
- Salesforce: offers Marketing Cloud, which is a customer experience platform that ties together its flagship customer relationship management solution with predictive analytics tools to help SMBs discover new segments, identify the likeliest consumers to engage, and make recommendations on how to engage them.
- Canopy Labs: focused on predictive analytics specifically to improve the customer experience. It provides cloud-based tools to help retailers analyze, segment, and predict any part of the customer journey.
- InsightSquared: gives you predictive analytics tools for “sales intelligence” that use your entire sales history and display multiple forecast models in intuitive visualizations.
- Stitch Labs: deploys predictive analytics to help with inventory management that helps SMB retailers gain inventory visibility and control across all retail and wholesale channels, systems, and fulfillment locations.
Many more start-ups are jumping into the market to make predictive analytics simpler for smaller organizations. Even those behemoths of the analytics world, IBM Watson and SAS are setting their sights on SMBs and offering cloud-based packaged solutions that are easy for non-experts to use.
A study by the Harvard Business Review found that only 20% of companies act on most of the data they collect. Since the data builds up day by day, the implication is that every time you open your physical or virtual doors, you know less about your customers than you did the previous day. Predictive analytics would help with that.
To win in today’s competitive retail space, it’s essential that you not only collect, but use your customer data to get actionable insights. This means adding predictive analytics to your suite of standard business management tools ASAP.