Improve Your Business by Leading the Transformation to Prescriptive Analytics
Forecasting what's coming is no longer enough. Modern enterprises must use prescriptive analytics to ensure users of all skill sets are able to benefit from a data-driven culture.
- By Guy Yehiav
- June 20, 2016
For a business, knowing every important detail about the pulse of the company in real time is powerful. The ability to prescribe solutions from the data and take action gives businesses greater control over outcomes and consistency; in some cases, these actions can change the future.
Imagine using this concept throughout your business to drive real value. This is prescriptive analytics, and IT departments should encourage business users to transform from predictive business intelligence or business analytics platforms to prescriptive analytics.
In a nutshell, predictive analytics looks at the past to forecast the future. It uses techniques such as regression, Bayesian techniques, and predictive modeling to anticipate what will happen next. Prescriptive analytics looks at the data and then advises what an enterprise should do in order to meet a particular goal. Simulation, machine learning, and complex event processing (CEP) are common prescriptive techniques.
Predictive analytics engages in analyzing past events, building models, and forecasting the future. For more advanced users, predictive analytics may help determine how to prepare for future actions, but this requires a user who deeply understands how to take raw data and execute on piles of metrics and also analyze data collected from predictive machines. Where most enterprises only have a few data scientists on staff, it just isn't realistic for those users to analyze all of the data for the entire company.
For these companies, especially those in less technical industries (for example, a retailer would be less likely to have access to a data scientist than a financial firm would), transforming to prescriptive analytics can help glean insights that may have otherwise gone unnoticed. Prescriptive analytics has the power to arm businesses of all sizes and skill sets with the ability to look at real, current statuses and take immediate action to hit the goal KPIs.
Here's an example of how shifting from predictive to prescriptive analytics can create value for an enterprise. Many retailers use predictive analytics models to produce reports on the state of the business. A grocer looking to understand the on-shelf availability of a product often works from a report that shows how much product is on hand, the amount sold that hour/day/week/month, and total revenue.
This information helps a retailer determine when it is time to order more of the product. However, if the retailer sold more product than expected, the shelf may be empty while there is still stock packed away in the back.
The predictive report will not alert managers that the product is not visible to consumers; it would typically require a highly technical person to draw this conclusion from the daily report. This means the grocer finds out that they need to replenish the product from an employee walking by (or worse, from a shopper). This could result in lost sales.
A prescriptive solution, however, will notice the sales trend before the shelf is empty and will alert managers via email, SMS text, or another preferred method of communication. This is where prescriptive analytics delivers value: by taking the guesswork out of predictive analytics and putting findings into clear language in real time.
Keeping this in mind, here are a few top benefits of prescriptive analytics for every part of a business:
- Information technology: Power up data investments by quickly loading any type of data, examining it to find patterns of behavior and identifying the root cause of anomalies.
- Asset protection: Get early detection of fraud and compliance using predictive indicators, including returns, voids, gift cards, no sales, discounts, coupons, and loyalty card activity. Look for employee, customer, and vendor fraudulent activities or compliance and training opportunities.
- Marketing: Analyze customer and promotional activity to identify opportunities originating from changes in customer behavior, maximizing a customer's basket, and the execution and effectiveness of promotions.
- Finance: Identify and act on any opportunity that will increase same-store sales, margin improvement, inventory control, and total revenue.
- Merchandising: Identify upstream opportunities with products, sales reports, allocation, on-shelf availability, vendor negotiation and basket analysis.
- Operations: Make store or warehouse visits more targeted, improve sales, increase margin with directed actions based on specific predictive indicators, and improve store audits and communication.
- Logistics: Identify and act on productivity improvements at distribution centers as well as in the delivery process.
- Planning and buying: Gain visibility and actions for high-priority vendors, stores, and products that impact margin performance based on KPIs and fine-tune forecasts and allocations.
eCommerce: Link stores with online buyer behavior, create seamless integration. Move aged inventory by targeting specific consumers, leverage stores as originating shipment centers without overloading labor.
Supply chain: Identify the root cause for on-shelf availability, late delivery, and on-time shipment-complete results. Leverage logistics employees by guided actions to improve KPIs.
Clearly, prescriptive analytics can make every department data-driven and therefore democratize analytics. IT leaders should take charge to inform users about the larger benefits of applying prescriptive data, such as no longer needing to rely on your gut or falling behind competitors who have the resources for an in-house data scientist.
Based on the needs of your business and the level of business user skills within your organization, you can also consider combining predictive and prescriptive analytics to identify, resolve, and measure opportunities for improvement with actionable prescriptions and insights delivered to the right person at the right time.
Once prescriptive analytics processes become part of a business culture, users are empowered to justify their actions with data, not feelings. This results in happier employees, better-informed partners and ultimately, more satisfied customers.
Guy Yehiav is CEO and chairman of the board at Profitect. Prior to Profitect, he served as vice president sales and strategy for Oracle’s Value Chain Planning Solutions where he was responsible for sales, strategy, and customer success. Guy was also founder of Demantra US, a leading global provider of demand-driven planning solutions, which was acquired by Oracle in 2006.