Neal Analytics is leading the way with machine learning and predictive analytics with the launch of its BidProfit and SmartStock solutions. Built on Microsoft Azure and Microsoft Azure Machine Learning, these two solutions are designed to enhance bottom lines when used with Decision Profit modeling from Neal Analytics.
Decision Profit modeling methodology enables Neal Analytics to take data from silos to solutions by federating internal data sets and then enriching the data sets with external sales-driving data. After data engineering, Neal Analytics conducts basic modeling and analysis and seek customer feedback to ensure it captures the key influencing factors driving business. This lets Neal Analytics refine modeling to the needs of each client’s business. Neal Analytics also automates solutions by blending model results into tools such as Excel, PowerPivot, and PowerView—allowing customers to focus on driving their bottom line.
BidProfit is a search engine optimization solution that interfaces with Google and Bing to predict bidding points for search results. David Brown, Neal Analytics Solutions Sales Director explains, “Initially, this solution was developed in conjunction with our work at a large travel company. We were predicting the bid position for their long tail of keywords. In the beginning phases of this project, we achieved success in spite of the scarcity of data of this set of keywords. Our model was able to predict correctly 33% of the time, which resulted in a 140% increase in traffic with only a 12% increase in cost. The bottom-line impact for new customer acquisitions was estimated at $40 million. Since our initial implementation, we’ve been working with the Microsoft Azure Marketing team to implement this solution for them. We’ve received very good initial results, with correct predictions occurring 66+% of the time, which represents a 100% model improvement in terms of performance.”
SmartStock is a demand prediction solution which optimizes stocking by predicting localized demand. “Our initial implementation of this solution was in a large cellular phone retailer, in conjunction with the launch of the iPhone 5,” continues Brown. “They had historical data on previous iPhone product launches, so we used this data in combination with local social media data to create predictions of demand for iPhones. These predictions had outstanding real-world results. Stock-outs were reduced by 70% and excess inventory on-hand decreased by 73%. These results represented an annual run rate improvement of $19M.” Since this implementation, the Microsoft Data Insights Incubation program has picked up SmartStock. Through this program, Neal Analytics has access to Microsoft resources and several modules which allow SmartStock to directly connect to ERP systems. This allows Neal Analytics to quickly and easily integrate customers’ data for SmartStock evaluations.
The key with both SmartStock and BidProfit is Neal’s unique capability of federating internal data—finding sales driving external data and collaborative model development with decision makers. With SmartStock, Neal Analytics creates a decision model based on sales drivers (weather, social media trends, or other key factors relevant to business). With BidProfit, Neal Analytics creates a decision model based around Bing and Google data, and incorporates real-time feedback data based on .NET development. Both SmartStock and BidProfit represent Neal Analytics’ first offerings based on Decision Profit Models™ with federated data sourcing and full project engagement methodology.
For more information, email firstname.lastname@example.org.