Sidecar today launched its latest machine learning capability for Google Shopping, Revenue Probability Scoring, to improve the way retailers drive sales in the channel.
With Revenue Probability Scoring, Sidecar takes a holistic look at an entire Google Shopping campaign, and identifies and invests in the next set of opportunities that will maximize campaign revenue at the retailer’s ROI goal.
“To maximize revenue and ROI in Google Shopping, it’s not just about bid management—it’s about investment management,” said Dave LeDonne, Director of Product for Sidecar. “Many marketers spend too little when a revenue opportunity arises, and too much when no opportunity exists, because they have no way to gain insight into those opportunities. There comes a point where marketers can’t be sure what investment will drive their future revenue.”
That’s what Sidecar’s Revenue Probability Scoring solves.
How Revenue Probability Scoring Works
Sidecar uses machine learning to emulate the way a marketer builds expertise in Google Shopping management, but with far more learning scenarios than a marketer could ever experience on his own.
Sidecar is constantly asking questions about every product to understand its revenue potential. What is its price? Is it discounted? What is the repurchase rate, and is it changing? What products are most similar to it? Is its brand popular across the catalog?
Rapidly and exhaustively, Sidecar identifies patterns in the relationships among millions of data points to develop a Revenue Probability Score for each product. The score indicates how much potential a product has to generate more revenue. Sidecar fuels every opportunity each day with the level of spend that will allow the campaign as a whole to achieve maximum revenue at the retailer’s ROI goal.
No Product Is Off the Table
Sidecar is not constrained by just historical product performance data, nor is it limited by intuition or if-then logic. Rather, it is reliant on data science, and inputs from the retailer’s site traffic, product performance goals, products attributes, consumer search behavior, and PLA channel trends.
This approach lets Sidecar gain a complete picture of every product, and give deserving products a chance to generate revenue, even if they have been overlooked by a marketer, gotten insufficient traffic, or created little revenue in the past.
It also allows Sidecar to build learning models for catalogs large and small, including those which have little or no Google Shopping data, because the technology learns about products using multiple, relevant data sources.
And Sidecar’s approach does not sacrifice revenue for the sake of maintaining ROI. Instead, Sidecar understands the tipping point of every investment, and avoids under-investing and over-investing.
Success in Google Shopping—like any product advertising channel—is about not just thinking smart or thinking differently. It’s about thinking smart and differently. With the ever-growing sophistication of Google Shopping—and the retailers advertising on it—marketers need new and intelligent approaches to stay ahead.