Spend any time listening to vendors of B2B pricing analytics solutions, and their pitches can be compelling. Readily peppered with phrases such as “scientific analysis,” “big data,” and “profit optimization,” they paint an alluring picture of growing revenues and accelerating profitability.
In short, the prospect they hold out is that by better understanding how prices are set, and how customers react to those prices, companies can not only extract higher prices and profits from existing customers, but pitch for business from new customers at price levels that are most likely to result in a profitable sale.
Put another way, by analyzing the myriad pricing decisions that a typical Fortune 500 company makes in a given month or quarter, pricing analytics can help spot:
• Sales representatives who are pitching prices too low, in order to meet internal target or bonus levels;
• Customers who receive prices and discount levels that aren’t appropriate for the volume of sales they bring to the business;
• Products that seem mispriced in relation to other products which offer a similar feature set or benefit proposition.
While the goals are clear, the paths to reaching them vary. The trouble is that when you speak to five different vendors of pricing analytics, you’ll hear five different approaches to pricing analytics. What’s more, each can provide detailed analyses as to why the approaches adopted by the rest of the market are mistaken.
“One of the issues with pricing analytics is that there’s no convergence in the solutions, such as we see happening with solutions such as ERP,” says Stephan Liozu, a pricing consultant and co-author of Innovation in Pricing: Contemporary Theories and Best Practices. “Vendors take very different approaches, with no agreement as to how to perform certain activities.”
So how is it done? Look carefully, and you’ll see the following three broad approaches to pricing analytics.
1. Eliminate deviant behavior.
Simply put, this approach looks for data outliers—outliers among transactions, products, customers, sales representatives and market segments.
Vendavo, for instance, has a “playbook” of recommended analyses for each customer vertical industry, each testing potential hypothetical opportunities, based on spotting and correcting deviant behavior.
Edward Gorenshteyn, Vendavo product marketing director, says the company has between 20 and 30 ready-to-run analyses for each vertical industry. “It’s about looking for things such as low margin by transaction, low margin by customer, price achieved by sales representatives, and excessive price variation across customers—which can point to instances where customers have pushed for a better deal, playing off the competition, and the sales rep hasn’t pushed back enough,” Gorenshteyn says.
Performance against commitment is another useful analysis to run, adds Gorenshteyn. “It’s the ‘big hat, no cattle’ syndrome,” he says. “It’s customers who aggressively negotiate a volume discount, but then don’t follow through with the volume—and so get lower prices, without the volume to merit those prices.”
2. Understand price elasticity.
To economists, the price elasticity of demand is the extent to which price affects demand: if you put the price up, by how much does demand shrink?
Gasoline, bread and beer might be thought of as having low price elasticity; foreign vacations, jewelry and sports cars as having high price elasticity.
And to advocates of price elasticity-based pricing analytics, the goal is to predict the price at which revenues and margins are maximized—in other words, if you increase the price any further, then overall sales, and profits, will fall.
“Looking backwards at past transactions, searching for outliers, only gets you so far,” says Barrett Thompson, general manager of pricing excellence solutions at Zilliant. “You can easily spot the big outliers—but how do you spot the small outliers? These are just part of the data. We believe that it’s better to make a lot of small changes to prices, going forward, and have those small changes add up to a big impact.”
How do you do that? It’s a three-stage process, says Thompson.
First, perform price segmentation, in order to identify those factors that are most influential in determining prices. Then calculate the price elasticity for all those combinations of price segments—which could be in the hundreds of thousands. Then search through the range of all possible pricing outcomes to select the one that maximizes profit and revenue.
Finally, says Thompson, the power of pricing elasticity-based analytics is that it can also be used to determine prices that satisfy multiple criteria—such as balancing revenue growth and profitability.
“Here, the idea is to answer questions such as: ‘What price will maximize profitability, while also delivering revenue growth of at least 2 percent?’” he says.
3. Segment the market.
According to adherents such as PROS, market segmentation is about using data science to capture and analyze all the attributes of a given transaction—customer attributes, product attributes, and transaction attributes.
For large companies, the datasets involved in capturing these attributes typically extend into the terabyte range—and at PROS’ large airline customers such as Lufthansa, multiple terabytes, says Patrick Schneidau, vice-president of product management and marketing at PROS.
Pricing analytics that emphasize market segmentation look for signals from customers, for example. “When a customer calls and wants overnight delivery, they’re probably willing to pay more,” says Schneidau. “Now, that might sound obvious, but most sales and marketing organizations—and their systems—aren’t set up to price products that way.”
Location is a commonly-overlooked pricing attribute, he adds. At an unnamed delivery service company customer—think FedEx, UPS, or similar—analysis showed that a major determinant of price was the coverage of a given location by the company’s competitors.
Now, ZIP code by ZIP code, the pricing strategy reflects this: where competition is extensive, default prices are lower than in ZIP codes with limited competition.
Different Approaches Each Have Their Place
So is there one “right” answer in deploying pricing analytics for competitive advantage? Probably not, say experts.
“If you’re looking for quick wins, then looking for outliers makes sense, because outliers are deviant behavior,” says Liozu. “But in a $2 billion business, individual outliers are just statistical noise.”
Equally, in a B2B environment—as opposed to B2C—then segmentation and price elasticity-based approaches don’t always work well, refutes Vendavo’s Gorenshteyn.
The problem: derived demand, where the demand for a product is based on the amount required by an end-customer, and not the price at which it is on offer. If you’re a fuel pump manufacturer in a supply chain that culminates in an automobile manufacturer such as Ford or GM, for instance, then demand for the components that make up a fuel pump is governed by the number of fuel pumps that Ford or GM is actually ordering, rather than on the prices available on those components. You don’t buy more, just because they’re cheaper.
In short, buy the “wrong” pricing analytics tool, and the consequence is likely to be shelfware, rather than informed insight.
“I know a couple of Fortune 500 companies that have invested in pricing analytics software, but don’t utilize it,” says Liozu, the author and consultant, who also is an advisor at LeveragePoint. “No one looks at it, and no one uses it, because it doesn’t make a difference.”
“Analysis on its own isn’t necessarily the answer,” says Jason Gordon, a London-based partner in the analytics practice at Deloitte. “It’s a hugely powerful enabler within a broader dialogue with the customer, but not the answer on its own. The key is to use the results to inform discussion as part of a mutually respectful relationship between supplier and customer.”
Malcolm Wheatley, a freelance writer and contributing editor at Data Informed, is old enough to remember analyzing punched card datasets in batch mode, using SPSS on mainframes. He lives in Devon, England, and can be reached at firstname.lastname@example.org.