B2B Firms Inject Data Analytics into Pricing, Where Instinct Has Ruled

by   |   July 2, 2013 5:56 pm   |   0 Comments

At British catering supplier 3663 (the name spells “F-O-O-D” on a telephone keypad), a target of delivering a 0.5 percent improvement in achieved gross margin when selling to new customers has been exceeded fivefold.

Meanwhile, at computer and printer manufacturer Hewlett-Packard, an initiative to incentivize junior- and middle-ranking sales staff on achieved gross margin—without actually disclosing those confidential margin figures to sales staff—is starting to bear fruit, informed by better analytics.

Related Stories

Price optimization firms bring dynamic pricing to retail aisles.
Read the story »

Real-time pricing service designed to arm retailers in showroom battles with online merchants.
Read the story »

SAP’s SmartOps acquisition signals move to add demand sensing to HANA for supply chain management.
Read the story »

Guide to Procurement Analytics.
Read the story »

Operations executives: soft skills indispensible for analytics pros.
Read the story »

And at a Johnson & Johnson medical device subsidiary, a six percent increase in average selling price has been achieved at a time when a competitor has suffered 11 percent price erosion.

These anecdotes come from a recent conference in Brussels, organized by PROS Inc., a 27-year-old Houston-based pricing analytics and optimization consulting and software firm.  With roots in the airline seat reservation market, the company has subsequently branched out into a broader set of industries: manufacturing, distribution, services and travel.  PROS plays in an analytics niche with others that include IBM’s DemandTec, SAS and Zilliant.

As executives related their experiences deploying pricing analytics, the real takeaways from this event emerged not in the shape of achieved uplifts in gross margin and average selling price, but in the lessons learned by pricing and analytics professionals in the trenches of pricing analytics, as they battle to impose data-based insights on pricing decisions traditionally reached through more informal means.

Salespeople, for instance, will tend to want to pitch the price that they think will close the deal.  Marketing veterans, meanwhile, simply feel they “know” the right price.  And the finance function, for all their spreadsheets and cost breakdowns, find it difficult to apply cost-based pricing dictats in a world where salespeople routinely apply volume discounts and other sweeteners in order to close bigger deals.  Likewise, each function knows that whatever the basis of the price charged, any price that is too high will risk losing the customer altogether: in the long-term, customer retention is vital.

Three distinct themes characterized what these real-life businesses found, as they imposed data-informed decisions on one of the most sensitive aspects of business: what to charge the customer.

1. Pricing Analytics Isn’t Just About Pricing.
Predictably enough, most users at the event began by focusing their efforts on fine-tuning price offerings to existing customers.  The objective: to avoid lost sales by pitching offers that were too high, yet at the same time avoid giving away gross margin by pitching offers that were too low.

But this should be just the starting point, says Craig Zawada, whose title is PROS chief visionary officer, in a keynote speech.

Zawada, quoting unnamed Amazon.com insiders, said the online retail giant attributes as much as 20 percent of its revenues to the firm’s well-known “customers who bought that also bought this” recommendation engine. He urged executives to see the data that informs pricing analytics as containing nuggets on not just customer price-reaction behavior, but on buying behavior in general.

Also important: using the same raw data to craft compelling ‘upselling’ and ‘cross-selling’ offers—and also pitching finely-priced offers to sales prospects so as to maximize the prospect of winning new business.

At British caterer 3663, for instance, sales to new accounts since the analytics implementation went live have achieved a 2.6 percent improvement in margin, better than the 0.5 percent predicted in the business case for the investment.  In addition, notes 3663 head of pricing Trevor Pearson, of 932 new accounts opened, none have subsequently been lost—evidence, he says, of improved retention.

2. Segment, Then Sell.
At both Hewlett-Packard and 3663, market segmentation through analytics has provided a tool for generating meaningful insights into likely customer behavior, both in terms of price proposition and product attractiveness.

“We were brilliant on understanding our top 1,000 customers, but we had a million customers we knew almost nothing about,” says Michael Immenschuh, senior manager pricing process and capabilities at Hewlett-Packard.

The problem: in a big data environment—3663 sells 19,000 product lines to 60,000 customers, for instance, delivering to each customer multiple times per week—segmentation tends to follow simple easy-to-understand rules of thumb, which don’t embrace the full subtleties of real-life customer behavior.

3663, for instance, grouped customers geographically, splitting its U.K. market into a simple North-South split.  Analytics, he points out, showed that not only would a three-way split be more accurate based on customer behavior trends, but that a better way altogether would be to segment customers by their willingness or ability to pay higher prices—independent of their location.

How is such segmentation achieved?  Through “attribute discovery and analysis.”  In short, by using past sales data, product data, and any data available on lost sales and customer attrition, analytics makes it possible to segment markets by key characteristics such as business size, likely sales volumes, transaction history and similar distinctions.

The result: statistically-grounded predictions as to the effectiveness of given price pitches for given products and bundles of products, together with a ‘don’t drop below’ floor-price, designed to counter salespeople’s tendency to drop prices to levels that might earn plaudits and bonuses, but which deliver little gross margin.

Typically, for instance, analysis might suggest an “expert” price, perhaps to be used as a starting point, or achieved by an expert salesperson.  At a floor price, the salesperson will be instructed to walk away.  And in between, there’s a “target” price, representative of a good deal with an acceptable margin.

And it’s through such offer “price bracketing,” says Hewlett-Packard’s Immenschuh, that Hewlett-Packard was able to impose margin discipline on its sales force, without actually giving away the crown jewels in terms of what individual product margins actually are.

3. The Organizational Challenge
Needless to say, such insights create ripples when applied in the workplace.  PROS’ Zawada likens it to replacing a map with a GPS: salespeople can zero in on optimal deals, pitching with a high degree of probability, even when the customer is relatively unknown.

And in a business environment where companies’ pricing executives are sometimes mocked as “sales prevention officers,” statistically-based estimates of a customer’s willingness to pay can add the confidence to pitch bottom-line enhancing offers.

For instance, 3663’s Pearson, during his presentation, recounts the reaction of Alan Payne, a customer-facing sales executive.

“From a tendering perspective, I have found that [the system] has guided me to strive towards a more realistic selling price,” Payne says.  “By providing accurate, current data, I now have the confidence to drive a higher asking price, without fear of embarrassment in my account base.”

“It’s a real game-changer,” adds Hewlett-Packard’s Immenschuh.  “For the first time, it will enable salespeople to have a real reference point, and their managers to have conversations with them along the lines of: “If 50 percent of your colleagues can achieve these prices, and these margins, then why can’t you?”

That said, for reasons of confidence-building and to manage customer expectations, companies have typically moved slowly and carefully while putting in place new pricing regimes.

Peter Faaborg‑Andersen, global marketing officer at Danish biotechnology manufacturer Novozymes, for instance, speaks of implementing pricing analytics as “a journey… eating the elephant one bite at a time.” In practical terms, he explains that means using analytics-based insights to first enforce existing policies, and then root out and eliminate price anomalies.

And while analytics can provide a wealth of information, Faaborg-Andersen advises that enterprises use it sparingly. “Let your priorities be driven by value-creation, and keep things simple,” he says.  “Avoid the in-built bias to ‘complexify.’ The sales force won’t appreciate it.”

Finnish chemicals manufacturer Kemira, another European customer, is also treading carefully when it comes to closing the 5-10 percent gap it sees between target prices and achieved prices, adds Kemira director of marketing and product management Peter zum Humel.

“We did a lot of interviews with customers while performing our segmentation, and these interviews gave us clues as to likely price elasticity,” he says.  “We’re trying to increase the prices, but not to the point of losing the customer.  And now, we’re running tests to see if we can increase prices to those that the theoretical models suggest.”

It’s an approach that goes right to the heart of the question of price-setting, and the issues that it poses for the bottom line.

“The software is the starting point,” says Chris Jones, PROS’ chief sales officer.  “The real story is about the leadership and changed mindsets necessary to deliver results.”

Björn Willemsens, a director with Deloitte Consulting, concurs. “Price analytics and optimization projects don’t fail on the technology; they fail because of organizational and change management issues,” he says.  “You need very good people as sponsors, and very good people working on the analytics and the implementation.  While in theory salespeople should welcome analytics-based pricing guidance, in practice it’s often in the sales function that the resistance has been greatest.”

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 editor@malcolmwheatley.co.uk.

Home page image by Flickr user CoCreatr, Creative Commons license.

Tags: ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>