Pricing optimization is one of the most important factors in a company’s success. Conceptually, it seems simple – it answers a very concrete question: “What’s the ideal price I should charge for my product or service that will generate the most sales?”

Easy, right? However, it’s often one of the most contentious and least optimized practices for any business – and one of the most impactful.

Few companies use predictive pricing optimization

In my experience, there are a lot of competing points of view within companies when it comes to pricing. Inter-departmental conflicts are common. Fortunately, quantitative, evidence-based pricing can be the antidote for these challenges.

Of the dozens of clients that members of our team have worked with, none were using predictive or prescriptive analytics for pricing optimization before we entered the picture. So how did they determine the ideal price? It was typically very subjective and non-evidence-based, such as using price ratios or historical trends, or deep discounting to win over new customers. Prices were often adjusted by individual geographies and product groups, so any systematic pricing was at best an outdated guide.

As a result, pricing guidance was often ignored, especially at quarter or year end and for new product launches. This approach erodes profit margin and trains customers to engage in behaviors that are problematic for both parties, such as waiting for quarter or year end to buy at a deep discount. It also sows discord when long-time loyal buyers learn that you’ve been selling to new clients at 40% off their best price.

Unfortunately, among analytics agencies and vendors, pricing optimization is not a popular offering, because the intellectual capital and man-hours are very demanding. It’s a minefield, and even if you get it right, it will be next to impossible to copy and paste your solution onto another client because the entire process is bespoke – that is, it must be tailored to each client’s unique business model, ERP and CRM systems, product and service features, etc.

Why would Fresh Gravity want to write a post on pricing optimization?

Well, because we love to take on projects no other agency will touch – the impossible missions.

Consider for example a company that sells only one product. What could be easier? Well, if the company charges too little, they may increase the number of units sold and breadth of customers, but end up losing money. If they charge too much, the profit margin on each unit increases, but the number of units sold and breadth of customers will decrease, harming market share. Things only get more complex when a company has thousands of SKUs and a product hierarchy that’s several levels deep.

Now add multiple customer segments to the equation and things really get complex. Even a company with just one product can have several distinct customer segments, based on their industry, geography, geo-demographics, psychographics, and their perceived value of that product or service. Should you charge them all the same price? If you do, you’re leaving money on the table, because some are willing to pay more, and some less. Intelligent price discrimination is a proven model.

What about timing? We all know that many companies become very flexible on pricing when it’s end of quarter or year. Deals that were impossible a week ago now become doable. This leads to conflicts between the marketing department, field sales teams, and merchandising or finance groups. Prices that were carved in stone are relaxed by sales executives to meet their quotas. The marketing executives blame the sales executives for teaching customers the wrong lesson, which is to buy late at a discount.

Maybe you’ve seen this happen with trade show and expo tickets. Marketing offers an early-bird special, which is smart. However, as the event date grows close, ticket sales and attendance looks lower than forecast, so the EVP of sales orders the marketing department to give tickets away to the best customers, or to deeply discount event tickets, even lower than the early-bird special. Consequently, in subsequent years, potential attendees have been trained to wait to the last minute, and this creates all manner of problems, from all the best sessions being sold-out to lower show revenue to hotel booking problems.

If pricing optimization for a single product or service is complicated, then imagine how much more complicated it is for companies that have different product lines, with thousands of SKUs and product hierarchies. For some companies, each new product is unique, or has a unique mix of disparate features. In those cases, there may be no existing product that is comparable enough to serve as a precedent for pricing the new product. This is true for manufacturers of SoC (System on Chip) devices or even PLD (Programmable Logic Devices) – not to mention those other cryptic acronyms like 3DIC, FPGA, CLPD, ARM, ASIC, and ASSP. Good luck using intuition and sales experience to price those.

So what’s the solution?

The process has become so complex, and the competition so smart, that it’s imperative companies apply advanced analytics to solving the multidimensional challenge of setting the right price, for every customer segment, across every geography, for any sales cycle, relative to the maturity of products and services (launch vs. plateau vs. sunset). To do that requires a variety of techniques which we will describe later on.

The Fresh Gravity approach

We always start by listening to our clients – in the case of pricing analytics, that means interviewing our client’s key sponsor to get their views on their product and service lines. In this interview we get your perspective on strengths, weaknesses, and challenges. We learn about your success stories and formulate a plan to build on those.

In subsequent meetings, we expand participation to include other key business managers who are either impacted by or have an influence on product and service pricing. From these interviews, we develop a gap analysis on how different stakeholders perceive your performance on various pricing-related KPIs. This often reveals areas where departments are and are not aligned in their thinking. One goal of our pricing optimization practice is to create greater alignment between groups, and land on more effective rules of engagement to minimize conflicts moving forward.

We also make connections with your pricing team, as well as the members on your team responsible for various data and system domains that are connected to pricing. Typically, those are ERP systems like SAP or Oracle, CRM/SFA tools such as Salesforce.com, and even your marketing suite experts for Eloqua, Marketo, or Adobe Campaign. We do a deep dive into the structure of your data, and the cycle of input to conversion to output. As part of our analytics, we focus on understanding the metadata (data about your data) and your unique architecture and process flow. This will include product hierarchies and attributes.

In some systems, a single order can contain hundreds of products and service line items at the SKU level, and it’s not uncommon to have associated tables that reveal quotations on prices changing dynamically or manually as negotiations take place daily between your sales team and your customers on a quote. These dynamics are key to our scientific approach to creating predictive models for pricing optimization.

The data will tell a story if you help it talk

If you’re in the B2B space, we may append firmographics from Dun & Bradstreet to your customer records, such as SIC or NAICS industrial codes, company size, and growth history. If you’re in the B2C market, we may append geo-demographics and psychographic clusters from Experian to your customer records. This type of third-party data can greatly improve our analysis and the accuracy of predictive models.

Where appropriate, we also bring to bear pricing landscape analysis to understand how your competitors set their prices, and to understand how adjusting your price is likely to affect sales volume and market share.

Along with your internal data on sales histories, sales teams, resellers and distributors involved, product or service mix, we use a variety of analytic techniques to reveal valuable patterns that we can leverage for pricing optimization:

• Factors analysis
• Neural network-based clustering
• Time-dimensional analysis
• Conjoint analysis
• Structural equation modeling
• Market basket analysis

This list isn’t exhaustive, but it’s representative of some of the techniques we have found very powerful for systematizing and automating pricing optimization. The insights surfaced by these methods are very revealing and, in most cases, highly actionable.

We take a similar approach in our partner-empowerment program for B2B companies that reveals which resellers are ideal for specific sales environments and how to empower those resellers with the tools and intelligence they need to maximize sales of your products and services.

Industrializing analytics

Once all the various internal constituencies are on-board with our analysis, recommendations, and predictive models, we scale the solution by systematizing the models across your global enterprise. Our tools enable inside- and field-sales and channel partners to select from a few drop-down choices on their mobile device to see the ideal pricing inflection point to generate the highest likelihood of a sale at the best profit margin. Our method can also automate quotation pricing across all channels and devices.

We believe in Analytics as a Service (AaaS), so the story doesn’t end after clients get their first rate card or pricing breakdown. We incrementally improve on our models over time, so that when old products are sunset or new products are launched, you’re armed with the best model to accelerate your success.