Part one of this two-part series discussed how Deloitte’s views on algorithmic risk management generally apply in distribution pricing, and key strategies to consider in managing pricing algorithms. This article, the second and final part of the series, illustrates how some of these issues play out in real-life distribution pricing projects, and offers relevant learnings from successful pricing transformation initiatives.
Algorithms are increasingly used across many business functions, but if not managed appropriately, too much of a good thing can backfire. This was the message of a recent Wall Street Journal report by Deloitte, On the Board’s Agenda: Board Oversight of Algorithmic Risk.
The first article of our two-part series, Managing Pricing Algorithms, discussed how Deloitte’s views on algorithmic risk management generally apply in distribution pricing. That article outlined some of the key factors that (rightfully) lead distributors to have strong interests in algorithmic approaches to price management.
Next, we offered some advice on how to identify and evaluate signals that algorithmic risks may exist in the pricing area. We then reviewed some common reasons that pricing algorithms underperform and may therefore increase algorithmic risks. These include the use of pricing algorithms that are not sufficiently market-informed and the use of outdated mathematical approaches in the design of the pricing algorithms. Finally, we warned distributors not to make false assumptions about the nature and level of sophistication of their pricing algorithms.
In this second and final part of our series, we illustrate how some of these issues play out in real-life pricing projects by reviewing the role algorithmic risk played for two private equity-owned distributors –ABC Distribution and Widget Distribution. Although both are fictional, their stories are reflective of pricing journeys of real-life distributors. The two distributors are similar in many ways:
- Both have about $60 million in annual sales.
- Both have management teams that appreciate the importance and powerful nature of pricing.
- For several years, both have experienced growing signs that their price structures may be inadequate for their businesses. Their reps regularly deal with customer objections – and these pressures seem to be gradually intensifying as customers increasingly research prices on the web.
- Both distributors have attempted to refine their price structures by way of internal analytics efforts, but these efforts underperformed versus targets.
- Both management teams decide to invest some time researching external service providers to see if they can find appropriate external help in this area. They have reviewed websites and marketing materials. They have talked to some peers. They have met with service providers attending industry conventions and scheduled discussions with them to have more in-depth conversations about how different providers might help improve pricing practices in their particular businesses. They have even seen some demos of different solutions.
- Both distributors decide to leverage outside experts and are ultimately successful. However, they take different approaches to solving their pricing challenges.
1. ABC Distribution’s journey
ABC Distribution takes this approach: “Our past ways of pricing have not worked, and this is evidenced by scatterplots that show random variation in invoiced prices/margins. So, we need to adopt a new pricing strategy that makes sense and is proven in the industry.”
Initial analysis of ABC’s transactional data shows that there is a “lack of bell curves” when prices/margins are mapped against volume. This reinforces management’s view that historical pricing practices were not strategically managed by the rep population. ABC decides to scrap its old, dysfunctional pricing models and chooses to adopt a pricing strategy that has been advocated by a pricing software vendor, which has proven to work at several other distributors.
The vendor ABC selects is quite experienced, and it shares a lot of learnings from other pricing projects, particularly stressing that change management and executive commitment oftenmake the difference between a successful and a failed pricing project. The vendor successfully maps ABC’s dataset into the vendor’s pricing cube/pricing matrix, and completes the analytics portion of project rather quickly, and with relatively little resource commitment from ABC Distributor’s already over-stretched management team.
ABC executives very much appreciate the timely and painless implementation: the business has several other initiatives going on and these stakeholders have little availability. Private equity ownership keeps the pressure on management to drive improvement in many functional areas, and pricing is just one of them.
Sales reps go through training that touts the strategic nature, mathematical sophistication and past success of the new pricing solution. Still, some sales reps remain skeptical of how effectively the new prices will work in practice. When the new pricing is launched, sales reps see many price changes. In some instances, the new prices seemingly override the reps’ pricing judgment, and they don’t always agree.
Though concerned with the impact of these price changes on their relationships with their customers, reps are instructed to follow the new pricing as management uses controls (and also some incentives) to force the adoption of the new strategies. Although management is fully committed to the success of the pricing initiative, they follow their vendor’s advice in some important aspect of transitioning to the new price structure.
The magnitude of the immediate price changes is kept smaller at larger accounts, particularly on more sensitive items. Reps also retain some (though much more limited) discretion to override system prices – and this helps them respond in select instances where price changes could damage strategic customer relationships.
Ultimately, the project is declared a success. Although some business leaves due to price, the favorable impact of the margin gains more than outweighs the impact of this volume loss. Management feels their pricing is now more strategic, allowing them to take risks in the right areas. Obvious pricing errors (such as selling at negative margins) are largely eliminated.
Reps understand that they have limited overriding discretion and are expected to “hold the line” in pricing. They also come to the realization that some noise around pricing, particularly resulting from customers making web-based price comparisons, remains a fact of life in the current business environment.
2. Widget Distribution’s journey
Widget Distribution takes this approach: “Our past ways of pricing have been inadequate, and this is evidenced by all the noise around pricing, and unexplained variation in invoiced prices/margins.”
The vice president of sales is vocal in his opinion that Widget is a sales-driven organization and that the sales team has over time accumulated a level of industry and customer pricing knowledge not easily replicated by standard pricing algorithms.
Although past efforts did not meet targeted results, Widget executives feel they were on the right track. But they lacked the expertise and time required to effectively devise a price structure that strikes the right balance between being simple enough to be manageable and understandable yet complex enough to reflect the many relevant aspect of marketplace realities that reps consider in pricing decisions.
Widget decides to look for additional expertise elsewhere to help them develop a pricing structure for its business. The company chooses to work with a highly credentialed external consultant with extensive experience in pricing in distribution experience, who was referred to Widget by its private equity owners.
In partnership with the distributor’s product management team, the consultant deploys multivariate statistical tools to devise a custom-built product segmentation tree. A council of seasoned sales reps works with the consultant on customer segmentation issues. The team decides to collect and build out some customer datasets to further inform segmentation and better profile prospects and customers. Widget also undertakes competitive price collection efforts.
The consultant conducts all analytics in a fully transparent manner. Pricing strategy discussions are held to identify, develop and document various pricing strategies that the business has used or is interested in adopting (as industry best practices are also considered in these discussions).
Although Widget’s data seems just as jumbled as ABC’s at first glance, the consultant is able to apply statistical tools to discern various drivers of past price variation from this otherwise “noisy” transactional dataset. As the transactional data is mined, the consultant confirms that Widget’s past pricing decisions already align with several sound pricing strategies. Segments with price/margin distributions resembling “bell curves” are formulated. From these analyses and discussions, a set of pricing strategies emerges, which build/expand validated pre-existing pricing approaches, while incorporating additional industry best practices.
This collaborative effort is somewhat taxing on the organization, as key organizational resources are spending a significant amount of time on the pricing project. The project ends up taking a month longer than initially planned for (four months in total). Given how the project unfolds, the delay is deemed necessary and acceptable. This is not without discussion: the distributor’s private equity ownership was already pushing for more compressed timelines. Ultimately, however, the board defers to the executive team in observing: “from our past failures we have learned that if we don’t do this right, we should not do the pricing project at all.”
When the new pricing is launched, it already has significant buy-in from the various stakeholders who were involved in its development. Sales reps from the pricing council voice their support in training sessions, describing how their input was considered in designing the pricing algorithms and how they reflect their own views on pricing in Widget’s markets.
As the pricing is rolled out, reps are comforted in that prices won’t change in most instances. These prices align with pre-existing pricing strategies that are retained (not scrapped) in the new pricing algorithms – strategies that the distributor and the consultant identified together and validated by analyzing past transactional data.
When prices are changed, prices at larger customers are tweaked more carefully, particularly on sensitive items. Pricing errors that caused undue and still unexplained price variation are eliminated. The results of the pricing initiative are carefully monitored. The distributor’s existing Microsoft tools are leveraged to create new reporting on price realization and other matters. The “noise” around pricing diminishes a great deal, and controls are found to be useful but not a major issue in implementation.
As system prices are competitive, the distributor begins to pick up more volume on its website as well. Targeted financial benefits are exceeded, and Widget’s project is considered a success as well.
As the distributor success stories suggest, there is no single right path to pricing transformation. The path can differ, depending on the trade-offs distributors make, knowingly or not, when they select their solution. Algorithmic risk is one of the factors in play; it is subject to trade-offs, just like other factors. In our tale, both distributors arguably made the right calls for their businesses.
ABC benefited greatly from transitioning to a more strategic approach to price management. Practically speaking, ABC might not have been able to timely and properly implement Widget’s type of pricing solution. There is more to business than just pricing, and many key ABC resources were consumed on projects to drive improvement in other functional areas. While they may have accepted taking on more algorithmic risk, they overcame the challenges thanks to executive commitment, a careful approach to gradual implementation of sensitive price changes, and the use of controls in price management.
Widget also benefited greatly from its pricing project. The company’s initiative arguably involved less algorithmic risk. This was more important at Widget. Given its sales-driven price management culture, taking on more algorithmic risk might have derailed adoption. But this came with important trade-offs: Widget’s project required more effort and more time to implement, and it also called for a level of pricing science expertise that is rarely found in the distribution industry.
Deloitte suggests we “should recognize the positive impact of algorithms.” We agree and strongly encourage distributors to use algorithmic approaches to manage their prices. In fact, market-aligned, advanced pricing algorithms can provide distributors with a competitive advantage.
We also agree with Deloitte that distributors should “use mechanisms to manage algorithmic risk.” Managing algorithmic risk in distribution pricing entails an understanding of trade-offs among different factors under consideration when distributors plan out and execute pricing initiatives.
An appreciation of the risk-reward aspects of these trade-offs can make the difference between failure and success, as distributors embark on their own journeys toward improved price management in their businesses.
Lee Nyari is managing partner of The Innovative Pricing Group, a consultancy offering strategic price management solutions for distributors. Reach him at email@example.com.