By Patrick Burns, Principled Strategies

How to Get Extra Value from Your Analytics Initiatives

With tight budgets and project timelines it becomes important to craft an analytics strategy that can quickly and cost-effectively address the question at hand. With forethought and careful planning, it is possible to design such a project to be flexible in its ability to answer other, related questions.

To be specific, the careful selection and development of a core analytic strategy creates an opportunity to answer a wider range of questions that may be directly or even indirectly related to the initial question at hand. This “invest-once-use-many-times” strategy may require a bit more time and budget to pull off, but will pay attractive dividends in the long-run.

The case study presented in this article took place in the pharmaceutical sales and marketing sector, though is conceptually extensible to similar sales and marketing questions in other sectors.

Introduction

The pharmaceuticals sector, like many brand-oriented sectors, uses a myriad of sales and marketing tactics to drive revenue.  A typical pharmaceutical manufacturer will invest a significant portion of its total Sales and Marketing budget on Sales Professionals who make sales visits (referred to as detailing in the industry) to healthcare practitioners. During a sales visit, samples are offered as a promotional marketing tool to provide practitioners an opportunity to initiate therapy with new patients on a short-term trial basis. Additional advertising and promotion tactics deployed in support of a brand include professional trade journal advertising, educational meetings and events, direct-to-consumer advertising in mediums such as television, radio, print, as well as host of complimentary, internet-based channels.

Sales Force Effectiveness (SFE)

By some estimates there are nearly 100,000 pharmaceutical sales professionals in the United States pursuing some 800,000 practitioners, a ratio of 1 sales professional for every 8 practitioners! Sales professionals often have a call list of 200 practitioners with 120 targets who qualify for a sales visit at the rate of 12-24 times per quarter. Thus it wouldn’t be unusual for a sales force promoting a large pharmaceutical brand (>$1B, or a “blockbuster”) to be staffed by 600-900 sales professionals divided among multiple sales teams.

The measurement of sales force effectiveness (SFE) presents a significant business challenge. There are multiple perspectives on what constitutes “effectiveness” and, in fact, effectiveness truly is a function of many activities, including targeting, contact planning, sampling strategy, sales force size and structure, sales force optimization, and sales professional training. With that in mind it is easy to see how a SFE project really isn’t about asking just one question. For example, you might initiate a project to explore sales force size and structure, and in the process of reviewing results develop insights about opportunities to improve targeting.

The trick to extending the value of your analytics investment dollar is to avoid boxing yourself into a corner. How? You must avoid situations in which the choice of a particular analytic method, which while well-suited to answering the core question cannot be used to explore other questions without retrofitting the data, analysis plans and/or methods.
This suggests that a flexible analysis plan is needed, as can be demonstrated in a case example for an SFE audit; which consists of a high-level assessment of the impact of key drivers on revenue.

Example SFE Audit Project

In Fall 2009 a well established, mid-tier pharmaceutical manufacturer commissioned an SFE audit with our firm. Instead of approaching the analysis with the sole purpose of assessing sales force effectiveness, we also considered a promotion optimization initiative. A promotion optimization analysis framework is capable of identifying substantial revenue lift opportunities from improved tactical management of sales visits and sampling, traditionally the two most expensive line items in a pharmaceutical brand’s advertising and promotion budget.

As a result, an analysis plan was recommended to perform the following specific diagnostics at a cost marginally higher than the originally proposed sales effectiveness audit:

  • Estimate the revenue lift associated with optimizing sales visits and samples
  • Estimate the effectiveness of the current target list
  • Compare the relative effectiveness of the sales force to internal and external benchmarks

SFE Audit Results

Within six weeks, sales visit response and sample response models were completed, SFE results were presented, and the methods were validated using hold-out data. The promotion response models both supported the measurement of tactic effectiveness and supported optimization of revenue associated with ongoing investment in the tactics. This process involved five key steps:

  • Build SFE (promotion response) models using data from Q1-Q3
  • Estimate optimal level of sales visits and samples for Q4
  • Use models to predict Q4 prescriptions given optimal sales visits and samples
  • Compare Q4 actual sales visits and samples to Q4 optimal sales visits and samples
  • Measure economic benefit associated with difference between actual and optimal

Revenue lift is accomplished by showing that sales representatives whose actual sales visits and samples are close to recommended values also experience the highest gains in motivating practitioners to write brand prescriptions.  Although the models are based on complex mathematics, the validation of the models is conducted using simple linear models so that the client can reproduce the validation results internally should they wish too.

So what is the SFE model doing “under the hood”? It employs a value metric for each practitioner that estimates the equivalent number of potential prescriptions possible through optimization. This is accomplished by creating an empirically constructed weighted market basket; that is, by combining prescription activity for the client brand with prescription activity for competitor brands into one number, so that all practitioners can be ranked in terms of their equivalent prescription-writing (revenue) potential.  The value metric is also informed by other relevant information about the practitioner, including their specialty (Internal Medicine, Primary Care, etc.), sales trend and seasonality within their state, and managed care access (e.g. practitioner accepts private or government insurance).

The SFE audit resulted in the following validated revenue lift opportunity, which includes three quarters of repeat purchases from patients:

  • 5.8% from improved sales visits
  • 6.3% from improved sampling

The core SFE audit models provided an opportunity to explore other analytical questions. For example, the weighted market basket analysis permitted practitioners to be rank-ordered by targeting value, even if they were not being actively promoted. Two key insights emerged from this secondary use of the models. First, the value of over 30% of targeted practitioners was significantly higher or lower than the client’s estimate; consequently, promotional investment dollars for these practitioners were being inefficiently deployed. Second, after inspecting the correlation between practitioner value segments and revenue, it was observed that the optimized value segments were much more highly correlated with brand revenue than were the current value segments (0.8 versus 0.45). This insight led the company to incorporate the optimized value segments into all tactical planning for this brand for the remainder of the year.

The pharmaceutical manufacturer gained an understanding of the opportunity associated with improved SFE as well as from optimization, but more importantly understood that a core model was in place that could help to answer other important questions.

Post-SFE Audit Question #1

Prior to the SFE audit, the company was developing a non-personal sales promotion campaign which included direct mail, e-mail, outbound sales calls, and samples. Targets for this campaign included high-value practitioners who were not currently receiving samples and sales visits. In light of the SFE audit, one smart manager wondered aloud whether the audit findings could help to answer an important targeting question:

  • Who are the optimal targets for this campaign?

Once that question was posed, the secondary usefulness of the SFE audit became clear to management, and other questions were posed:

  • How many practitioners should we target?
  • How many samples should we provide?
  • Given the fixed campaign budget, how do we evaluate the cost-benefit of creating multiple promotion tiers within the target practitioner universe?

Since the equivalent weighted market basket (as described above) could be computed for any practitioner, it was relatively straightforward to identify the top 60,000 (8%) practitioners not currently serviced by the sales force.  With an optimal sample model in place, a better quarterly sample target could be produced for each of the 60,000 practitioners. And via the creation of a financial calculator which used, among other inputs, the results from the SFE model, the answer to the question of how to design multiple promotion tiers in order to maximize ROI was within reach.

Post-SFE Audit Question #2

With the value of an “invest-once-use-many-times” strategy proven to management, a SFE audit was commissioned for a second brand. The subsequent SFE audit produced results even more compelling than the first (e.g. higher revenue lift potential):

  • 6.2% from improved sales visits
  • 9.6% from improved sampling

Soon after completion of the SFE audit for the second brand, the company unexpectedly lost patent exclusivity protection for the first brand, which resulted in a decision to reallocate sales and marketing investment resources from the first brand to the second brand. The natural starting point for questions about this strategy included:

  • Who are the optimal targets for this brand?
  • Given the revised sales budget, how many practitioners could be visited and sampled with regularity?
  • What was the optimal number of visits per quarter?

Once these questions are answered, the appropriate number of sales professionals could be estimated, and the alignment of targets to sales territories could proceed.
Because a flexible promotional response model had already been established and validated as part of the SFE audit, it was both cost- and time-effective to create a prioritized target list, and then use SFE model response results to explore reach and frequency scenarios.  It took less than two weeks to provide sufficient information for management to create a new sales force appropriately sized and structured to maximize ROI since the SFE model was validated and in place.

What’s next? Once the sales force is deployed, the natural extension of the SFE audit model is to create optimal recommendations for both sales visits and sampling.

Summary

The key lessons that can be learned from this experience are:

  • Organizations should carefully develop their analytical strategy to be capable of answering as wide range of questions as possible. Many additional insights are possible from flexible analytic strategy and planning.
  • Marketers can maximize internal project ROI and reduce project timelines by reusing models and analyses, and building on prior outcomes.
  • Models and analytics can provide useful organizational guidance beyond informing a specific strategy, tactic or promotional activity. Insights can even be helpful for appropriately structuring an organization’s sales force.
  • Marketers should fully explore the intelligence derived from modeling and apply it to their business use cases whenever possible.

Patrick Burns is President of Principled Strategies, a firm that provides their clients with the intelligence and insight they need to make smarter strategic, tactical, and clinical decisions. He can be reached at patrick.burns@principledstrategies.com.