6 questions to ask when growing revenue using analytics – July 2017

By: Matt Lindsay, Xavier van Leeuwe and Matthijs van de Peppel
President NRC Media and Mather Economics, Director of Marketing and Data, and Manager of the Data Intelligence
and Customer Relationship Management Team

Pricing decisions involve not only questions of revenue and volume but also the overall relationship with customers. Some businesses prefer less price differentiation, and others prefer more, depending on their strategies and market circumstances.

These questions have been addressed by many of our customers, and we felt it would be helpful to share them here.

News media is an industry near and dear to our hearts. Often, publishers and CEOs are not familiar with the dynamics of audience revenue because they rose through the editorial or advertising departments.

In addition, news media companies have often been very profitable in the past, so optimising audience revenue was not a priority. The company could rely on advertising revenue, and as a result, publishers often just raised customer prices across the board without big concerns over the lost customer volume.

Similarly, editorial teams tend to focus on readership and do not have much experience with audience revenue, retention, and acquisition. Only recently, as the industry has been disrupted by new digital platforms, have the issues of customer revenue and engagement received significant attention.

Here are some issues to consider as you develop a pricing strategy for your recurring customer revenue.

1. Volume or margin?

Companies consider several factors when answering this question, particularly firms that derive other revenue streams from the size of their customer base. These factors include advertising for publishers, interest income for banks, or in-app purchases for game developers. Where other revenue streams are significant, having higher customer volume is helpful.

It is important to optimise the revenue derived from these revenue streams holistically. On a related note, it is important to include different departments in the pricing strategy discussion to understand the effect of changes to one revenue stream on all other revenue streams.

2. Slow and steady or full throttle?

If additional revenue is needed quickly, companies can implement an accelerated test-and-learn process. An ideal implementation is to use a small but statistically valid sample of the customer base in a test of possible pricing options, but this will typically require several months to produce meaningful results.

Alternatively, several test groups that include a greater number of customers can be implemented simultaneously so insights are derived quickly and revenue is increased sooner rather than later. The risk from this strategy is an adverse reaction from one of the pricing test groups that negates some of the benefits.

Also, the degree of pricing changes plays a big part in the risk and return calculation in the slow versus fast testing decision.

3. Light tech or high tech?

You may be worried that you and your IT group have to invest a lot of time in connecting systems to implement targeted recurring revenue pricing strategies or deal with vulnerabilities in your payment processing or other operational systems.

The IT investment is often far less than what most companies expect, but where there are significant IT challenges, a pilot programme can be implemented. Again, this can be done without committing too many resources.

4. Size and nature of control group?

Using a control group to evaluate the effect on pricing and retention is an incredibly valuable tool. We strongly recommend this approach to our customers even after the testing phase is completed. There is an opportunity cost from using a control group and that is the foregone revenue from not including these customers in the price increase.

To minimise the lost revenue, the control group size can be selected as efficiently as possible to obtain a statistically valid sample. The pricing tests can be designed so this group also sees an increase of some sort in an across-the-board manner. Although not as accurate as a no-increase control group for measuring price elasticity, this method does provide insights into relative price elasticity across customer segments.

5. Differentiated or per product?

You will get the best results in customers saved (volume) and correct pricing (margin) if you have an individual customer pricing strategy. One customer’s rate may go up 20% while another’s rate may only increase 2%. However, if you worry about clients having a bad experience or about bad public relations, then you can choose less granular pricing strategies, such as separate increases per product or per bundle (like NRC did).

Once you have an understanding of the variation in price elasticity across your customer base, the trade-off between pricing at the customer, product, region, or other level of granularity can be estimated. In Mather Economics’ experience, the benefits of targeted pricing outweigh the costs.

6. What data to use in the pricing decision?

What type of customer data will be included in the analysis of pricing elasticity and the yield management process? Are any data fields sensitive for you or your customers? Are any practices prohibited by law?

Typically, gender, race, and religion are excluded from pricing decisions. Other variables can be considered for inclusion or exclusion, such as online consumption, where cookie permission is often needed. Customer tenure is typically a very valuable factor for estimating price elasticity, as is income. Age group can be helpful, too.

In most cases, demographic data on customers is aggregated to the ZIP+4 level in the United States and comparable geographic areas in other countries. Additional customer data will make your modeling and revenue strategies more accurate and efficient.


For full article, click here.

Optimize your data. We’re here to help.

Subscribe to our newsletter!