Share with your friends










Submit

Analytics Magazine

Basic sales analysis

Twelve ideas for anyone assigned the task of analyzing a firm’s sales data.

Jerry W. ThomasBy Jerry W. Thomas

Most companies have massive databases of historical sales data, but few firms invest the money and staff time to mine the intelligence hidden in those databases. It seems that everyone has sales data, but almost no one does a good job of analyzing that data. The purpose of this article is to present some basic ideas on sales analysis that might serve as a starting point for any novice who might be assigned the task of analyzing a firm’s sales data. The discussion is written from the perspective of sales data at a manufacturing company, but the concepts apply equally to retailers and service companies.

1. Sales by SKU. The first, most basic level of analysis is examining sales data by product and stock keeping unit (SKU); that is, each size and variation of each product by state or smaller geographic areas and across time. What is the trend in sales of each SKU over several years for each state, metropolitan service area (MSA) or other geographic areas?

As marketing directors study changes across time, the core question is always, why? Why are sales trending up in California but down in New York? Answering the “why’’ should be the ultimate goal of all sales analyses because knowing the “why’’ allows marketers to project into the future.

2. Sales by channel of distribution. The next level of analysis involves distribution channels. What is the trend in sales by type of “retail’’ outlet? Large individual retailers should be looked at separately, and retailers should also be grouped by type (supermarkets, convenience stores, warehouse club stores, distributors, vending machines, online, etc.).

Sales data take on new meaning and reveal new understanding when examined on some type of relative basis. Photo Courtesy of 123rf.com | © christianchan

Sales data take on new meaning and reveal new understanding when examined on some type of relative basis.
Photo Courtesy of 123rf.com | © christianchan

Sales of all products might be aggregated to study the distribution of total product sales by different channels. The changes from year to year, or quarter to quarter, in sales by channel can be quite revealing. For example, several years ago we were able to detect the growing influence of Walmart as a distribution channel for groceries by analyzing a food manufacturer’s historical sales by channel. At the time, Walmart was not on that food manufacturer’s radar screen, and it probably was not on the radars of other food manufacturers. By pinpointing the growing importance of Walmart early on, the client was able to focus attention on Walmart and take advantage of Walmart’s growth.

While looking at sales in total by channel is a great starting point, marketing directors can also examine sales by SKUs, brands or product categories by channel of distribution, especially the trends over time.

3. Per capita sales. Sales data take on new meaning and reveal new understanding when examined on some type of relative basis. Per capita sales analysis is one of these relative techniques. For example, marketers might take the annual dollar sales – or unit sales – of one of their brands state by state, divided by the total population of each state. This would provide the annual sales per person by state.

We once did this type of analysis for a greeting card company and discovered that per capita sales of the brand declined as a function of distance from corporate headquarters. States close to corporate headquarters had higher per capita sales, while states farther away had lower per capita sales. The per capita differences by state reflected the historical development of the company, as well as the greater attention the closer states received from the firm’s higher-level executives.

The analysis provided evidence that sales levels per capita in distant states could be increased with greater managerial attention.

Per capita sales analyses can be based on total population, male population, female population, adult population, children population, individuals earning more than $50,000 or any other population group. The analyses can be run by country, states, groups of states, MSAs, designated marketing areas, by individual cities or areas within cities. The results are often quite revealing and can lead to improved sales efforts or better targeting of advertising.

4. Per comparable economic data. Another way to analyze sales data on a relative basis is to compare sales to various economic data. For example, marketing directors might compare annual dollar sales or unit sales to gross domestic product (GDP) by state, or to total electricity consumption by state, or to total gasoline consumption by state, or to hundreds of other economic variables. These types of analyses help answer the question: How well is our product, service or brand doing by state (or other geographic areas) in relation to other measures of economic potential? It may be learned, for example, that the superstar salesperson in Southern California who is being considered for promotion to national sales manager is actually selling less in relation to economic potential than salespeople in other areas. Analyzing sales in relation to economic variables can also be valuable in helping set sales targets and quotas by geographic areas.

5. Category development index. A widely used measure of product category development is the category development index (CDI). The index can be calculated for any geographic entity, such as city, county, state or groups of states.

Let’s suppose marketing executives are interested in the CDI by state. If they have total annual sales for a product category by state, they can calculate a per capita category sales number for each state. Next, they would calculate an overall average for per capita sales for the United States. Then they would divide each state’s per capita sales by the overall average per capita sales, and multiply by 100.

What is created is an index number where 100 is the average index score. Any state with a CDI above 100 is an above average market, while those with CDIs below 100 are below average markets. The CDI provides a reliable and consistent measure of market potential, or market development, by state for the product category.

6. Brand development index. The brand development index (BDI) is a very similar measure to CDI, but it focuses on an individual brand within the product category. The method of calculating the BDI is analogous to that for the CDI: The brand’s per capita sales by state are determined, and the brand’s per capita sales are calculated for the whole U.S.

Then each state’s per capita brand sales are divided by the average U.S. per capita brand sales, and multiplied by 100. For the resulting BDI number, 100 is the U.S. average. Above average states enjoy BDIs greater than 100, and below average states have BDIs below 100.

BDI need not be based on per capita sales for the whole population. BDIs could be based on adults only, teenagers only or men only, if per capita sales data are available for each of these groups. BDIs provide a measuring stick to help marketing managers allocate advertising and promotional expenditures across various states or other geographic areas.

7. Competitive trends. If companies monitor their market share via a syndicated service, or some type of tracking study, then they know what the total market is doing, and they know what their competitors are doing. For example, if they know their brand of peanut butter has a 6 percent market share in Iowa and annual sales in Iowa are $2.5 million, then they know that the size of the total peanut butter market in Iowa is $41,667,000 annually (100/6 times $2.5 million). This competitive share data also allows companies to calculate how much peanut butter each of their major competitors is selling in Iowa per year.

Tracking market share is vital because it allows marketing directors to know how well the company is doing in relation to major competitors. The company may be growing 20 percent per year, but losing ground to competitors.

8. Analytical database. The sad truth is that virtually no company has the important and relevant sales data and related variables neatly organized and maintained in a pristine database, and that’s why building an analytical database is often the first step to any type of sophisticated sales analysis.

An analytical database would contain all of the historical sales data, of course, but it would also include related competitive data, if available, plus demographic data and economic data by the smallest geographic units feasible. If budget and time permit, marketing variables such as advertising spending by media, pricing information, promotional spending by type, distribution levels, etc., would be continuously added to the analytical database.

The more inclusive and precise the analytical database, the more valuable it is as a foundation for advanced analyses. With an accurate analytical database, sophisticated marketing mix modeling is an analytical possibility.

9. Cross-tabulations. Once the analytical database is in place, some very powerful analytic work can be accomplished by creating cross-tabulations. One method is the “contrast of extremes.” How do the geographic areas with the highest BDIs contrast to the areas with the lowest BDIs? What do these differences reveal about the reasons for a product’s success or failure? What about the contrast between geographic areas receiving heavy media advertising versus those with little media spending?

What about comparing geographic areas with high unemployment versus those with low unemployment, or high economic growth areas versus no-growth areas? What do these comparisons say about a product’s likely performance during periods of economic distress?

10. Multivariate analyses. What’s really driving the sales of a given product category or a specific brand? Companies always have explanations for the success – or failure – of a product, but these explanations are typically mythology. Few companies have done the scientific econometric analyses to measure “cause and effect’’ and determine what variables are driving a company’s (or brand’s) growth or decline.

Once marketing directors start building their company’s analytical database, they can enhance it with historical economic data. A great source for high-quality, free economic data is Federal Reserve Economic Data, offered by the St. Louis Federal Reserve.

The American Community Survey is another source of high-quality government data, as is the U.S. Government’s Consumer Expenditures Survey. Or economic data can be purchased from a number of commercial sources. Marketers must identify all of the demographic and economic data they suspect might be causing their brand’s sales to increase or wane, and include that data in the analytical database.

Multiple regression is the technique that’s most widely used for this type of analysis. Marketing directors can use commercially available software, or they can go to www.decisionanalyst.com and download STATS 2.0 – a free, easy-to-use statistical package. The dependent variable (the thing marketers are trying to explain or predict) is a brand’s sales or category sales. The independent variables – those things that might be causing sales to go up or down – include such information as employment trends, household income trends, demographic trends, inflation trends, rail shipments, electricity consumption and so on.

The variables marketing directors choose will vary by industry and product category, of course. If they have chosen the right variables, then multiple regression should indicate which variables are most important in explaining sales of their brand. They will want to experiment with different types of regression and, by lagging some of the variables, to see if they can improve the model, as measured by R-square – the percentage of variation in the dependent variable explained by the independent variables.

11. Loyalty programs. Many retailers, but not manufacturers, own reams of loyalty program data about their customers. Retailers often know exactly what an individual household is buying, and how those buying patterns have shifted over time. If the analyst is careful and cautious, some of this loyalty data can be used to supplement the other types of sales analyses discussed in this article.

By choosing a representative sample of loyalty program records, the analyst can examine patterns, trends and changes in sales by demographic variables and by time and season of the year. This micro-level data can often shed light on trends reflected in overall sales.

12. Marketing research. Marketing research is often the “missing link” that helps analysts understand the patterns and shadows they see in their sales analyses. Sales analysis tends to be backward looking, since it is based on historical data, but marketing research can help marketing directors look into the future.

At some point marketing leaders will need to bring in qualitative research (focus groups, depth interviews, ethnography, etc.) to find answers to the questions raised by insightful sales analysis. Marketing directors might need to do surveys among their target audience to measure their brand awareness, advertising message recall, brand image or trial. Or they might need to test their advertising, promotions and new products among target-market consumers. Sales analyses can only take you so far.


Jerry W. Thomas (jthomas@decisionanalyst.com) is president and chief executive of Dallas/Fort Worth-based Decision Analyst Inc. (www.decisionanalyst.com), a global marketing research and analytical consulting firm. The company specializes in strategy research, advertising testing, new product research and advanced modeling for marketing-decision optimization. Thomas welcomes comments, suggestions and corrections.

Related Posts

  • 64
    Companies across industries admit to the growing importance of data analytics to improve sales and marketing effectiveness and decision-making. However, many struggle to piece together siloed data, properly define the problem or design the solution. As a result, they often fail to realize widespread business impact from their efforts, according…
    Tags: sales, data
  • 61
    The Internet of Things (IoT) is considered to be the next revolution that touches every part of our daily life, from restocking ice cream to warning of pollutants. Analytics professionals understand the importance of data, especially in a complicated field such as healthcare. This article offers a framework on integrating…
    Tags: data
  • 54
    To download a PDF of a white paper please click on cover images and fill out the brief information form.   To find out how your company can benefit from posting White Papers on Analytics-Magazine.org contact one of our advertising sales personnel: Companies A-L Aileen Kronke 888.303.5639, ext. 212 Direct:…
    Tags: sales
  • 52
    It’s long been popular to talk about customer interaction data such as clickstream, social activity, inbound email and call center verbatims as “unstructured data.” Wikipedia says of the term that it “…refers to information that either does not have a pre-defined data model or is not organized in a pre-defined…
    Tags: data
  • 50
    Today, we live in a digital society. Our distinct footprints are in every interaction we make. Data generation is a default – be it from enterprise operational systems, logs from web servers, other applications, social interactions and transactions, research initiatives and connected things (Internet of Things). In fact, according to…
    Tags: data

Analytics Blog

Electoral College put to the math test


With the campaign two months behind us and the inauguration of Donald Trump two days away, isn’t it time to put the 2016 U.S. presidential election to bed and focus on issues that have yet to be decided? Of course not.


Headlines

Three keys for organizations to gain value from information

In the current information-driven society and increasingly digitalized world, Gartner, Inc. says that sentiments are shifting from the economics of tangible assets to the economics of information – “infonomics” – and other intangible assets. Infonomics is the theory, study and discipline of asserting economic significance to information. It strives to apply both economic and asset management principles and practices to the valuation, handling and deployment of information assets.  Read more →

Burtch Works study on ‘Salaries of Predictive Analytics Professionals’

According to the recently released Burtch Works study on “Salaries of Predictive Analytics Professionals 2017,” senior-level executives saw the largest increase in salaries from 2016 to 2017, and industry diversification of employment has diluted the concentration of such professionals from financial services and marketing/advertising to consulting and technology. Read more →

New study asks, ‘Is your business AI-ready?’

Despite fears that robots will replace human labor, the majority of artificial intelligence (AI) leaders (79 percent) expect their employees will work comfortably with robots by 2020, according to a new Genpact survey of C-Suite and senior executives titled, “Is Your Business AI-Ready?” Read more →

UPCOMING ANALYTICS EVENTS

INFORMS-SPONSORED EVENTS

2017 Winter Simulation Conference (WSC 2017)
Dec. 3-6, 2017, Las Vegas

CAP® EXAM SCHEDULE

CAP® Exam computer-based testing sites are available in 700 locations worldwide. Take the exam close to home and on your schedule:


 
For more information, go to 
https://www.certifiedanalytics.org.