So You Think You Know Your Customers
Nearly every small business owner thinks they know their customers. But years of research yields a different story. They do know company names, approximate billing and key people, but having that information alone is like saying you know a book by its cover.
Many additional elements need to be known about customers to make the best business decisions. These include aggregate information, cohort information and individual information. Each level of information contributes to different strategies and tactics that help better target and communicate with existing customers about building existing business and identify possible new customers.
Different businesses in different industries have different customer profile. Some become self-evident with the data, while others respond to classic principles. One such principle, called the Prado principle, states that 20% of something will account for 80% of the activity or business. In this case the Prado principle, or 80-20 principle, will be that 20% of your customers will contribute 80% of your business. This of course is a general or average across many data sets, but is serves as a benchmark.
Collecting the right amount of information from customers is clearly important to knowing more about them. Many companies just record dollars bought. Perhaps they don’t even know the customer’s name. BtoB (business to business) customers generally expect to provide this information, while BtoC (business to consumer) sometimes want to remain anonymous. Each type of customer must be dealt with differently.
Consumers are all individual persons who (except for children) make their own decisions about what to buy. They buy for very complicated reasons that behavioral market researchers are continually studying for insight. This has been the subject of many entire books and we will not take this up in detail here. So suffice it to say now that different types of consumers have different buying habits. In order to understand them better, marketers group them into cohort groups, or groups of similarity, to analyze how these similar groups behave toward your products or services.
Small businesses often have a small number of cohorts as customers because they are serving small market niches. Large companies service larger markets and therefore normally have a wider selection of cohort groups. This is illustrated by comparing a department store to an independent dress shop, where the owner has a certain taste level that her customers like.
Capturing data must occur to create cohort groups. Name and address is normally all that is needed to find out an enormous amount about a person. This is because in most locations in the US, the neighborhood separates people based on income (house value) and lifestyle. Lifestyle is quite different in suburban neighborhoods compared to apartments in major cities. These people tend to have different family structures, shop in different stores, and spend leisure time differently.
Capturing a name and address is relatively simple today. Without anything but a spreadsheet, these two items can be stored. Other information can also be added such as what was bought (category and/or SKU), how much was spent, and when.
Even accounting programs such as QuickBooks allows the capturing of this type of information.
What to do with this
Sorts and filters can be applied to spreadsheets to create lists, totals and counts of different items. Believe it or not, all this was done very successfully using paper journals for well over a hundred years, so it is certainly possible today. If you don’t have information at this level, your understanding of customers is greatly diminished.
In a retail store one of the most critical numbers is the total of each category of merchandise sold. All retailers must understand this so they can replenish inventory in an intelligent way. Those who try running business without this information or on the fly, almost never stay in business over several years.
To use the example of a specialty women’s store, it is critical to know how much was sold each month in dresses, jeans, tops, shoes, accessories, skirts, jackets, shorts and swimsuits. Knowing what type of customer cohort bought each category provides additional insight about customers that I think is obvious. Knowing sizes, colors, styles, price points and markdowns is much more valuable.
Address information alone can also provide an incredible amount of information on each cohort. In the US, market researchers and data companies have carved up the entire population into between 45 and 64 cohort groups, depending on the model. This means that all people in the country fit into about 60 groups of people with an incredible similarity for buying behavior, lifestyle behavior and demographic range.
The models vary from the wealthiest who live like Hollywood stars, to high income families that live well below their means, to young single people starting out in the city, to struggling moms with children. Certain behaviors require money, some do not, but if you have money you can buy boats, cars and expensive dresses that many others cannot. A single mom with a low income has typical behaviors with her cohort group, too. She might buy a dress from a nice women’s store, but she will likely not buy many expensive ones.
This information is used by even small companies to see who their customers are, and with the information comes full profiles and many additional behavioral activities. If you have many of one type of customer, it helps to know if her cohort reads certain magazines, how much time she spends shopping on the internet, if she responds to coupons, if she has an active lifestyle. If the store owner is thinking about adding workout gear, she would know exactly how much she works out and if she is health conscious.
In a real life example, a milk delivery company had 80% of their business from just 12 cohort groups. This sounds a lot like the Prado Principle, huh? Knowing this, they stopped soliciting in the other 45 cohort neighborhoods and saw their selling efficiency soar. They also increased the density of customers which increased the brand recognition and created whole neighborhoods of supportive customers. They focused efforts on carefully chosen schools to support in the same neighborhoods, instead of randomly donating money to various causes, supporting the customers who support them.
Because this dairy keeps great individual records of what customers buy, new products can be targeted to specific groups having a better chance of acceptance rather than just randomly adding new products. Marketing is specific to these customers based on the lifestyles they lead, and media could be used that is known to be favorites of these customers. By separating out customers from prospect lists, the message can be tailored to how each cohort groups thinks about their products. Some see milk as a healthy drink for the kids, while others use chocolate milk as a re-hydration drink following a run with more protein than sports drinks. These messages resonate with customers who then develop a kind of relationship with the company and its products.
Cohort information is also useful to understand how many more are in the market. Counts by cohort group are compared to the number of customers to yield market share. If the share is low, it means many more likely customers exist in that area. If it is high, it means growth of new customers will slow, so more must be sold to each family to grow that area. The strategy changes based on the reality of the market driven by the numbers. Unbelievable waste can occur when advertising and promotional money continues to be spent where no market exists. It is one of the underlying reasons why much advertising simply does not work.
As the old saying goes, fish in a bigger pond to get bigger fish. If you’ve caught most of the fish, switch to another pond. Well’ that’s not the exact saying, but you get the idea. Data drives so many marketing decisions that not having it is like driving blind.
B to B Customers
Business customers are a completely different issue than consumers. Buyers in businesses generally buy for more logical reasons. They have a need and are necessarily price conscious. They buy tens of thousands or millions of dollars of goods and services and get very good at the professional practice of buying.
This level of sophistication creates a different selling environment. The level of service that comes with large or sophisticated goods is also much higher than for consumers, and many times requires more sophisticated support than consumer support.
But companies also can be placed into categories of industry type and size. This information is also important to determine the importance of each type, what each type tends to buy and how many more companies exist like the customers.
Lots of list companies can provide counts by industry, company size, employees etc. The US Commerce department and Census Bureau also have data on businesses and industries. Comparing how many accountants are customers for your software can help determine better ways to communicate value to accountants specifically. Certain industries and sub-industries will also be the top 20% and provide 80% of revenue. If not, it raises questions about your mix of customers that can provide insight about them and how to grow your business. It also is one way to measure interest and create new innovative products for specific groups of companies.
Comparing the percentage of companies at each revenue level in your local or regional economy against your own can indicate a strength or weakness in a certain size company. This insight can help marketing efforts.
Using this type of data is a discovery method. Not all data or summary data means something, but discovering a new insight can be valuable and is usually worth the effort.
Individual Customers – Surveys
Knowing more about individual customer is also very helpful. This data generally falls into three types: meta data, data that is gathered due the transaction and data that is gathered outside the transaction. The latter is most often collected in a survey.
Meta data is information that is gathered when a customer is new, or from a third party. Information can include company revenue, number of employees, headquarters vs. SBU (Strategic Business Unit, a sub business of a large corporation), executives and industry type. Data gathered due to the transaction is company (address), buyer name, vital information to do the order and the products, revenue, margins, quantity and frequency of orders, returns, number of contact touches, etc.
Data that is gathered in surveys or through conversations with customers includes attitudes and awareness of products, services and product preferences. Each industry and business will have several specific kinds of information to gather.
Today keeping this information attached to the company name record is more important than ever, and easier than ever as well. Still, much information is not well collected and creates “bad data” that can slow the process of analysis. Organizing this data requires logical thinking about what will be useful and what is extraneous.
Used together, all this data continues to be extremely useful in all kinds of company decisions that lead to tactics and strategy. The competitor with more and better data will have an advantage over you because they have the ability to make better decisions with the data than competitors without the data.
Other expected results might be that 20% of your SKU’s (stock keeping units) will contribute 80% of your revenue. Falling outside of this baseline is not necessarily a problem, but it does provide thinking fodder about the advantages, disadvantages or opportunities for changing your mix.
Combining just these two data summaries can provide further insight. If both of these examples hold true for your company, then what does it mean when one of your main customers, who are in the top 20% for providing revenue, has a product mix that is opposite of your normal expectation(i.e. they buy 50% of their revenue from products in the lower 80% of your performing SKU’s) ?
This would make them very different from the average product mix from most customers. What would it mean? How would you respond to such a difference?
Another important bit of summary data is to determine the contributed margins by customers. Does the largest customer also contribute the largest margin dollars? If not, this might mean that they are the gorilla that demands more service or lower prices. What does that say about your strategy for that customer next year? Can you attempt to reverse this, or do they still contribute so much to fixed costs that you can’t afford to let them go?
These and many other questions can be answered by understanding customers better.