The GIGO Rule Applies to Marketing Too

Garbage in, garbage out. It is a tried-and-true rule in accounting. Inadequate or inaccurate financial data equates to inaccurate and inadequate financial information.

Full stop. What about data excellence for consumer marketing?

The reality is that data excellence in consumer marketing is more challenging than when applied to accounting.

In accounting, companies freeze results as a snapshot in time to report trends and earnings. The numbers are the numbers.

However, a marketer’s interaction with a consumer is fluid. Immediate past results might not be indicative of a consumer’s next action. Great marketing is driven by analytics that guides the marketer to deliver the right message at the right time.

Consumer marketers, especially retailers, have often relied on consumer buying data of recency, frequency, and monetary (RFM) value to segment customers for targeted communications. RFM uses data related to how recent a customer’s last purchase was, how frequently a customer purchased from a brand, and the monetary value of the product(s) or services purchased. In today’s world of data, it is a recipe for GIGO.

Marketing Technology News: MarTech Interview with Dmitry Kudrenko, CEO at Stripo

The reason? RFM may have been state-of-the-art analytics for online retail sales years ago. Today it is outdated and insufficient to optimize customer loyalty and the brand’s profit. RFM limits a marketer’s segmentation capabilities because the data it relies on is limited in scope. And limited data is garbage.

Consider that recency and frequency measures can be significantly influenced by the recency and frequency of the discounts a brand has offered. So, in relying on RFM, marketers are effectively training customers to purchase at a discount. The higher the ratio of these customers a brand has, the lower the profit margin.

Examine a customer’s discounted purchases divided by total purchases. The closer this figure gets to the number one, the more the customer’s buying decisions are based on discounts. In simple terms, if a customer makes a total of two purchases on discount, the divided ratio is 2 divided by 2, equals 1 (2/2 =1). If another customer makes a total of five purchases, four at full price and one at a discount; the divided ratio is 1 divided by 5 equaling 0.2 (1/5 = 0.2). The first example is not a high-value customer.

The three variables, RFM, comprise inadequate information to anticipate how to best market to complex humans. People are creatures of habit yet can change on a dime.

The ultimate question for marketers is, “how do you create customer segments of one person?” So, a retailer with 250,000 customers should aim to have 250,000 individual campaigns.

Going beyond RFM

The ultimate goal is to run campaigns to individuals. This was highlighted in 1993 with the book: The One to One Future: Building Relationships One Customer at a Time. That was before AI, before the internet – but the future was clear. Driving to one-to-one marketing – especially with existing customers is expected by today’s consumers.

With that said, brands armed with customers’ zero- and first-party data can at least find similarities between customers to create micro-segments: and not three, five, or ten segments, but hundreds.

Our data from working with hundreds of online brands revealed that brands that market to more than 300 customer segments per week outperform those marketing to only 100 segments per week by 74 percent.

To increase the number of customer segmentations, brands must go beyond RFM. Successful brands now use the following measures to target different sets of customers:

• Discount ratio – total discounts / total order amount,
• Churn factor – time since last activity/activity frequency,
• Item ratio – total items purchased / number of purchases,
• Returns ratio – total items returned / total items purchased,
• Retail ratio – total purchases in-store / total purchases,
• eCommerce ratio – total purchases online / total purchases.

RFM measurements, like discount ratio, still have value to help weed out the most ardent cherry-picking shoppers.

Marketing Technology News: 5 Ways to Create a Breakthrough Moment For Your Brand

Beyond the discount ratio, here is some guiding information that can be gleaned from select segmentation analytics:

1. Churn Factor

While several studies show that retaining a current customer is much more profitable than recruiting a new one, customer retention is even more critical in light of a mercurial economy with an on-and-off recession threat. Recession-conscious customers often reduce spending and will want incentives to remain loyal customers.

Not all consumers engage with brands with the same cadence. Being able to identify when consumers next “beat” is coming up or when they just miss their expected one, can help modern marketers improve their one-to-one marketing, significantly.

Targeted discounting or other incentives for your highest value (and highest margin) customers are warranted to retain their business. But you need to know when these people are in true danger.

2. Retail Ratio/E-commerce Ratio 

For purposes of segmentation, these two ratios are interrelated. Marketers should examine a customer’s purchase history based on the place of purchase. All online? All in-store (increasingly uncommon)? A hybrid mix includes customers who buy online and at physical stores and those who order online but sometimes or always pick up at the store.

Segment customers who fall into the hybrid models as omnichannel customers. Similarly, segment the online-only customer and the in-store-only customer. However, keep close tabs on either of the last two changing (most likely to omnichannel) and be ready to shift your marketing to these customers if such a shift occurs.

The point is for marketers to know how each customer shops. For example, a retailer should not ask an online customer only if they want an item delivered or a store pickup – especially if the nearest store is 100 miles away.

3. Returns Ratio

A returned item can often be the gateway to a best new customer for a retailer.

Returns are a fact of business. And it is a chance to build rapport with the customer by making the return efficient and easy. Marketers learn more about the customer’s product preferences making a return a golden opportunity.

It starts a deeper conversation with the customer than an RFM model. Customers with similar returns for the same reason may create a customer segment. Plus, customers with a high percentage of returns based on a particular issue, such as product damage, wrong size, etc., can indicate a problem with a supplier, the supply chain, online sizing (for clothes) recommendations, or more.

Marketers must segment their customers to offer more targeted, personalized marketing. RFM data is needed, but on its own, it is inadequate. RFM is limited data that can be misinterpreted, resulting in garbage communication to customers. Basing marketing on recency, frequency, and monetary value to segment customers for targeted communications is a dull instrument in the era of one-to-one marketing. The idea of one-to-one marketing has been around for 30 years. Now we have the technology to deliver on it. No reason for GIGO marketing anymore.