Most people look at retention as something they measure annually – which misses the opportunity to respond to early warning signs in your installed base. So when I read this article over on Chief Marketer I got suitably excited … because it is right on in pointing out how to measure retention using two metrics that are much more immediate:

• dollar retention rate
• replacement rate

Dollar retention rate
A customer who spent an average of $1000/year drops to $50/year. Is the customer retained? She didn’t leave, so traditional retention measures would show no churn. Clearly this doesn’t capture the erosion in value. Instead, use “dollar retention rate,” which measures the percentage of revenue retained over a prior period. So where you may have an 80% traditional retention rate in your best customer segment, if half of those retained drop 95% in value like the example above, your dollar retention rate is only 42% (40% retained at prior value and 40% retained at 5% of prior value). This means you need to replace 58% of your best customer revenue from somewhere else.

Replacement Rate
You had 10,000 customers in your top segment. You lost 500 in the last month. How many new ones entered the top segment? If its 500, you have at least held your ground. Any fewer and you have to account for the lost revenue elsewhere.

Source
Retention Metrics You Should be Using

It’s been said many times: “you get what you measure.” The choice of which metrics to use is a strategic one and should be made with care. Do not measure results based on volume alone e.g. the number of leads generated, number of new customers acquired. Volume alone tells you very little about either efficiency or effectiveness.

The most meaningful metrics are those that get to the value of the customers you acquired in a given program. Examples include ROI and LTV. If your company does not have the infrastructure to measure ROI or LTV, you may need to rely upon other metrics. A best practice is to balance your metrics to include data on both effectiveness as well as efficiency.

The testing plans provided elsewhere on this site are designed to give you guidance on which metrics your company uses when evaluating different types of campaigns.

Best Practices
The best metrics are those that are normalized for dollars spent, which is just a fancy way of saying that the metrics use the dollars spent on the campaign as the numerator. Normalization is essential to allow for results to be read and compared across campaigns.

Another best practice is to develop a standard set of assumptions you’ll use consistently in analyzing and reporting on campaign results.

The conversion rate is a critical input into ROI calculations. More recently, “conversion” has taken on a special meaning in Internet and Search Engine Marketing. Click here for a definition of the term from that perspective. Click here to check out “new school” thinking on conversion rates.

Where x is the marketing activity of interest (leads, prospects, inbound calls, unique visitors, in-store traffic), y is the time period of interest, and z is the buying window. The buying window is typically determined by interviewing sales personnel as well as customers to determine how long it takes to close the sale.

A prerequisite for measuring conversion is a closed-loop tracking system, that enables you to match back today’s sales to yesterday’s marketing activities.

This can be easy or hard depending on the length and complexity of your sales process. All things being equal, the conversion rate tells you how effective your marketing is at generating a volume of leads, traffic, or qualified prospects and in turning that volume into sales.

Best practices
The conversion rate can be used as a diagnostic: as you move from unqualified prospects to qualified leads the conversion rate should increase. If the conversion rate does not go up in this manner, there is likely to be something wrong with the criteria you are using to qualify leads. Likewise, ATBE (all things being equal) the conversion rate should go up over time — as your company gets better at separating the wheat from the chaff (prospects from leads). Comparing conversion rates in different types of selling situations can also be useful. For example, it is often thought that different types of customers are more effectively handled through different sales channels. You can test this hypothesis by taking leads of equal quality and tracking the conversion rate through Channel A vs. Channel B. Note that this only works if the leads you are tracking are of equal quality, which means developing a knowledge-base around what characteristics of your customer drive sales.

This metric measures the efficiency of a given program when it comes to delivering orders of your product or service.

Best Practice
The cost per order (CPO) isn’t the best metric to use when comparing the results of different campaigns. Why? Because CPO does not differentiate between the size of the order. ROI is a much better measure as it gets to both effectiveness and efficiency in one number.

Customer Placement Value (CPV) is an relatively new alternative to customer lifetime value (LTV). While LTV requires that you calculate the retention rate and complete a detailed analysis of cohorts, CPV is much simpler. Calculations are based on the notion that the value of a customer can be estimated based on the bundle of goods and services sold to them at one point in time. Some of the items may be purchased in the current period. Some of the items in the bundle may be committed to up front but will not be purchased until the future. An example is the purchase of a printer with a 3-year service contract.

Proponents claim that CPV is a better metric to use in the “real world” where it may be difficult or impossible to calculate “true” lifetime value.

Best Practice
The people who created CPV argue that it is a worthwhile substitute for LTV particularly in situations where the data needed to calculate a true LTV is missing. We’re not sure what to think about CPV. On the one hand, this metric comes closest to capturing what we here at Firewhite call franchise value, by looking at the value created when you sell a customer a bundle of products and services that they will use (& pay for) over multiple time periods. On the other hand, CPV doesn’t really value customers per se but instead puts a value on the products we sell to them. This can lead to erroneous decision making and – in the worst case – ultimately bankrupt the company through a phenomenon known as product death spiral. In short, use at your own risk.

Relevant Link

• How to calculate CPV*
• CRM Metrics Framework

*Free site registration may be required to access this material

The conversion rate is a critical input into ROI calculations. More recently, “conversion” has taken on a special meaning in Internet and Search Engine Marketing. Click here for a definition of the term from that perspective. The “old school” thinking on conversion rate is is available here

Where volume is measured based on the marketing activity of interest e.g. leads captured, software downloads, sales of a specific product promoted. Often conversions are measured by creating a specific landing page for each promotion and for each media choice within the promotional plan and measuring results by landing page. With search engine marketing, a best practice is to create specific landing pages for each search engine term or phrase included in the campaign.

Related Links

• Landing Page Optimization
• Conversion Rate Optimization
  Research note and literature review from Marketing Experiments

Lifetime value (LTV) is used to answer the question of how much a marketing manager can afford to spend to acquire a new customer – and still make money on that customer. To estimate lifetime value, you need to be able to estimate how long a new customer will remain with you, what purchase behavior will this new customer have over the lifetime of the relationship, and what the contribution margin is for different types of customers and/or products.

In this example, customers were acquired in year 0 and are known to have a life expectancy of three years.¹

• n0 – is the number of new customers acquired.
• m0 – m3 – is the contribution margin delivered by each customer (on average) in years 0 – 3, respectively.
• r – is the average retention rate seen over the period.
• i – is the discount rate used in net-present value calculations … representing the time-value of money.

A complicating factor here is that acquisition and retention are a continuous process. Analytically, this means that you need to isolate customers who came onto your file at the same point in time and look at the retention rate and contribution margin within that cohort — something called “cohort analysis”.

Best Practices
LTV has been around for 30 years. Many companies don’t understand LTV models and are therefore reluctant to use what these models tell us about the value of different customers to drive decision making. We think that is too bad. LTV is most useful when used in conjunction with segmentation, to identify customers you need to treat differently so as to maximize their value to the business. This is best illustrated by example:

• An online investment company classifies its customers into more than 10 segments based on their profitability to the company. Different segments are treated very differently in terms of the frequency of communications, the type of offers that can be used to intervene when a customer plans to defect, as well as the offers and fee structure that applies.
• An email service provider now rates its 2 million customers, with top customers getting special attention from service reps and senior management, including an annual call from the CEO. This has reduced customer attrition by 50% in the last five years and has trimmed the number of unprofitable customers by 6%.

The calculations used in LTV can get complicated very quickly.

Related Links

• Academic white paper from Columbia Business School
• Lifetime Customer Value Calculator from Harvard GSB
• Actual example of a lifetime value calculation from NetFlix – presented as part of their IR (investor relations) materials.

If you sell on a subscription model that expires once every period, a key metric is the renewal rate. Renewals are typically thought of in terms of a “block of expires” where each block represents a set of subscriptions of equal value that is set to expire at the same time.

Think of a company like TiVo which sells not only hardware (Personal Video Recorder or PVR) but also a monthly service (program guide). The subscription service can be purchased on a monthly basis or for the lifetime of the PVR. Calculating the renewal rate is very straightforward as the above calculation indicates.

Best Practice
The renewal rate measures how effective you and your organization are at getting customers to sign up for your service when it expires. A low renewal rate is negatively related to both customer value and profitability.

Related Links

• 2003 Summit on paid-content subscriptions
• Marketing paid-content subscriptions

Typically calculated for one period like so:

The length of the period may be measured in months or years depending on the particulars of your product and how it is purchased and/or consumed.

Retention starts out at 100% and decays over time. Retention tends to hit a trough after a certain number of periods. See – for example – the retention curve on the right which is typical for demand deposit accounts (a.k.a. “consumer checking accounts”) in financial services. Retention goes from 100% in year 1 to 80% in year 2 to 75% in year 3 to 70% in year 4 to 60% in year 5. After 5 years, retention stabilizes and stays flat at 60%.

In a perfect world, your company would track purchasing behavior by cohort. A cohort is a group of distinct customers who entered your file (or started buying from your company) at the same point of time. Of course, most of us don’t live “in perfect.” Recently JD Powers looked at the retention rate for new buyers of cars. Here they analyzed the behavior of 117,000 new car buyers in 2003 regardless of when these buyers were last in the market for a new car. Measured in this way, the retention rate is more akin to brand loyalty. This methodology can make sense when it not possible to track customer behavior by cohort or where such tracking is highly imperfect.

The retention rate measures how effective you and your organization are at holding onto customers. Retention is a key driver of customer value and profitability. The impact of changes in retention don’t always show up in your bottom line immediately.

If you sell on a subscription model, most likely you don’t look at retention as a key metric but instead look at churn.

Best practices
Well designed retention metrics can be used for:

• Diagnostic purposes Decreases in the retention rate over time may signal a drop in product quality, a problem with your pricing, and/or a new competitor with an offering that is a better fit for the customer’s needs than your offerings.
• Predicting & Managing Customer Defections By analyzing customers who have defected you may be able to identify predictable patterns. These patterns – in turn – can be used to built predictive models that allow you to identify customers who are likely to defect early enough that you can do something about it.

This metric measures how successfully a campaign encourages opt-ins for additional communications via email. Permission-based marketing is the least expensive way to create an ongoing dialogue between your company and large numbers of customers and/or prospects. The opt-in rate tells you how effective your EM or DM efforts are at encouraging end users to opt-in to further permission-based marketing.

Best Practice
Obviously the time period used here – at 10 days – is arbitrary. Pick a time period that makes sense based on when the majority of opt-in requests come in following a flight of communications.

Trend data can be very useful here. If you see the opt-in rate going down over time, it could mean that you are overcommunicating with your customers and/or prospects.

This metric measures the efficiency of a given campaign when it comes to delivering new customers. Note that this metric assumes you have a way to attribute new customers acquired back to specific campaigns using match back or via some form of a closed-loop marketing system.

Best Practice
Most companies spend more on customer acquisition than they see in initial revenue. Knowing how much you are spending on new customer acquisition can be critical to your profitability.

Return on Investment (ROI) analysis generally started by looking at the incremental revenue generated by a particular vehicle. Adjustments to revenue are typically made to account for COGS, known service and support costs, and for any any sales incentives that are fielded in conjunction with a particular mailing program.

ROI is generally expressed as number ranging from 0 to infinity. An ROI of 2x means that for every $1 invested in the program we got back $2 in incremental contribution.

Best practices
Not all of the revenue that comes in off a given program is truly incremental. Deciding how much revenue to count as incremental is more of an art than a science. The best practice is to be consistent with your assumptions so as to allow for comparison of ROI over time.

The best way to attribute revenue to a given mailing program is through a procedure called match back. Of course, not every business has the closed-loop measurement system that are a prerequisite for match-back analysis. With a match back, we take the list we mailed to and match it back to actual sales transactions. If the customer we reached in our mailing purchased in the time period of interest, we attribute the revenue to the mailing program. The time period of interest should correspond to the normal purchase cycle, generally measured in days. Be aware that many customers buy as a result of your communication even after the promotional window has expired.

One nuance here is to make sure that you are properly accounting for frequency. If your sales cycle is a 6-month period, you are likely to touch customers multiple times during any given 6 months. Attributing sales to any one mailing program or promotion may not make sense. Instead, consider analyzing results within each group of customers who received a certain frequency level.

You might call this a natural experiment as opposed to a controlled test, in the sense that you didn’t necessarily plan to expose different customers to different frequency levels. Unless there is a random control group, use care when interpreting results. Customers who received a higher-level of frequency may differ in some important ways to those customers who received a lower-level of frequency.

Notes

1. Be careful here to measure conversion based on the number of unique customers you touched through the mail↩
2. Calculated either based on the types of products or services purchased or based on the type of customer↩

If you sell on a subscription model, most likely you don’t look at retention as a key metric but instead look at churn. Churn can be calculated on a monthly, quarterly, or annual basis.

• where D = days between initial and current customer count
• and NCA = new customers acquired during the period
• and C1 = initial customer count
• and C2= current customer count

As an example, a carrier exits December with 5,000 subscribers. In January 100 of these subscribers leave and 600 new customers are added. As a result, the carrier exits January with 5,500 subscribers.

The churn rate for January is 2.0% or 100/5000. This amounts to an annualized churn rate of 24.3%. (While churn is a negative for your business, typically the churn rate is NOT expressed as with a negative sign in front of it.)

Alternatively, you can look at churn relative to when the customer must resubscribe to your service. This is called the renewal rate.

Best practices
Churn is a negative. The less churn you have the better your organization is at holding onto its customers. All things being equal, a reduction in churn will drive incremental customer value and profitability.

When your churn rate is dropping it’s cause to celebrate, right? For example, recently WebEX reported that churn had dropped from 4.0% in Q3 2003 to 2.8% in Q4 2004. Before you break out the champagne, take an in-depth look at the numbers. In a growing business, churn rates tend to move in only one direction: down. Even so, churn can be useful as a diagnostic, particularly when it comes to looking at trends over time as well as how churn varies by segment:

• Trends over time If your churn rate suddenly increases this may signal a drop in product quality, a problem with your pricing, and/or a new competitor with an offering that is a better fit for the customer’s needs than your offerings.
• By Segment Certain segments may be associated with a higher rate of churn than others. Understanding this can allow you to decide how much to invest in loyalty marketing for different types of customers. For example, the average consumer changes wireless carriers every 18 months* versus 36 months for business-to-business customers.
• Competitive Offers Many companies address churn directly with offers that match the competition. Examples in high technology include competitive upgrades, which are often designed to get customers using Intuit’s QuickBooks – for example – to upgrade to Microsoft Money. These types of offers can be an effective way to keep customers in their seats. That said, it is important to understand the value of the customers you keep. Bottom feeders – that is customers who routinely switch from one company to the next to get the best deal possible – may not be worth keeping.

This metric measures the efficiency of a given campaign when it comes to generating leads as opposed to prospects. 

Best Practice
Typically, it is a best practice to define prospects as meeting a lower standard of qualification than leads.  Each
company defines the various stages of the sales funnel somewhat
differently.  Companies also vary in terms of the standards they
use when determining when a prospect (sometimes called an inquiry)
becomes a qualified lead.  For example, your company might define
a prospect as anyone who fills out a entry form for more information on
your website or in response to one of your direct mailing
efforts.  A lead – on the other hand – might be defined as anyone
who provided us with the following 5 pieces of information:

• Name, Job Title, Company Name, Website, Telephone Number, Email
• Role in buying process
• Time frame when purchase is expected
• Size of budget available for solution set
• Preferred method of contact – by email or phone and if by phone at what time

Cost per Lead (Qualified)

This metric measures the efficiency of a given campaign when it comes to generating leads as opposed to prospects.

Best Practice
Typically, it is a best practice to define prospects as meeting a lower standard of qualification than leads. Each company defines the various stages of the sales funnel somewhat differently. Companies also vary in terms of the standards they use when determining when a prospect (sometimes called an inquiry) becomes a qualified lead. For example, your company might define a prospect as anyone who fills out a entry form for more information on your website or in response to one of your direct mailing efforts. A lead – on the other hand – might be defined as anyone who provided us with the following 5 pieces of information:

• Name, Job Title, Company Name, Website, Telephone Number, Email
• Role in buying process
• Time frame when purchase is expected
• Size of budget available for solution set
• Preferred method of contact – by email or phone and if by phone at what time