Glossary

Accelerator

A customer whose rate of purchasing is increasing

Active customer

A customer whose most recent purchase falls within a specified period, typically the last 12 or 24 months; that is, a customer whose recency is less than the inactivity period

At risk

Customers whose buying behavior is deteriorating; they may be purchasing from a competitor

Attrition

Gradual, negative erosion of a customer's score; such a customer is sometimes called an ‘attritor’ or ‘defector’ (if they are no longer active)

Back test

A way to validate predictive analytics results by predicting purchasing behavior for a period in the recent past for which actual transactions are known and then comparing the predictions with the known behavior

Cross-sell

Offering customers products or services they have not previously purchased

Control group

In a marketing campaign, a set of customers who do not receive the campaign offer; used to determine the efficacy of the offer

Deceleration

The rate at which a customer's velocity is slowing down

Deceleration percentile

Where a customer's deceleration ranks among all customers; a deceleration percentile of 97% means that only 3% of all customers are decreasing their purchasing at a higher rate

Decelerator

A customer whose rate of purchase is slowing down

Defector

A customer who is no longer active, whose recency is longer than the inactivity period; see also attrition

Download

Transferring data from a remote computer to your computer; see also upload

ERP

Enterprise Resource Planning, a set of programs to manage all aspects of a business including Accounts Payable, Accounts Receivable, General Ledger, Manufacturing, Inventory, etc.

Factor

A variable of interest in an experiment; for example price, discount level, expiration date, etc.

Factorial design

A type of experiment design with two or more factors, each with discrete values or “levels”, where all possible combinations of levels across all factors are tested simultaneously

Finite State Machine

A mathematical model of an entity that describes its behavior as a result of its past history and current inputs. If an example entity is a customer, inputs might be direct mail promotions and responses could be a purchase or changed loyalty score.

Fractional factorial design

A variation of a factorial design experiment in which only a subset of all possible combinations of factor levels is tested; less runs are needed to conduct the test, but at the cost of a loss of information

Frequency

The number of purchases made in a defined time period, ignoring differing interorder wait times; for example, five times per year

FTP

File Transfer Protocol, a standard way of exchanging files over a network that supports the TCP/IP protocol, such as the Internet

Inactivity period

A company-determined time interval without a purchase beyond which the customer is no longer considered to be active; typically set to one or two years

Interorder wait time

Time between a customer's consecutive purchases; also called Time Between Sales

Lifetime value

The future revenue expected from a customer; usually the revenue expected in the next one or two years, since that is what can be predicted with reasonable accuracy

Likely buyers

Those customers who are expected to buy something in the next period (month or quarter)

Likely Buyer Score

The percent probability, from 0 to 100, that a customer will buy something in the next period (month or quarter)

List select

The process of selecting a subset of the entire customer population for a marketing campaign

Loyalty

In the context of ‘customer loyalty’, a measure of the likelihood of a purchase in the near future

Loyalty decile

One-tenth of the customer population; after customers are segmented by their loyalty score, those in the highest 10% (i.e. the best customers) are said to be in the tenth decile, those in the lowest 10% are in the first decile

LRank

The relative percentile rank of a customer compared to the rest of the customer population based on the customer's loyalty score, scaled from 0 to 100

LScore

The absolute loyalty score assigned to a customer based on the Loyalty Builders analytical model, scaled from 0 to 1000

Mean

The average value of a group of numbers, most meaningful (and near the median number) when the numbers have a bell-shaped distribution but often skewed (and made less meaningful) by a few extremely large values near the end of a distribution of values

Median

The middle value in a set of numbers; if 13 numbers are arranged in order of size, the 7th number is the median

Merge/Purge

The process of cleaning a customer data file by combining records that represent the same customer (merge) and deleting records that are duplicates, inaccurate or no longer needed (purge)

Monetary value

The revenue received from a customer so far, usually in the most recent period; also the ‘M’ in RFM, an older segmentation scheme

Nonlinear

A relationship between numerical quantities is called nonlinear if there is not a constant proportion relating changes in one quantity to changes in the other. Exponential growth, as occurs when compounding interest over time, is an example of a nonlinear relationship.

OFAT

(An acronym for One Factor at A Time, also known as a Split Run); a type of experimental design in which only one factor is changed at a time while the other factors are held constant

Pay-as-you-go pricing

Pricing is based on size of downloaded file and payment is due upon download, rather than a fixed fee paid at regular intervals

Predictive analytics

The science of predicting what customers will do based on analysis of their past behavior

Probability horizon

The time between the date of a purchase probability analysis and some point in the future within which a purchase might be made; typically one or two fiscal quarters

Purchase delay

How many times a customer was expected to purchase between the date of their last purchase and the latest analysis date, used to detect changes in customer purchase frequency

Probability threshold

The percentage purchase probability below which values are ignored (and customers discarded) when building a list of targeted customers likely to purchase a specific product within the probability horizon; identical to probability cut-off

Purchase probability

The percentage likelihood of a customer making a purchase within the defined horizon; sometimes called purchase propensity

Quintiles

The 20, 40, 60, and 80th percentiles are called quintiles

Recency

Length of time since a customer's last purchase

Recipe

A distinct combination of factor levels used in an experiment; for example, if there are 3 variables, each at 2 levels, there are 2x2x2 = 8 distinct recipes

Response rate

Percent of total recipients in a campaign who purchase or otherwise answer the promotion

Retention

Length of time between a customer's first purchase and today

Revenue rank

How the total amount of money a customer has spent compares with that of other customers; a revenue rank of 71% means that 29% of the customers have spent more money during their history with the company

RFM

An older, linear modeling technique that looks only at Recency, Frequency and Monetary value when scoring and segmenting customers; see RFM vs. Loyalty Builders Modeling for more details

Risk probability

The probability, between zero and 100%, that a customer will not make a purchase in the time between the most recent analysis date and the date when their recency equals the length of the inactivity period (in other words, when they become inactive)

Satisfaction

A measure of a customer's past experiences, most likely related to their most recent transaction

Segmentation

The process of dividing a customer population into groups based on their loyalty score

SKU

Shop Keeping Unit, the alpha-numeric code used to designate a specific product

Split run

see OFAT

Standard deviation

A statistical measure of the spread of a set of numbers about their mean value

Suppression

The process of removing certain names from a list of customers to be targeted in an upcoming marketing campaign because of pre-established rules, for example they were recently contacted or do not qualify for the offer

Upload

Transferring data to a remote computer from your computer; see also download

Up-sell

Offering customers more of the products or services they have previously purchased

Velocity

The rate at which a customer is making purchases