Skip to content

Audience discovery: metrics and calculations

Understand the terminology, definitions and figures on Audience discovery, and how they're worked out.

All figures are forward looking, what happens in word-of-mouth clusters after detection, not a record of what's already happened. See how we measure over time.

Areas

Word-of-mouth areas are postcodes detected as sitting inside a word-of-mouth cluster for your organisation, at a particular time.

Baseline / non-word-of-mouth areas are everywhere else: postcodes where we haven't detected word-of-mouth for your organisation at a particular time.

Mosaic Group

One of Experian's 18 Mosaic UK Groups (A - Lavish Lifestyles through R - Elderly Essentials). See What is Mosaic 8?.

A postcode is rarely 100% one Mosaic Group, so a postcode's new customers, revenue and households are split across every group it contains, in proportion to that group's share of households in the postcode.

New customer

Anyone whose first transaction with your organisation inside the sales history you've uploaded falls inside the measurement window. For better measurement, upload your full sales history.

The three metrics

Each is calculated separately for a group's word-of-mouth households and its non-word-of-mouth (baseline) households.

New customer rate: new customers per 1,000 households, per quarter.

\[\text{New customer rate} = \frac{\text{New customers}}{\text{Households}}\]

Lifetime value: the average lifetime spend of the customers acquired in the window. That's everything they've spent with your organisation since, not just what they spent during the window itself.

\[\text{Lifetime value} = mean(\text{Lifetime value})\]

New revenue: Lifetime value and new customer rate combined. How much all time revenue a postcode is generating per household.

\[\text{New revenue} = \text{Total lifetime spend of new customers, normalised per 1K households} = \text{Lifetime value} \times \text{New customer rate}\]

Index and the all-area average

Every figure is shown as an index rather than a raw number. 

100 is your organisation’s own all postcode average for the metric you're looking at. Every household, word-of-mouth and non-word-of-mouth, across every Mosaic Group combined. It is the average behaviour towards your organisation not the make up of the population overall.

\[\text{Index} = \text{round}\left(\frac{\text{Group's word-of-mouth or baseline value}}{\text{All-area average}} \times 100\right)\]

An index of 150 means that segment of the group is worth 1.5 times your brand's own average household. An index of 80 means it's worth 0.8 times the average.

BASELINE and WOM

The two index numbers shown against each row, BASELINE is the non-word-of-mouth index, WOM is the word-of-mouth index.

The bar

The lower value fills it first and the higher one continues from where it ends, so the colour at the end tells you which has the higher value. Magenta if word-of-mouth is adding value, dark grey if it is a shortfall.

Word-of-mouth lift

How many times higher the WOM index is than the BASELINE index.

\[\text{Word-of-mouth lift} = \frac{\text{WOM index}}{\text{BASELINE index}}\]

A lift of ×1.3 means the word-of-mouth areas are worth 1.3 times the baseline. A lift below ×1 means word-of-mouth areas are worth less than the baseline for that group.

Rank

The 18 groups are ordered by their overall value for the selected metric. This is word-of-mouth and non-word-of-mouth households combined, not the WOM or BASELINE figure alone. figure alone.

Rank is calculated separately for each of the three metrics.

All-area average line

The dotted vertical line, by definition it's at index 100.

Selected WOM line

When you select a group, a yellow line will go to that group's WOM index across every row, so you can compare it against every other group.

Discovery card

The card in the selected-group panel headed DISCOVERY. It tells you where this group sits relative to a demographics-only ranking:

  • Strong on its own: this group is ranked first overall, and word-of-mouth still adds further value on top.
  • Overlooked audience: word-of-mouth lifts this group above its own baseline, but other groups rank above it overall. A demographics-only plan would target those groups first and miss this one. The card names all of those higher ranked groups.
  • Weaker in WoM areas: word-of-mouth areas are worth less than the baseline for this group, so there's no discovery opportunity here.

Refine

The percentage of a group's households that sit in word-of-mouth areas.

\[\text{Refine %} = \frac{\text{Households in word-of-mouth areas}}{\text{Total households in the group}} \times 100\]

A low percentage means the word-of-mouth premium for that group is concentrated in a small share of its households - useful to know before you put your spend towards it.

How we measure it over time

Rather than one snapshot, each figure is the average of many windows across your sales history.

Each window covers three months and steps forward a month at a time, so one strong seasonal quarter doesn't skew a group's headline figure.

The earliest part of your history is set aside before this starts, to reduce the impact of large numbers of new customers in this period in cases when a full sales history is not available.

Word-of-mouth status is recalculated for each window, using only the sales data available up to that point, so later windows never use anything that wasn't known at the time.