Customer Flow Segmentation

Purpose

Make Customer Flow Segmentation assignments to all card identified shoppers in the timeframe of the report.

CustomerFlowSegmentation2_segflow

When this report is defined with the TIME selection of complete quarter

  • the report is automatically re-run whenever a new Quarter is complete
  • any shopper groups saved from the report will be re-run with the new Quarter so they stay current, automatically

The segments are assigned after examining the shopper’s performance over three periods of time

  • a “current” period, the quarter selected – below referred to as C
  • a “break” period, the 182 days (half year) preceding the “current” period – below referred to as B
  • a “prior” period, the 182 days (half year) preceding the “break” period – below referred to as P

Instance

A Quarter (3 month timeframe).

Rows

The rows are the segments.  The segments are designed to recognize certain changes in shopping behavior. The segments are assigned based on purchasing in the three defined timeframes P|B|C.

To begin with there is an overall assignment to a segment called “PEANUTS“.  These are shoppers whose average monthly spend in any of the P|B|C timeframes has not exceeded the “peanuts breakpoint” of $20 per month.  These are really low spenders who we may want to think twice about before spending marketing dollars on them.

Then there are the real shoppers.  Here is a table that outlines the logic used to assign segment values.  The notation p,b,c designates purchasing (1), or lack thereof (0), in each of the defined timeframes.

P, B, C, the segment assignment
0, 0, 0, no purchasing so no participation in the report
1, 0, 0, a shopper we have not seen in a year is a LAPSED shopper.
0, 1, 0, a shopper who shopped in “b” but not currently is in DECLINE
1, 1, 0, a shopper who shopped in “b” but not currently is in DECLINE
0, 0, 1, a shopper who shopped in this quarter but we hadn’t seen for a year is considered a NEW shopper.
1, 0, 1, a shopper who left us for 6 months or more but who has returned has been REVIVED
0, 1, 1, a shopper who shopped in both “b” and “c” could be GROWING, DECLINING or STABLE -see more..
1, 1, 1, a shopper who shopped in both “b” and “c” could be GROWING, DECLINING or STABLE -see more..

More..  In those cases where the shopper appeared  in both “B” and “C” we apply these rules.

  • if their spending in Current was less than <= 80% of their spending in Break they are DECLINING
  • if their spending in Current was > 120% of their spending in Break they are GROWING
  • if their spending in Current was between 80% and 120% of their spending in Break they are STABLE

Columns/Metrics

The columns are grouped under two headings.

All HHs and $ Change. This section quantifies, in sales dollars, the change each shopper segment has made to the performance of the store.

  • All Segment HHs – This is a count of all the household currently identified as being in this segment. Different from the #HH metric in the following section which is only those segment Households who shopped in the Quarter of the report.
  • $ Change – This is a time adjusted change in the Dollar spend.
    • For a NEW or REVIVED shopper it is the Dollar spend in the current quarter indicating an increase in the total business of the store.
    • For a GROWING, STABLE or DECLINING shopper it is the Dollar spend in the current quarter less half the half-year spend in the Break. (to equivalize the timeframes) indicating an increase|decrease in the total business of the store.

Shopping in selected Quarter.  This section gives us an understanding of what these shopper segments are doing in the store in the selected quarter. To be included in this section a Household must have sales volume in the selected quarter.

  • #HH – Count of distinct households assigned to the segment.
  • %HH – Count of distinct households assigned to the segment expressed as a percent of the total.
  • Cume %HH – Cumulative percentage.
  • $ – Total dollar spend for the households in the segment.  Dollar spend is the taken from the transaction log as the List Amount + Tax Amount – Markdown Amount.
  • % of $ – the Total Dollar Spend expressed as a percentage of the total.
  • Cume % of $ – Cumulative Total Dollar Spend
  • $/HH – The average dollar spend per household in the segment.
  • $/Trip – The average dollar spend per trip, or transaction.
  • Trips/Week – The average number of times a household in the segment makes a shopping trip in a week.
  • % of of Total Discounts – The discount (markdown) dollars used by this segment, expressed as a percentage of the total discount dollars used.
  • Basket Discount % – The average percentage discount realized per transaction.
  • Use of Discounts Index – 100 * (% of discounts used by the segment) / (% of households in the segment), where 100 would represent a “fair share” use of discounts.
  • Discount $ – The sum of discount  (markdown) dollars used by the segment.
  • Value Index – 100 * (% dollar spend by the segment) / (% of households in the segment), where 100 would represent a “fair share” of dollar spend.

More on Metrics

Household Counts

There are two Household Count columns; All Segment HHs (all), and # HH (hh). The counts can be different. Here are the reasons why, and what those counts represent by segment.

NEW, REVIVED, GROWING, STABLE – The all and the hh are the same because their definition has those HHs shopping in the current quarter.

DECLINING – The all includes shoppers who shopped in the b timeframe and whose purchasing in the c timeframe was <80% of their b shopping.  The counts are not the same because the households who DID NOT SHOP in the C timeframe (so that spending is definately <80%) are included in the all count, but having zero spend in C do not show up in the hh count.