Skip to main content

ES Loyalty Boost 2.9 release notes

note

These release notes cover features introduced or enhanced since ES Loyalty Boost 2.8.1.


What's in this release


New features

GenAI Module — Member LLM Score

A generative AI module evaluates members across a set of profile values and past product purchases to produce an aggregated score. This score is used to determine how to optimize offers for individual members.

Member profile values considered:

AttributeDescription
GenderMember's gender
Average Household IncomeEstimated household income bracket
Province of ResidenceMember's province
RFM Level RatingRecency, Frequency, Monetary Value rating — for example, Bronze, Silver, Gold, Platinum, or No Purchase
Last Purchase Date RangeNumber of days since the member's last purchase

Product affinity:

Past product purchases are used to calculate affinity scores for related products. Affinity is expressed on a 0–1 scale. For example, a member who purchased DIAPERS may have the following affinities:

  • WIPES — affinity score: 1.0
  • BABY_FOOD — affinity score: 0.8

See also:


Budget Burn Down

Campaign budgets must be managed to ensure spending stays within defined limits. The Budget Burn Down feature provides controls for tracking and enforcing budget usage across campaign iterations.

What is managed:

  • The list of vendors governed by the budget
  • The duration of the budget
  • The maximum allowable amount for the budget
  • Iteration scheduling (budgets can repeat on a cadence, with each repetition counted as an iteration)
  • Key metrics such as current points earned toward the budget and the start and end dates of each iteration

Budget cap enforcement:

To prevent the budget limit from being exceeded, a configurable threshold — expressed as a percentage of the total budget — determines when further points are blocked from being awarded. The default value is 75%. Once the budget is consumed to this threshold, offers are blocked from being selected.

Example budget parameters:

Start:      2022-07-1T20:13:05.637Z
End: 2022-09-30T20:13:05.637Z
Cadence: WEEKLY
Amount: 100000
AmountType: POINTS

In this example, the cadence is set to weekly. Between the start and end dates, there are eight iterations of the budget, each beginning with the full budget amount specified.

See also:


Redemption Offers

This new offer type incentivizes members to redeem their points during a campaign period. Encouraging redemption has been shown to create more loyal customers and demonstrate the value of the loyalty program.

Example: Redeem 25,000 points, receive 5,000 points.

How it works:

To identify the optimal configuration for each member, a Global Test group is created with different redemption threshold permutations. The permutation selected for each member is the highest available for their current points balance.

For example, for a member with 60,000 points, an offer to redeem 50,000 points is generated if the member meets all offer qualifications — including that at least X days have passed since their last redemption and that they have sufficient points available. If the criteria are not met, the slot is left empty.

Global Control group members with an appropriate points balance receive the lowest possible redemption offer for which they qualify.

A new optimization objective — Redemption — has been added alongside the existing objectives (Affinity, Stretch, and Penetration) to support this offer type.

Success metrics:

  • Number of members redeeming
  • Percentage of members redeeming
  • Percentage of members redeeming at the optimized or offered redemption threshold

Reporting:

Because the objective and success measures for redemption offers differ from traditional spend-and-get offers, they are excluded from Standard Report calculations. Redemption offer data is available in:

  • The campaign configuration section of the Campaign Comparison dashboard
  • A new Redemption Offers dashboard, which includes metrics such as:
    • Spend over redemption threshold
    • Number of members in redemption test versus control groups
    • Offer and member acceptance and completion rates for redemption offers

The goal of this feature is to make redemption offers attractive, reducing rewards liability for the client while increasing member engagement.

See also:


Fixes

Boost Product Affinity Name Mismatch

Fixed a naming inconsistency for products when retrieving product affinity scores for reporting. Mismatched product names were preventing affinity scores from being used correctly in Boost.

For example, the Boost controller stored a product name as L'OREAL while the analytics function expected LOREAL. All instances of this mismatch have been corrected to protect the integrity of reporting.