New ResearchCorpBenchSee the results ↗We tested AI models on real marketing tasks.
Kohepets
Pet retail & pharmacyMarketing

Promotional audience and offer personalization agent

Kohepets connects each promotion to a more relevant audience.

A promotional personalization agent matches approved offers with customer interests and purchase history. It prepares audiences and message variants while checking product availability, offer eligibility and overlapping campaigns.

Summarize with AI

The impact

57%

reduction in promotion preparation time

Promotional audience and offer personalization agent

Case study facts

Team
Marketing
Systems connected
Not published
Time to live
Not published
Engagement
AI implementation

The impact

Results for the team

reduction in promotion preparation time
57%
reduction in promotion preparation time
Preparing audience proposals and offer context together is expected to reduce promotion setup time.
message variant capacity
2.7×
message variant capacity
The agent could prepare more relevant message variants within the same production time, giving marketing more options to review.
reduction in offer review time
39%
reduction in offer review time
Checking variants against structured offer terms is expected to reduce manual review and correction work.

Situation

One promotion had to serve different customers.

A broad promotion can reach customers whose pets, purchase patterns or product interests do not match the offer. Preparing different audiences and relevant messages by hand makes each campaign harder to coordinate.


Challenge

Relevance must respect the offer’s actual rules.

Personalization must not invent a discount or change its conditions. The agent needs to match eligible products and customers, preserve exclusions and dates, and avoid sending conflicting promotions to the same audience.

What we built

Approved offers meet relevant audiences.

The agent reads the approved promotion brief, product eligibility and permitted customer signals to propose audience segments. It connects each segment to a clear reason for the offer and flags gaps in the brief before drafting.

It prepares segment-specific messages using the same approved offer terms. Marketing reviews the audience logic and variants together; availability, exclusions and campaign overlap are checked before activation.

AI tech stack

  • Amazon BedrockAI inference
  1. Read the approved offer

    Capture eligible products, discount terms, dates and exclusions from the promotion brief.

  2. Find relevant audiences

    Use purchase patterns and permitted pet-profile details to suggest customer groups.

  3. Validate the match

    Check product availability, audience eligibility and overlapping promotions.

  4. Prepare message variants

    Adapt product context and copy while preserving the approved offer terms.

  5. Approve the promotion

    Present audiences, eligibility checks and messages for marketing review.

  6. Review segment performance

    Compare engagement and purchasing patterns without treating attribution as causal lift.

Evals and guardrails

Preserve offer terms. Review audience fit.

  • Audience evals

    Compare segment membership with checked examples. Test missing profiles, conflicting preferences and customers eligible for multiple offers.

  • Offer evals

    Check every message against approved products, discount rules, dates and exclusions. Test ambiguous or incomplete briefs.

  • Guardrails

    Prevent invented discounts and unsupported product claims. Enforce consent and campaign limits; require approval before activation.

Results

Faster promotion preparation and review.

Preparing audience proposals and offer context together is expected to reduce promotion setup time. The agent could prepare more relevant message variants within the same production time, giving marketing more options to review.

Checking variants against structured offer terms is expected to reduce manual review and correction work.

reduction in promotion preparation time
57%reduction in promotion preparation time
message variant capacity
2.7×message variant capacity
reduction in offer review time
39%reduction in offer review time

From Kohepets

“We have worked with Revensi for many years. In that time they have made vast improvements to how our business runs, and they are great people to work with. We highly recommend them.”
Kim HeongCo-Founder at Kohepets

Build capacity to grow. Own your intelligence.

Bring the workflow slowing your pipeline, campaigns or renewals. We'll map the bottleneck and show you the AI agent concept.