Breed-aware product personalization
Kohepets tailors product discovery to each pet.
Breed, life stage and stated preferences shape a personalized shortlist, with catalog rules keeping every recommendation grounded in what each product actually offers.
The impact
26%
higher recommendation engagement
Breed-aware product personalization
Case study facts
- Industry
- Pet retail & pharmacy
- Team
- Marketing
- Systems connected
- 3
- Time to live
- 2 weeks
- Engagement
- AI implementation
The impact
Results for the team
- higher recommendation engagement
- 26%
- higher recommendation engagement
- Owners engage more with recommendations matched to their pet's breed, life stage and stated preferences.
- higher average order value
- 18%
- higher average order value
- Average order value tracks the amount customers spend per order as recommendations make relevant products easier to find.
- more product discovery clicks
- 31%
- more product discovery clicks
- Shoppers click through to more products from the personalized shortlist, extending discovery across the catalog.
Situation
A broad catalog. Different pets.
A broad catalog asks owners to do the filtering themselves. Generic recommendations miss useful differences in pet size, life stage and preferences, while breed alone is too limited to describe an individual animal.
Challenge
Breed alone was not enough.
The challenge was to turn pet information into useful recommendations without assuming every animal of a breed has the same needs. Product suitability, stated preferences and missing information all needed explicit handling.
What we built
Pet context connected to product discovery.
The recommendation engine combines owner-provided pet information with structured product attributes. Eligibility rules narrow the catalog before ranking products for relevance.
Missing information prompts a useful question or a general fallback, and every match comes with an explanation. The system never infers a diagnosis or recommends prescription products as ordinary retail purchases.
- 01
Capture pet context
Use the details owners choose to share about each pet.
- 02
Structure the catalog
Map every product to attributes like size, life stage and intended use.
- 03
Enforce eligibility
- Eligibility rules
Unsuitable or restricted products are filtered out before ranking begins.
- 04
Rank the best matches
- Contextual ranking
Blend stated preferences with purchase and browsing context where permitted.
- 05
Explain every pick
Each recommendation shows the product attributes behind it.
- 06
Learn from feedback
Preferences and purchases sharpen future recommendations.
Evals and guardrails
Rank relevant products. Enforce eligibility.
Recommendation evals
Measure how many top recommendations match reviewer-checked pet profiles. Test missing details, conflicting preferences and new customers.
Evidence evals
Check explanations against approved product facts. Test whether breed information alone leads to unsupported health or suitability claims.
Guardrails
Filter out unsuitable and restricted products before ranking. Use general recommendations when pet profile details are missing.
Results
Owners get a more relevant shortlist.
Owners engage more with recommendations matched to their pet's breed, life stage and stated preferences. Average order value tracks the amount customers spend per order as recommendations make relevant products easier to find.
Shoppers click through to more products from the personalized shortlist, extending discovery across the catalog.
- higher recommendation engagement
- 26%higher recommendation engagement
- higher average order value
- 18%higher average order value
- more product discovery clicks
- 31%more product discovery clicks
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.”
Build Capacity To Grow. Own Your Intelligence.
Bring the workflow slowing your sales, orders or delivery. We'll map the bottleneck and show you what the AI system could look like.