What Petbarn deployed
Petbarn, the pet retail arm of Green Cross Pet Wellness and one of Australia's largest pet chains, has automated its planogram management using Crisp's AI-enabled planogram technology integrated with Blue Yonder's Category Knowledge Base. The retailer operates 267 stores with up to 70 bays of shelving each, and had been producing planograms, the detailed shelf layout plans that dictate where every product sits, largely by hand for categories like pet food that see frequent SKU changes and heavy localization by store.
The results are the kind of concrete operational numbers that rarely show up in AI retail announcements. Planogram creation for a prescription pet diet category dropped from three weeks to one day. The system now processes more than 240 planogram updates every hour, and mass revisions that used to take hours to push across hundreds of stores now complete in roughly 30 minutes. Store-level feedback reportedly rated the automated planograms as the best-received layouts the chain has shipped, which addresses the usual worry that automation trades speed for merchandising quality.
Why planogram work was the actual bottleneck
Space planning is one of the least visible parts of retail operations, but it directly gates how fast a chain can open new stores, respond to a supplier's new product line, or react to a category reset. Petbarn had added 13 new stores recently on top of its existing 267, and manual planogram production simply could not keep pace with that expansion rate while also servicing the ongoing SKU churn inside categories like pet food, where new formulations and package sizes arrive constantly. A three-week turnaround for a single category reset is a real constraint on how quickly a retailer can act on anything, from a supplier deal to a competitive response.
That is the pattern worth noticing across this category of AI deployment: the highest-leverage automation targets are often not customer-facing at all, they are the internal workflows that quietly cap a retailer's operating speed. A chatbot that helps shoppers find a product is visible and easy to demo, but a planogram engine that turns three weeks of manual layout work into one day changes how fast the entire business can move, which is a much larger structural advantage than most AI pilots retailers announce.
What the automation actually does
The system automatically positions new products within existing brand blocks, color-codes what changed for easy human review, and surfaces sales performance, ranking, and stock-level metrics directly inside the planning workspace rather than requiring planners to pull that data from a separate system. That design choice matters: it keeps a human planner in the loop making merchandising judgment calls, while removing the purely mechanical work of manually placing products and cross-referencing performance data from multiple systems by hand.
This is a meaningfully different automation pattern than the fully autonomous AI agents getting most of the retail industry's attention right now. It is narrower, it targets a well-defined operational task with clear inputs and outputs, and it is far easier to validate because the output, a shelf layout, is something a human can inspect and approve before it ships to stores. That combination, real speed gains plus a built-in human checkpoint, is likely why store feedback on quality held up rather than degrading.
What this signals for retail operations AI spending
Most of the AI investment conversation in retail right now centers on customer-facing shopping assistants and agentic checkout, categories where the return on investment is genuinely difficult to isolate and where competitive pressure often justifies the spend before the ROI case is fully proven. Petbarn's planogram automation is the opposite: a narrow, back-office workflow with a before-and-after time measurement anyone can audit, three weeks against one day, and a throughput number, 240 updates an hour, that a finance team can turn directly into a labor-hours calculation.
That auditability is exactly what makes this kind of project easier to fund and easier to defend after the fact than a customer-facing AI initiative. A retail technology leader building next year's AI budget request will have an easier time getting approval for a project with Petbarn's shape, clear inputs, clear outputs, measurable time savings, than for a more ambitious but harder-to-measure generative AI shopping experience, even if the latter gets more attention at industry conferences.
The evidence a CFO will actually believe
Before chasing the next customer-facing AI feature, retail and CPG technology leaders should inventory which internal operational workflows, space planning, replenishment, labor scheduling, still run on manual or semi-manual processes that cap how fast the business can move. Those workflows are often the highest-confidence place to deploy AI first, precisely because the task is well-bounded and the value is measurable in hours and dollars rather than in harder-to-prove metrics like conversion lift or customer satisfaction.
Petbarn's specific numbers, three weeks to one day, 240 updates an hour, 30 minutes for a mass revision across hundreds of stores, are the kind of evidence a CFO will actually believe, because they describe a process getting objectively faster rather than a customer experience getting subjectively better. Any team building a business case for operational AI spend should be collecting and presenting exactly this type of before-and-after data, not general efficiency language, if they want the next budget cycle to say yes.
The takeaway for the reader's roadmap
That discipline compounds over time. A retailer that can point to two or three narrow, well-measured operational AI wins builds far more internal credibility for its next, harder-to-measure AI initiative than one that leads straight with an ambitious customer-facing bet and no track record to back it up when a board member asks for proof. Sequencing matters as much as the technology choice itself, and the sequencing that works starts with the workflow easiest to measure, not the one easiest to demo at a conference.
Petbarn's own next move is worth watching too. Having proven the model on planogram management, the natural extension is applying the same narrow, measurable automation pattern to adjacent workflows like labor scheduling or promotional resets, where the same combination of well-bounded inputs and auditable outputs should make the business case just as easy to build and just as easy to defend to a skeptical finance team six months from now.
