The pet food industry’s relationship with artificial intelligence (AI) is best described as cautious pragmatism: half-curious, half‑conservative. Fresh polling of 139 professionals (fielded August 7–21, 2025) captures that ambivalence. Just over a third of companies (36%) say they actively use AI—only 16% across multiple functions and 21% in one or two areas—while 26% are not using AI at all and another 26% are unsure. A further 12% are still evaluating tools, suggesting the balance could shift over the next several quarters as pilots mature. Published August 22, 2025, the report frames an industry still in early adoption, with sentiment and capabilities varying widely by company size, category and maturity. PetfoodIndustry
The most decisive signal in the data is where leaders expect AI to pay off first: marketing. When asked about future investment priorities, 35% rank marketing and sales—especially customer targeting and campaign optimization—as the domain most likely to benefit from AI. Market insight applications (trend scanning, competitive analysis) follow at 21%, while production efficiency (18%), formulation (13%), and regulatory/compliance (13%) trail. That pecking order implies executives see AI first as a growth lever—finding and persuading the right pet parents—before they see it as a cost lever or R&D accelerator. PetfoodIndustry
Why the tilt toward marketing? Three reasons. First, the data foundation is already there. Digital storefronts, retailer portals, DTC sites, email service providers, and social platforms stream rich first‑ and third‑party signals. AI can exploit these to model high‑value segments (e.g., multi‑pet households, sensitive‑stomach dogs, indoor-only cats) and predict next‑best offers without major plant or ERP overhauls. Second, the marketing stack is modular. Teams can pilot an AI copy assistant, a lookalike‑audience tool, or an MMM/attribution service without re‑architecting the entire enterprise. Third, the feedback loop is fast. A/B tests on creative or audience selection deliver measurable lift—or not—within weeks, turning skepticism into support or prompting quick pivots. (These are interpretations based on the ranked priorities reported in the August survey.) PetfoodIndustry
Just as telling are the headwinds. The top obstacle is talent: nearly a quarter of respondents cite lack of in‑house expertise as their main barrier. Legacy systems (15%) and “no perceived need” (17%) also loom, and nearly one in ten point to data‑quality problems. Cost and unclear ROI—often presumed to be the biggest hurdles—together account for only about 7%, underscoring that capability, not capital, is today’s gating factor. In other words: leaders aren’t waiting for cheaper tools as much as they’re waiting for the right skills and cleaner data. PetfoodIndustry
For operations and R&D, slower prioritization does not mean limited potential. On the factory floor, quality‑assurance anomaly detection, predictive maintenance, and yield optimization are well‑established AI use cases in adjacent food categories. In formulation, generative design can suggest ingredient swaps to hit nutritional targets and cost constraints. But these domains demand validated data layers (sensor fidelity, recipe libraries, lab results) and robust governance—investments that take longer than spinning up a marketing pilot. The August polling simply reflects time‑to‑value realities: marketing returns are faster to see and easier to attribute, so they rise to the top of the near‑term roadmap. PetfoodIndustry
Strategically, the split points to a two‑speed AI transformation:
- Front‑of‑house acceleration (2025–26): Most brands will start with AI that sharpens demand creation—audience modeling, content variants, offer personalization—because KPIs are familiar (CAC, ROAS, repeat rate) and stack integrations are bite‑sized. The September follow‑on poll confirms that among active users, data/insight work dominates initial marketing use, while creative tasks remain secondary. PetfoodIndustry
- Back‑of‑house modernization (mid‑term): As data hygiene improves and teams acquire skills, attention will drift to production, supply‑chain planning, and formulation optimization—areas where AI can compound margin over time.
The implication for leadership is not just “do more marketing AI.” It’s “sequence wisely.” Use marketing pilots to build a repeatable playbook—data standards, model evaluation criteria, human‑in‑the‑loop review, and measurement discipline—then port that muscle to production and R&D. This crawl‑walk‑run approach aligns with the August survey’s portrait of an industry still sorting out answers to basic questions: What problem are we solving? What data do we trust? Who owns model risk? PetfoodIndustry
Finally, don’t mistake “marketing first” for “marketing only.” The same investment in first‑party data capture (consented profiles, purchase and LTV signals) that fuels personalized campaigns can also inform demand forecasting, new‑product sizing, and channel mix. If leaders design for reuse—standard taxonomies, shared feature stores, and governance that spans CRM and operations—the initial marketing wins become a flywheel for enterprise AI. The polling hints that this is exactly where the industry is headed, just on different clocks. PetfoodIndustry+1
