By: Eric Okimoto, CEO, Widge.ai --
The business of equipment manufacturing is highly complex and very technical. That’s why commercial-off-the-shelf AI service and support solutions don’t make the grade.
For many equipment manufacturers, getting a customer a clear answer can take days of sales and support teams checking the ERP, waiting on engineering, and confirming specs and lead times. One industry benchmark study shows industrial and manufacturing companies average roughly 62 hours for a first response on general inbound inquiries due to legacy routing processes and field-based sales teams.
The slowdown comes from how manufacturers manage information across the product lifecycle. The average manufacturer pulls data from 4 to 12 software applications, each of which may hold a complete or partial answer to a complex sales or support question. Teams then spend time on routine work, such as entering the same data into multiple systems and checking it for accuracy, instead of helping customers.
Not only is this process frustrating and time-consuming for both employees and customers, it can also impede your business. That’s because the ability to provide timely information and support impacts revenue, customer satisfaction, and referrals. Delays create friction that reduces expansion revenue and makes room for competitors to capture the opportunity.
Artificial intelligence is setting new expectations for speed of response from customers, shifting the baseline expectation for service response times from hours or days to absolute immediacy.
It’s also setting new expectations for employees. JLL's Workforce Preferences Barometer reveals 47% of manufacturing frontline employees cite "the right work environment to be productive" as their top expectation.
But most AI isn’t a fit for most equipment manufacturers, because it doesn’t meet their unique requirements:
AI must be technical grade. AI must be fully capable of ingesting and retrieving complex product information, including formats that have typically been difficult for AI systems to read, such as static PDFs, complex tables, and product schematics.
AI must be highly accurate. Poor answers can be even worse than no answers. Generic AI models struggle with accuracy on technical queries, with accuracy levels ranging from 40 to 60%. This is because generalized AI models are optimized for breadth rather than precision. That’s unacceptable for manufacturers and industrial companies.
AI must be easy to deploy quickly. AI value realization is often stalled due to significant issues with data readiness and/or the need for configuration. Despite recently doubling AI investments, only 37% of manufacturers say they are fully prepared to operationalize AI at scale, according to a recent Riverbed study.
AI must maintain manufacturers’ data integrity. Manufacturing data is the key to competitive advantage. Manufacturers need to be vigilant to safeguard their IP, such that it isn’t being traded to train AI models.
In 2027, equipment manufacturers must up their sales and support game. While AI holds great power and promise for sales and support, it must be applied to the same exacting specifications that manufacturers work with in their design and shop floor operations.
Eric Okimoto is CEO of AEM member company Widge.ai. Widge makes industrial information easier to access and act on, combining intelligent retrieval with an agentic framework that surfaces accurate answers, connects information across systems, and automates workflows. Its multi-contextual understanding handles complex manufacturing conversations that generic AI struggles with, while enterprise-grade security, fast deployment, and out-of-the-box ERP integrations provide peace of mind and fast time-to-value.
