itstargetnews
At the 77-year-old promotional products company Gold Bond Inc.CIO Matt Price knows generative AI adoption won’t come from flying a chatbot. Employees need AI embedded in the work they already hate doing: messy ERP usage, document processing, and call follow-up.
Instead of setting benchmarks, Price built a small group of “super-users” to show Gold Bond-specific examples and train the rest of the org. Then they wired Gemini and other models into high-friction workflows, backed up by sandbox testing, guardrails, and human review for anything public-facing.
The payoff shows as behavioral change, not hype: Daily use of AI has increased from 20% to 71%, and 43% of employees report saving up to two hours a day. “I want to bring everyone along for the ride,” Price told VentureBeat. “After we reset some expectations, people started to trust it. Our adoption took off.”
ERP streamlining, product visualizations
Gold Bond, Inc. – can’t go wrong with the skin care company – one of the largest suppliers of $20.5 billion promotional products industrycreates custom swag and corporate gifts for 8,500 active customers.
Orders, quotes, and sample requests come via website, email, fax, and more — in every format imaginable. “That’s why it’s so chaotic,” Price said.
AI has proven to be a natural fit. Previously, employees manually entered order details into the ERP. Now, Google Cloud ingests incoming documents and normalizes them, while Gemini and OpenAI extract and structure fields before pushing a complete purchase order into the system, Price said.
From there, Gold Bond evolved into a pragmatic multi-model approach: Gemini within Workspace, ChatGPT for backend automation, Claude for QA/reasoning checks, and smaller models for edge experiments.
"We’re pretty agnostic on utilizing AI technology,” Price said. Gold Bond is largely set up as a Google shop, with implementation and change management led by Google premier COMPANIONS Promevo.
Early wins include phone call summaries, email drafting, and contract review. A more advanced use case is AI-assisted “virtual mockups” of branded products; used by teams Recraft to repeat visual samples before sending previews to customers, Price said.
Employees also use AI to create formulas in Google Sheets (including Excel-style XLOOKUP logic), while NotebookLM helps create an internal knowledge base for procedures and training.
Other ways Gold Bond uses AI internally:
-
Presentations: The job that took four hours now takes 30 minutes, said Price.
-
Code audit: Developers run scripts in NetSuite, then use two models to check them before moving to testing.
-
Research: Track importer trends and tactics in response to tariffs.
AI also compresses early-stage planning. “We went back and forth with AI and came up with a high-level project that we could build for implementation,” Price explained. “We get to concepts more quickly. We have a lot of smaller meetings, which is good.”
To quantify the impact, Price’s team runs Kaizen events – short workshops that document baseline workflows and compare them to versions aided by AI and automation.
To validate multi-LLM workflows, Gold Bond tests changes in a sandbox environment and runs QA scenarios before launch. “Our technical team, along with subject matter experts, sign off before sending changes or going into production,” Price said.
Change management is required
Adoption is not automatic – in a legacy company, change management is the job. “It’s a little scary, it’s something different,” Price said.
Most users start with Gemini because it’s built into Workspace, then switch to ChatGPT, Claude, or Mistral if they need different capabilities — or a second opinion.
Price relies on a “small cool group” of about eight early adopters to test the bleeding-edge devices; once they have a use case, they train the rest of the team.
“You can’t just look at something like a new piece of software," said Promevo CTO John Pettit. "You need to change people’s thoughts and behavior around it. “
But even though Price’s team advocates widespread use, blind trust is not an option, he stresses.
Gold Bond adds policies, DLP controls, and identity layers to reduce the use of shadow AI. It also uses LibreChat to centralize access to approved tools, enforce paid/approved usage, and block certain models when necessary.
Human-in-the-loop is mandatory: Public-facing content goes through approval, and outputs must be verified. “You have to set the right temperature of trust, but verify,” he said. Even with strong prompts, the outputs still require verification. “You’ve got the data, you can’t just brilliantly take it and use it.”
For example, he asks for sources and reasoning – “Give me all the work cited, where you got this data from” – and treats the verification step as part of the work flow, he said.
Price also warns against overreach. “Agent solutions can only go so far — you still have to have people on the edge,” he said. “Some people have bigger visions than what technology can do.”
His advice for other businesses: Don’t get too carried away with the hype. Start simple. Start with the basics. “Give a detailed prompt, try it, play it.”
