Promotional-products companies are no longer asking whether anyone in the industry uses AI.
ASI's 2026 State of the Industry coverage says 64% of distributors report using AI, while supplier adoption is higher. PPAI's research also describes firms moving through AI onboarding and toward measurable business impact.
But “uses AI” is a broad category.
A company that generates product descriptions with AI and a company that lets an agent reconcile hundreds of transactions are both counted as adopters. Operationally, they are in very different places.
The first wave was naturally individual
Generative AI entered the industry through the easiest door: work one person could improve immediately.
Write an email. Generate an image. Rewrite a product description. Brainstorm an idea. Research a prospect. Summarize a meeting.
These use cases are valuable because they require little infrastructure. An employee can try them today.
They also leave the underlying workflow mostly unchanged.
The person still receives the order, moves the data, checks the supplier, updates the ERP, sends the status, and reconciles the result.
The second wave is organizational
Operational AI is different because the unit of value is not a better output from one prompt. It is a recurring business process that moves without waiting for someone to initiate every step.
commonsku's 2026 roadmap reflects this shift inside its own platform: from AI tools that respond when asked toward agents and automations that notice stalled work and act. PRINTING United is using the language of implementation. PromoStandards' Tech Summit is discussing advanced AI use cases and integration with industry infrastructure.
The market is moving from AI as a tool to AI as part of the operating system.
Cross-system work is the hardest part of the second wave
Native platform agents are powerful when the workflow is contained inside that platform.
The branded-merchandise business rarely is.
A project can span email, commonsku or Syncore, a supplier portal, QuickBooks, a custom webstore, a spreadsheet, and a shipping site. That is where company-level automation becomes difficult.
The next AI architecture has to work across system boundaries, including systems that do not expose useful APIs.
The new maturity question is “what runs without you?”
Companies should stop measuring AI maturity by the number of employees who have ChatGPT accounts.
A more revealing question is: Which recurring workflows now complete without a person performing every step?
That could be an order moving from a store into the ERP. A supplier status being checked and communicated. A bill being compared against the job. An overdue invoice receiving a follow-up. A repeated approval being routed automatically.
That is when AI starts changing operating capacity.
Creation made AI visible. Execution will make it valuable.
The industry's creative use of AI will keep improving. It should.
But the largest economic impact may come from the less glamorous side: the thousands of clicks, lookups, re-entries, checks, and handoffs that keep every order moving.
The second wave will be quieter than the first. It may also matter more.
AltOps focuses on that execution layer: agents taught from real operator workflows that can move between applications, apply learned rules, ask for help on exceptions, and keep running without requiring a new system of record.
Sources: ASI, “State of the Industry 2026: Adopting an AI Strategy,” July 29, 2026; commonsku, “What's Next in commonsku: Every Feature Coming in the Second Half of 2026,” Sept. 4, 2026; PromoStandards Tech Summit 2026 content.
Written by Kavish Soningra, teaching agents to run the back office at AltOps.
