New Findings Shed New Light

By September 2025, 700 million people worldwide were using ChatGPT weekly, representing about 10% of the planet’s adult population. That level of adoption is unprecedented for a technology that launched less than three years ago.  

How is this reflected in US small and medium businesses (SMBs)? A major study analyzing 1.5 million ChatGPT conversations was released this month (Sep 2025) by the National Bureau of Economic Research (NBER) that analyzed how people actually use ChatGPT — and the data has direct relevance for SMBs.

Why? Because the data was drawn from the Free, Plus, and Pro plans, excluding the Enterprise subscriptions. SMBs overwhelmingly use these plan levels first. Few jump straight into the hefty Enterprise contract; instead, owners and employees are mostly using Free, Plus or Pro accounts. With ChatGPT’s broad adoption, this dataset is arguably the best mirror of early SMB adoption behavior.

And the results are striking: while much of the daily headlines focus on AI for coding, enterprise-scale applications, and therapy, the reality is that most work-related usage centers on practical guidance, decision support and communicating.

In other words, the tasks that define everyday SMB life.

What People Are Actually Doing

Forget the sci-fi scenarios. Nearly 80% of all ChatGPT usage falls into three perhaps mundane categories:

1. Practical Guidance (32%) tutoring, how-to advice, creative ideation.

2. Seeking Information (26%) searching for facts, products, or current events.

3. Writing (21%) drafting, editing, summarizing, translating, communicating.

Notice what's missing from the top three? Coding represents just 4.2% of usage (although is much higher in the Claude world). Companionship and therapy? A mere 1.9%. The sensational AI apocalypse looks suspiciously like… getting help with everyday work tasks. The study makes clear that humans are still at the center —making decisions, interpreting results, drafting text. AI amplifies their efforts.

SMB Relevance: Daily productivity struggles are where many SMBs can see broad improvement in efficiencies: sales proposals, marketing content, RFQ responses, HR policies, polished client emails. AI is being used as a savvy business guide, research assistant, digital editor — not just a generator of new material.

Some employees are already AI-capable, and some need guidance to embrace AI in their everyday work. And we all need steadfast upskilling for the foreseeable future. The question isn't whether to adopt AI—it's how to systematically scale from individual use to business-wide deployment before your competitors do.

Work Hard, Play Hard

YThe use of AI in the workplace is seeing rapid adoption across industries, departments, and job levels.

• 40% of US adults use AI at least sometimes at work.

• Work-related usage of ChatGPT grew 200% year-over-year.

• Yet, work-related ChatGPT usage represented 27% (so over 70% personal!).

Unpacking that for SMBs: This reflects the dual nature of ChatGPT adoption in SMBs: owners, leaders, and employees blur personal and professional use. Even if “official” usage looks small, the actual work usage is rising steadily in the background.

In raw numbers, personal use is at 70%. So how is this related to business? Remember, work-related use TRIPLED year-over-year. This means non-work-related usage grew even faster (the proverbial pie grew). This is still a positive. Think about it: when your sales manager uses AI at home to research vacation destinations, they're building AI fluency that transfers to work. When your operations director uses AI to plan a major house remodel they’re evaluating complex projects, timelines, suppliers, costs, and impact. When your marketing manager uses AI to plan their wedding they’re orchestrating a major event down to every important detail and navigating social dynamics with diplomacy. Sound familiar? These capabilities translate to the workplace.

Decision Support Is the Core Value

The majority of work-related messages fell into just two activity groups:

• Obtaining, documenting, and interpreting information

• Making decisions, giving advice, solving problems, and thinking creatively

SMB Relevance: This goes beyond efficiency. It means ChatGPT is acting as a decision-support engine for knowledge workers — helping them think through problems, not just produce outputs. This certainly resonates with me and reflects my use for building my company. For SMBs without layers of analysts or consultants, that’s a direct competitive edge.

Unlike large enterprises, SMBs have more agility; they don’t need bureaucratic committee approval to experiment. The fact that ChatGPT conversations skew toward practical, applied tasks shows SMBs can turn AI into real outcomes faster.

SMBs Have a Hidden Advantage

TLarge enterprises are still forming AI committees and elaborate governance regulations, while SMBs that are using AI, shadow or sanctioned, can realize the benefits more readily. If you aren’t using AI, here are some ways your competition likely is:

• Practical guidance as a force multiplier (32% of usage): SMBs often don’t have dedicated in-house legal, IT, or HR teams. Asking AI for how-to guidance — whether it’s building a project plan or troubleshooting Excel — saves time and reduces dependence on costly outside consultants.

• Decision support for leaders (information seeking, 26% of usage): Test scenarios (“What are the risks if we expand into this region?”), compare vendor proposals, draft internal policies, brainstorm strategies. It becomes a virtual thought partner.

• Writing as a strategic lever (21% of usage): Emails, proposals, reports, marketing copy — this is where SMBs are gaining immediate productivity. Training staff to use ChatGPT effectively for editing, tone adjustment, and translation is low-risk, high-reward.

• Train employees on demand (Educational tasks, 12% of usage): SMBs can get a higher percentage of their workforce upskilled faster than enterprises, giving them an outsized advantage.

SMB Relevance: This creates opportunity. Your five-person professional services firm can now produce content, research, and customer communications that rival Fortune 500 companies. Your 300-person manufacturing business can strategically enter new markets faster and with more agility without six-figure (or 7!) consulting retainers.

Small and medium businesses occupy a unique sweet spot—they have more sophistication than pure consumer use, and more flexibility than large enterprise deployments.

What NOT to Do (the expensive mistakes to avoid)

Here are some risks and blind spots for SMBs to be aware of.

Don't ban AI while pretending to study it. Your competitors' employees may already be months ahead of yours in AI fluency. Every day you delay is a day they're getting better at using these tools.

Don't overthink enterprise-grade solutions first. The study reveals that consumer ChatGPT drives massive business value. Start where your people are already succeeding, add early guardrails, then scale where the impact is greatest.

Don't ignore the personal usage. That "non-work" 70%? It's building the skills your employees bring to work every day. Companies that recognize this will develop more sophisticated AI capabilities faster. Responsible AI will quickly become a priority, but don’t let it halt progress now.

Don't assume this is just for knowledge workers. The study shows AI finding utility across diverse job categories and organization types, from factories to professional services.

Early Strategic Moves for SMB Leaders

Acknowledge the Reality: Your employees are using AI, whether or not you’ve sanctioned it.

1. Audit Your Hidden AI Usage: Survey your team anonymously about current AI tool usage. You'll likely discover more adoption than you expected—and valuable use cases you haven't considered.

2. Develop Lightweight Policies Now: Start with guidelines around sensitive data and customer confidentiality and train staff quickly. Then fold in broader governance guidelines and training.

3. Train for Impact: Show teams how to use AI for editing, summarizing, and decision support — the areas with highest ROI.

4. Start Pilots in Key Workflows: Marketing, sales proposals, HR policies, and customer support scripts are all ripe for experimentation. Workflows and no-code agents are a natural progression.

5. Measure Outcomes: Track whether AI use saves time, improves quality, or drives revenue. Use this data to decide where to invest next.  

Net, Net: SMBs Live in the AI Sweet Spot

Large enterprises move slowly. Consumers play casually. But SMBs? They’re in the sweet spot: resource-constrained enough to need AI, agile enough to adopt it quickly.

This perspective focuses on the immediate economic value in easy reach today: guidance, decision support, and communicating — the exact areas where SMBs can leap ahead immediately. Some SMBs are already expanding into workflow automations and deploying no-code agents, while more sophisticated SMBs are building custom agents and applications with impressive results.

That’s a strategy SMB leaders should embrace: you don’t need a technical team or a big IT budget to unlock value. What you need is clarity: where AI helps most, how to use it responsibly, and how to scale from experimentation to systematic adoption.

The question isn’t whether your business will use AI. It’s whether you’ll harness it deliberately, or let your competitors do it first.

Sandra Peterson, CEO & Founder
Sandstone Partners
September 22, 2025

Sources:

How People use ChatGPT: National Bureau of Economic Research, September 2025.

Work tasks (US adults) — ~40% use AI for work tasks at least sometimes. The Associated Press July 225.

Among U.S. adults who have used an LLM: ChatGPT 72%, Gemini 50%, Microsoft Copilot 39%, Meta (Llama) 20%, xAI Grok 12%, Anthropic Claude 9%. Elon University March 2025.

Not sure where to start? We can customize a 2-hour workshop for your business or organization to get started.