76% of SMB owners are automating work. The next step is making sure they're doing it safely.

I don't do new hobbies halfway. Baking, sewing, whatever it is—I'll disappear into inspiration boards, overbuy supplies, and find myself fully invested in a niche I joined only four hours ago. Unfortunately, that's when I hit a wall. Turns out enthusiasm and a can-do attitude only get you so far. 

Many businesses are approaching AI exactly like I approach a new hobby: all enthusiasm, shaky planning. And it's happening at scale. 

The majority of SMBs (76%) have adopted some form of automation, ranging from simple rule-based workflows (38%) to AI that makes decisions across multiple systems with limited human input (30%).

Unlike my hobbies (which I can shove into a drawer if they go wrong), AI running across your business isn't something you can abandon if it goes sideways. That's why a governance foundation is no longer optional, and it's why this guide exists.

We surveyed over 500 SMB owners and operators about how they're using AI and automation, what's holding them back, and what's gone wrong. This is what they told us, plus what to actually do about it.

Table of contents:

• 55% of SMB owners have work they've been meaning to automate. Here's how to pick what to tackle first.

• Most agree that AI offers a competitive edge. Governance turns that edge into a long-term advantage.

• Nearly all SMB automation users have seen a positive payoff. Here are 3 ways to protect that advantage.

55% of SMB owners have work they've been meaning to automate. Here's how to pick what to tackle first.

With 76% of SMB owners using automation, it's no question it's a priority—but small and medium-sized businesses are implementing automation differently, both with and without AI.

• 38% use rule-based automation that connects apps and runs predictable, "if this happens, then do that" workflows.

• 52% have AI features built into the tools they already use to assist with specific steps in a process.

• 30% use AI-powered automation that decides what to do across multiple systems or apps and takes action with limited human input.

Among the owners who've adopted automation, most aren't just experimenting with the technology. More than three-quarters (74%) rely on automated processes daily or use automation to run many of their core processes, including multi-step or AI-driven workflows. Only 26% use just a few basic automations like scheduled emails or invoice reminders.

This strategy is paying off, too. Among automation users, 94% report benefits such as improved consistency and fewer errors (44%), cost savings (44%), and time freed up for higher-value work (43%).

And yet, 55% of businesses say they have repetitive manual tasks or processes they've thought about but haven't automated yet. These are the top three reasons why: 

• 43% aren't sure which tools would actually work for their business.

• 34% say the cost isn't justified for their use case.

• 30% say automating those tasks feels too complicated or technical for their team. 

A few holdover manual tasks won't make or break a business. But the cost can add up. In the last year, 35% of SMB owners say someone worked past their regular scheduled hours to keep a critical process running. Another 27% say someone made an error in a manual process that cost them time or money, and 26% say they turned down or delayed work because their team was stretched thin. 

If you're stuck in the same rut as these businesses, you don't need to automate everything at once to see meaningful results. 

• Consider versatile tools with a low barrier to entry. If you're not sure what tools your team needs, start with no-code platforms that connect the apps your team already uses. Zapier MCP, for example, can be set up in minutes within any AI assistant.

• Start with your most repetitive task. Look for work that's done the same way every day or every week. Things like follow-up emails, moving information between apps, creating invoices, or updating spreadsheets. And instead of trying to automate an entire team, pick a single process and measure the results. Track how much time it saves, whether errors decrease, and how employees respond. 

• Involve the people actually doing the work. Employees who handle manual tasks every day usually know exactly where the biggest bottlenecks are. Asking for their input can help you prioritize automations that solve real problems rather than adding unnecessary complexity.

• Calculate the cost of doing nothing. It's easy to weigh a monthly subscription fee against your budget. It's harder to compare that fee to line items that never actually appear on an invoice, like losing money due to a manual error. Documenting these specific losses can help you determine if automation is the more profitable path, not just the more efficient one. 

Most SMB owners agree that AI offers a competitive edge. Governance turns that edge into a long-term advantage.

According to most SMB owners, AI has officially moved from being an experiment to an advantage that will set their business apart. Over a third (35%) told us that AI will be essential to staying competitive over the next two years. Another 35% say it will provide a meaningful advantage even if businesses can succeed without it. 

Combined, that's 70% of SMB owners who expect AI to strengthen their competitive position in the next two years. If you've used the internet in the last year, this probably doesn't surprise you. But with this explosion of enthusiasm comes the need for a clear governance framework that helps employees use AI while protecting sensitive information and producing high-quality work.

Don't assume your team has the right tools and training to use AI safely. Make sure they do. Hope is a beautiful thing in literature and a catastrophic substitute for onboarding.

Create a governance policy your team will actually use

AI governance doesn't need to start as a 50-page document or a months-long project. The most useful policies are the ones employees can understand and reference when they're making decisions in the moment.

Start with a simple one-pager that answers the questions your team is already asking: Which tools can we use? What information can we share? When does AI need a human review? Who should we go to when we're unsure?

Here's a quick checklist of things to include in your first policy draft: 

• Approved AI tools. List the tools employees can (and cannot) use and create a clear process for requesting new ones.

• Data guidelines. Define what information should never be entered into an AI tool, including sensitive customer information, confidential business documents, credentials, or proprietary data.

• Ownership and support. Give employees a clear person or channel to contact when they have questions. A named owner is more useful than a vague instruction to "ask IT."

• Error reporting. Create a simple way for employees to flag incorrect outputs, security concerns, or unexpected behavior. The goal is to find and address issues early.

• Set a review schedule. AI tools and best practices change quickly. Set a date to revisit your policy and update it as your team's needs evolve.

• Establish human review requirements. Identify which types of AI-generated work need a second set of (human) eyes before they're shared externally or added to important workflows.

Nearly all SMB automation users have seen a positive payoff. Here are 3 ways to protect that advantage.

For many businesses, automation is already delivering results: 94% of automation users reported at least one benefit. The next challenge is protecting that investment as more employees, tools, and workflows are added to the mix.

That doesn't require a massive governance overhaul or a team of AI specialists. Start by creating practical guardrails that help your team use AI consistently, catch issues early, and keep workflows running smoothly.

1. Make your workflows accessible so everyone can use them correctly

A safety measure that only your most technical employee understands is less of a guardrail than a decorative railing in a dream sequence.

If employees don't understand how a tool works, they may use it in ways that create new risks or avoid it entirely. Nearly one-third (30%) of SMB owners told us they haven't automated certain tasks because their teams find the processes too technical.

Prioritize tools that make automation accessible, including no-code tools like Zapier that let employees build and manage workflows in plain language rather than complex technical instructions. When more people understand the systems they're using, your business can get more value from automation without losing control over how it's applied.

2. Build in checkpoints and keep a record

AI-powered automation works best with clear instructions and a human in the loop to catch mistakes before customers see them. People work better that way, too.

The 29% of owners worried about errors happening while they're not watching are describing a solvable problem. 

Add checkpoints at the moments that matter: notify someone before a client-facing decision goes out, and log it when a process touches sensitive data. Both of these steps give you something to trace back when you need to answer "what happened here?"

3. Keep your systems and policies flexible and up to date 

Over half of SMB owners (52%) use the AI features built into their existing tools. Those tools are constantly shipping new AI capabilities, and the stack that fits your team today may not fit in six months.

This same principle applies to your AI governance policy. Write it so it can be amended rather than replaced, revisit it when your tools get major updates, and keep asking your team for feedback. If you're not already running occasional AI training sessions, start—and use them to collect the use cases people have figured out on their own.

With Zapier, you can build automated workflows that flag sensitive data before it leaves your systems, add human approval steps before decisions go out to clients, and log activity across every tool in your stack—all without writing a single line of code. If your stack changes (and it will), your automations flex with it: swap one app for another without rebuilding from scratch. That's 9,000+ apps, built-in governance controls, and a workflow layer that keeps running no matter which AI your team is using next month.

Related articles: 

• 34% of people shipping software using AI tools have no formal programming background

• AI security: How to protect your tools and processes

• 5 systems you should automate today so they run themselves tomorrow

• Human in the Loop: Pause Zaps for human review and approval