There are three reasons small businesses haven't automated certain tasks, at least according to the owners who'll admit it: 43% aren't sure which tool would work, 34% say the cost isn't justified, and 30% say it feels too technical, according to a survey Zapier commissioned from Centiment. Only one of those three is actually a problem you need to fix.
Why Small Businesses Haven't Automated Yet: The Short Answer
Here's the verdict on each, before we get into the weeds. Tool confusion (blocker 1) is usually solvable with a narrower decision, not a bigger research project. Cost not justified (blocker 2) is often just correct, and that's fine. Too technical (blocker 3) is a coin flip: it depends on the specific task, not your team's ability.
One number needs unpacking right away, because it's easy to misread. The same survey found that 55% of businesses have repetitive tasks they've thought about automating but haven't gotten to. That is not “55% of small businesses haven't automated anything.” It sits next to a different figure from the same survey: 76% of SMB owners already automate some part of their work. Both are true at once. Most businesses automate something, and most also have a short list of tasks they've considered and left alone. You're probably somewhere in that second group right now, which is exactly why you're reading this.
This isn't a push to automate everything on that list. It's a diagnosis: which reasons for leaving those tasks alone actually hold up, and which ones are worth revisiting.
What the Survey Actually Found (and What “55%” Doesn't Mean)
Here's the full methodology, because it matters more than the headline number: Centiment fielded the survey for Zapier May 19 to 20, 2026, collecting 508 completed responses from US-based owners and operators at companies under 500 employees. Zapier published the results September 22, 2026. The data is unweighted, margin of error roughly ±4% at 95% confidence.
of SMB owners already automate some part of their work
have repetitive tasks they've thought about automating but haven't gotten to
The 76% figure breaks into three overlapping categories, not exclusive buckets: 38% use rule-based automation (the classic “if this happens, do that” setup), 52% use AI features already built into tools they own, and 30% use AI-powered automation deciding across multiple systems. A business can fall into more than one category, so don't try to make these three add up to 100%. They won't.
The 55% figure measures something different: specific repetitive tasks an owner has thought about automating but hasn't touched yet. It sits alongside the 76% figure, not against it. A business can already automate its invoicing and still have a backlog of tasks it's meant to get to since last year. Both are normal at once, and neither one is a verdict on how well you're running things.
Worth naming plainly: Zapier sells automation. Of course it does. The numbers are real and checkable on its blog, but its own conclusion naturally leans toward “automate more, just do it safely.” This article's framing doesn't share that incentive.
The 3 Reasons Owners Actually Give for Not Automating
The three blockers, exact figures: 43% aren't sure which tools would work, 34% say the cost isn't justified, 30% say it feels too complicated or technical for their team.
The next three sections take each one in turn and give it an honest verdict, not a pitch. The scannable version first:
| Blocker | What it sounds like | Verdict |
|---|---|---|
| Not sure which tool (43%) | “There are too many options and I don't have time to test them all.” | Usually solvable: narrow the decision to one task, not a whole platform. |
| Cost isn't justified (34%) | “It would cost more to set up than it would ever save.” | Often correct, and that's a fine answer for low-frequency tasks. |
| Too technical (30%) | “My team can't handle this.” | Depends on the task, not the team: sometimes the tool is wrong, sometimes the task is a bad first automation. |
We flag which of these blockers are worth revisiting in the newsletter, not just when a new tool launches.
Blocker 1: “I'm Not Sure Which Tool Would Work” (Verdict: Usually Solvable)
Tool overwhelm is real. Too many tools to evaluate, vendor marketing that hides what a tool actually does, a fatigue that sets in before you've even opened a spreadsheet to compare them. But “which tool” is a decision problem, not a research project, once you stop trying to solve it at the category level.
The de-scoping move: pick one task, not a platform. Ask what this one task needs, not which tool is best overall. Which two apps does it connect? Does the step need a judgment call, or the same fixed path every time? A task that just moves a form submission into a spreadsheet needs almost nothing. A task deciding whether a lead is worth a follow-up needs a tool that can actually make that call.
Answer those two questions and the tool list narrows to two or three candidates instead of forty. A pick-by-situation guide once you know the task goes deeper from here. So does a no-hype framework for choosing a tool.
Blocker 2: “The Cost Isn't Justified for My Use Case” (Verdict: Often Correct, and That's Fine)
This is the sharpest verdict in the piece: the blocker most likely to be a legitimate, good decision, not a gap to close. A task done a handful of times a month, taking a few minutes each time, often costs more in setup time and a new subscription than it will ever save. Not exactly what you want to hear after reading yet another “why aren't you automating” headline, but it's true.
Free tiers exist, so “no budget” usually isn't the real constraint. Zapier's free plan runs 100 tasks a month. Make's free plan gives 1,000 credits a month. The real constraint is that those caps don't survive more than one lightweight workflow, and a paid tier earns its keep only once a task runs often enough to clear that bar.
A plain gut-check: low frequency, plus low time-per-instance, plus low stakes if it's late or off, usually adds up to leave it manual, at least for now. Nobody selling automation software will tell you that, but it's true often enough to say plainly.
Cost isn't the only side of this, though. Weigh it against what staying manual is actually costing you, covered below, not instead of this gut-check. The real cost breakdown, tier by tier, is in the dedicated cost guide.
Blocker 3: “It Feels Too Complicated or Technical for My Team” (Verdict: It Depends on the Task, Not the Team)
Two honest sub-cases, and they call for different fixes.
The first is a genuine tool mismatch. Hand a developer-built platform to a non-technical team and it will feel technical regardless of the team's ability. Some tools live up to this reputation: based on our analysis of Trustpilot reviews of n8n (47 reviews, collected June 2026), complexity was the main complaint in 19% of them (second only to support), with one reviewer writing it's “very hard to debug problems since it's all UI. Credential connections expire quickly, making requests start failing.” That's a signal about a specific tool, not small teams in general. Swapping to a no-code tool often resolves it.
The second is a real signal worth listening to. Some tasks need judgment calls and edge-case handling that make them a poor first automation, no matter how simple the tool is. “This feels too technical” is sometimes the team correctly sensing the task isn't ready to be handed off. That's not a skills gap. That's good instinct.
Running the 5-minute test on that specific task settles which case you're in. Reading up on whether it's the tool or the task covers the distinction in more depth.
What Staying Manual Is Actually Costing You
The same survey asked about the other side: 35% say someone worked past regular hours because of manual work, 27% say a manual error cost time or money, 26% say the business turned down or delayed work because the team was stretched thin.
That's a real cost, not a reason to automate everything on reflex. Weigh it against the blocker-2 gut-check above, not instead of it. A low-frequency, rarely-wrong task doesn't suddenly become urgent just because this section exists.
For contrast, here's the payoff case: among businesses that automate, 94% report a benefit, most commonly fewer errors and more consistency (44%), cost savings (44%), and time freed up (43%). The pattern holds both ways. Automation earns its cost on tasks with real frequency or real stakes, and doesn't on tasks without either.
Which Tasks to Leave Manual on Purpose
Leave a task manual if:
- It happens a handful of times a month or less.
- Being late or slightly off doesn't cost much.
- It needs a real judgment call or a personal touch.
- The process itself is still changing shape and hasn't settled into a repeatable pattern yet.
This isn't an anti-automation piece. It's about spending setup effort where the survey's 94%-benefit number actually shows up, not spread evenly across every task.
For the flip side, see which specific tasks actually pay off. For a broader check once you've sorted your blockers, the fuller readiness checklist covers that ground.
FAQ
Why don't small businesses use automation?
Most actually do: 76% already automate some part of their work, per a survey Zapier commissioned from Centiment. Specific tasks stay manual for three reasons: 43% aren't sure which tool would work, 34% say cost isn't justified, 30% say it feels too technical.
What are the barriers to automation for small businesses?
Per the same survey: tool uncertainty (43%), cost that doesn't seem justified (34%), tooling too technical for the team (30%). The first is usually solvable, the second is often correct, the third depends on the task.
Is automation worth it for a small business?
Depends on the task. There's no blanket yes here, whatever the vendor pitch says. If it's done rarely and being late or slightly off doesn't matter much, automating it usually costs more than it saves.
What tasks should a small business not automate?
Leave it manual if it happens only a handful of times a month, being off doesn't cost much, it needs real judgment or a personal touch, or the process hasn't settled into a repeatable pattern yet.
How do I know which automation tool fits my business?
Start from one task, not a platform: which two apps does it connect, and does the step between need judgment or follow a fixed path. A pick-by-situation guide and a no-hype tool-choosing framework both go deeper from there.
Is AI automation too technical for a small team?
Sometimes. It can be a tool mismatch, a developer-built platform handed to a non-technical team, which a no-code tool usually fixes. Or it's a real signal the task needs judgment calls unsuited to a first automation.
How much does small business automation actually cost?
Free tiers exist (Zapier covers 100 tasks a month, Make covers 1,000 credits), but they only handle one lightweight workflow before a paid tier makes sense. The real breakdown, tier by tier, is in the dedicated cost guide.
What percentage of small businesses use AI or automation?
76%, per a survey Zapier commissioned from Centiment: 38% rule-based automation, 52% AI features built into existing tools, 30% AI-powered automation deciding across multiple systems, with overlap between them.
The Short Version (and Where to Go Next)
- Three named blockers, per a survey Zapier commissioned from Centiment: 43% unsure which tool, 34% cost not justified, 30% too technical.
- Verdicts: tool confusion is usually solvable, cost not justified is often correct, too technical depends on the task, not the team.
- 55% have specific tasks considered but not automated; that sits alongside, not against, the 76% who already automate something.
- Staying manual has a real cost too (overtime, errors, turned-down work). Weigh it against the cost-not-justified gut-check, not instead of it.
- Leave a task manual on purpose when it's infrequent, low-stakes if late, needs judgment, or hasn't settled into a repeatable pattern.
We cover this kind of honest verdict in the newsletter too, not just when a new tool launches.
By blocker: not sure which tool, a pick-by-situation guide once you know the task. Cost math unclear, the real cost breakdown, tier by tier. Not sure a task is a fit, run the 5-minute test on that task. Want a broader check, the fuller readiness checklist. Want payoff examples, which specific tasks actually pay off.