|
|
Welcome to Issue #1 of Put AI to Work — a weekly briefing built for small business owners who want real results from AI, not a theoretical overview of what AI might become.
This week: the adoption gap closing in on your competitors, which automation platforms SMBs are actually switching to, why 95% of AI projects fail and why that's genuinely good news for you, and what the FTC's 2025 AI compliance priorities mean before they mean anything worse.
Let's get into it.
|
|
In This Issue
01 → Lead: 91% of SMBs Using AI Report Revenue Growth
02 → Feature: The Automation Platform You Don't Have to Commit To Yet
03 → Core: 95% of AI Projects Fail — Small Businesses Are in the 5%
04 → Deep Dive: FTC 2025 AI Compliance — Your SMB Checklist
|
|
01 / Lead Story
|
91% of SMBs Using AI Report Revenue Growth — The Adoption Gap Is Closing
|
|
91%
of SMBs using AI report measurable revenue growth · Salesforce, 2024
|
Nine out of ten small businesses that are actively using AI report real revenue growth as a result. That data is from Salesforce's 2024 SMB research — and it doesn't stand alone.
PayPal's 2024 small business survey found that 82% of SMB owners now tie AI adoption directly to their competitive survival — not growth, survival. And the latest data shows that 58% of small businesses are already using AI in some part of their operations — up from 40% at the start of 2025.
That last number matters. It means more than half of the businesses in your market are already running with an AI advantage. The window to get ahead of this is narrowing fast.
The timeline is also shorter than most people expect. SMBs deploying AI for core functions — customer service, scheduling, marketing, administrative tasks — are reporting measurable productivity improvements within 90 days. Not a year from now. Ninety days from the day they start.
The businesses seeing results right now didn't run complex AI projects. They picked one operational problem — answering customer inquiries, writing follow-up emails, scheduling appointments — and deployed a specific tool to solve it. That's the entire playbook at this stage.
Identify one place where you spend time on something repetitive. Find the right tool. Deploy it. Measure it.
The Numbers |
91% of AI-using SMBs report revenue growth — Salesforce 2024 |
82% of SMB owners tie AI adoption to competitive survival — PayPal 2024 |
58% of SMBs are already using AI in operations — Business.com / JP Morgan 2025 |
90 days — average time to measurable productivity improvement for SMBs deploying AI on core functions |
|
|
02 / Feature
|
You Don't Have to Use Zapier. Here's What SMBs Are Switching To — and Why.
|
|
For years, Zapier was the default answer when a small business wanted to connect two apps and automate something. It still works well. But the market has matured and the alternatives are compelling enough that you should know what's out there before you sign up for anything.
Three platforms are getting the most traction with SMBs right now: n8n, Zapier, and Make. They're built for different situations. Here's how to figure out which one fits yours.
|
n8n — Best for: Cost-conscious businesses with some technical comfort
n8n is open-source. You can self-host it for free with unlimited executions, or use their cloud version starting at $20/month for 2,500 executions. Either way, your per-automation costs drop sharply compared to task-priced tools. It has a visual workflow builder that doesn't require code, but it rewards users who are comfortable with settings, conditions, and logic.
Right for you if: You're running lean, you or someone on your team isn't intimidated by a settings menu, and you want to connect a lot of tools without a bill that scales with every task you run.
Not the right fit if: You want something running in an afternoon with zero technical setup.
|
|
Zapier — Best for: Speed, simplicity, and the largest app library
Zapier connects to more apps than any platform — 6,000-plus. It's the fastest way to get a basic automation live: pick a trigger, pick an action, done. No hosting, no configuration beyond the workflow. It's also been adding AI-powered steps, including natural language workflow creation, which makes it increasingly useful if you want to layer AI in without building from scratch.
Right for you if: You're connecting popular tools — Gmail, HubSpot, Slack, Shopify, QuickBooks — and you want the fastest possible setup. Cost-per-task pricing works at your current volume.
Not the right fit if: You're running high task volume and the per-task billing is adding up faster than you expected.
|
|
Make — Best for: Complex workflows at mid-range volume
Make (formerly Integromat) starts at $9/month for 10,000 operations and sits between Zapier and n8n in complexity. Its visual scenario builder is well-suited to multi-step, branching workflows — it's clearer than Zapier's interface when your logic gets complicated, and it's significantly cheaper per operation at higher volumes.
Right for you if: Your automations involve multiple conditions and branches, you're hitting cost ceiling on Zapier, and you want a more visual picture of your entire workflow at once.
Not the right fit if: You just need simple two-step automations — Zapier will get you there faster.
|
Quick Selection Guide |
Need something live today, no friction → Zapier |
Complex multi-step workflows, cost matters → Make |
Technical confidence, want to cut per-task costs → n8n |
|
|
03 / Core Story
|
95% of AI Projects Fail to Deliver. Small Businesses Are in the 5%. Here's the Difference.
|
|
95%
of AI initiatives fail to scale beyond pilot stage or deliver promised business value · MIT Sloan Management Review, 2025
|
That number is real and it's not an outlier. McKinsey's 2025 research tells a similar story. Most AI projects at large organizations don't deliver the revenue impact that was promised when they were approved.
Here's why that's actually good news for your business.
Why Large Organizations Fail
Enterprise AI failures follow a consistent pattern. The scope is enormous — initiatives that touch multiple business units, require integration with legacy systems built decades ago, and involve months of change management before a single real interaction happens. By the time the first test runs, the use case has shifted, the internal momentum has stalled, or the original budget is gone.
McKinsey points out that the most common failure is not technical — the AI often works. The problem is organizational. BCG's research puts it plainly: 10% of AI success comes from algorithms, 20% from technology and data, and 70% from people and processes. The business isn't structured to use it.
Why Small Businesses Win
Small businesses have two structural advantages that enterprise budgets can't buy: speed and freedom from legacy constraints.
When you decide to deploy an AI customer service tool, you don't route it through three departments, a legal review, and a security audit. You set it up, run it for a week, see whether it works, and adjust. The decision cycle that takes an enterprise six months takes you six hours.
There's also no legacy system problem. You're not trying to connect a modern AI tool to a 20-year-old database no one wants to touch. You're connecting it to the tools you're already using — Gmail, Calendly, QuickBooks, Shopify — most of which have native AI integrations available right now.
The Failure Patterns Worth Avoiding
Starting too broad. Trying to use AI across everything at once rather than one defined task. You can't measure results you can't isolate. |
Picking a tool before defining the problem. "We need to use AI" is not a use case. "We spend four hours a week on customer follow-up emails and want to cut that to 30 minutes" is a use case. |
No baseline measurement. If you don't know how long something took before AI, you won't know whether AI helped. Measure first, then deploy. |
Quitting after one bad week. AI tools have a calibration period. Give any new deployment two to four weeks before deciding whether it's working. |
The 5% Profile
PwC's research on AI implementations that actually deliver finds the same profile every time: single well-defined problem, measurement before deployment, time given for calibration, expansion only after the first use case is proven. That is a playbook any small business can run. The 95% failure rate is an enterprise problem. You're already built for the 5%.
|
|
04 / Deep Dive
|
The FTC's 2025 AI Compliance Plan Is Out. Here's Your SMB Checklist.
|
|
The FTC's 2025 AI enforcement priorities have been published. Short version: if you're using AI in ways that affect customers — in pricing, hiring, credit, or customer service — the FTC has signaled it's paying attention to small businesses, not just tech companies.
This is not a reason to stop using AI. It's a reason to spend 30 minutes getting clear on what you're doing with it and what you'd say about it if someone asked. And it's not just the FTC — states like Colorado, California, Texas, and Illinois now have their own active AI regulations, and NYC's Local Law 144 requires bias audits before deploying AI in hiring. Here's what this means at your level.
What the FTC's 2025 Focus Actually Covers
The FTC's AI enforcement priorities in 2025 center on three areas: transparency — are you disclosing when AI is making decisions that affect customers; fairness — is your AI treating different groups of customers the same way; and accuracy — are any claims you make about your AI's capabilities actually true.
For most small businesses, the practical exposure is narrow. If you're using AI to draft emails, you're not in scope. If you're using AI to make individual decisions that affect customers differently — pricing, credit, service access — you are.
The Employment Discrimination Risk — Know This One
If you use any AI tool in your hiring process — resume screening, interview scheduling, candidate scoring — the EEOC has made it clear that employers are responsible for AI bias in hiring even when the AI was built by a third party. The software vendor's responsibility doesn't eliminate yours.
If you use an AI hiring tool that inadvertently screens out candidates based on characteristics that correlate with protected classes, you're on the hook. Ask your vendors whether bias testing has been done. Keep records of your hiring decisions. That's the practical requirement.
What "Audit Trail" Means for Your Business
The FTC's guidance references being able to demonstrate what your AI did and why. For small businesses, this means being able to answer one clear question if it's ever asked: what AI tools are you using, what decisions do they influence, and how would you explain those decisions to an affected customer?
If you can answer that clearly, you're in good shape. If you're genuinely not sure what AI is doing in your business or what it's deciding, that's worth sorting out — not because the FTC is coming for your specific business tomorrow, but because not knowing is also just a business risk.
Your 2025 AI Compliance Checklist |
☐ List every AI tool you use and identify what decisions or outputs each one produces |
☐ Flag any AI tool in your hiring process — ask the vendor what bias testing they've done and document it |
☐ Review customer-facing AI — chatbots, pricing tools, recommendations — confirm customers can tell when they're interacting with AI |
☐ Check any marketing claims about your AI — the FTC is scrutinizing "AI-powered" claims that aren't substantiated |
☐ Resources to stay current (all free): FTC.gov AI updates, EEOC AI hiring guidance, your state's consumer protection AI bulletins |
The businesses that will face compliance issues aren't the ones using AI — they're the ones using AI without knowing what it's doing. The checklist above takes 30 minutes. Do it this week and it's done.
|
|
That's Issue #1.
Next week: more practical AI for your business — tools, tactics, and real results. Same format, no fluff.
If a fellow small business owner would get something from this, forward it. That's how this grows.
— Put AI to Work
|
|
You're receiving this because you subscribed to Put AI to Work.
Unsubscribe · View in browser
|
|
