Most small businesses don't decide to automate. They get forced into it because their best person is about to quit.
The signs that you need automation show up in people's behavior, not in dashboards. By the time it shows up in a dashboard — revenue flat, margins compressed, customer complaints rising — you've already lost months of runway. Here are the five behavioral signs I see over and over, in order of how early they show up.
1. Your best person is the bottleneck for everything
There's always a person who "just knows how it works." They run the onboarding. They fix the errors. They know which columns in the spreadsheet matter and which ones are old. They reconcile the weird stuff the system gets wrong.
When that person is on vacation, everything slows down. When they're in a bad mood, everything slows down. When they finally hand their notice in, you lose three months of institutional memory in one afternoon.
Cheapest first move: document what Sarah does. Literally sit next to her for a day and write it down, step by step. Nine times out of ten, half of what she does is rote and automatable. The other half becomes an SOP the next person can follow without inheriting her brain.
I'll build you a documentation template for this for free if you ask. It's the single highest-ROI first move any SMB can make.
2. You have three tools doing the same job
Not four. Not five. Three. Three is the inflection point.
- HubSpot (bought because sales said they needed a CRM)
- Mailchimp (kept from the old website launch)
- Intercom (added last year for support, but the chat is now doing "campaigns" too)
Each of these has a partial view of your customers. Each has partial marketing data. Each has partial notes from sales calls. Nothing matches. Your team has to check all three to answer a simple question like "when did we last email this person?"
Cheapest first move: pick the one tool that you're contractually locked into or that has the data you can't move, and make it the source of truth. Then either (a) cancel the other two and migrate, or (b) wire them to the source of truth so data flows one direction only.
Most people try to "sync" them — bad idea. Bi-directional sync has edge cases that will eat your life. Pick a winner.
3. Onboarding is a fire drill every single time
Every new client, you do some variation of:
- Manually create accounts in 4-6 tools
- Copy-paste their info from the contract into each one
- Email them welcome templates, one at a time
- Set up a recurring meeting
- Create folders in your file system
- Add them to your billing system
- Forget one of the above about half the time
And then a week later you realize you never added them to the email list. Or you gave them the wrong Slack invite. Or their billing didn't go through because you typed one digit wrong.
Our onboarding isn't a process. It's a checklist held together by one person's memory.
Cheapest first move: an onboarding automation. This is the single most common automation I build. When a new client signs, the system:
- Creates their records in every system (usually via API or Zapier)
- Sends the welcome email sequence on a schedule
- Creates the shared folder with your template structure
- Adds them to the billing system with the right plan
- Creates the kickoff meeting
- Notifies the right person on your team
Cost to build: typically $1,500 – $3,000. Time savings: 2-3 hours per new client, at zero error rate. If you onboard one client a week, the build pays for itself in under two months.
4. You're paying for SaaS you don't use
Go pull your credit card statement for the last 90 days. Highlight every software subscription. Now answer for each one:
- Who on the team actually logs in at least weekly?
- What's the last workflow that used it?
- What would happen if you cancelled it tomorrow?
The honest answer for 20-40% of your SaaS line items is some version of "nobody" / "I don't remember" / "probably nothing."
This isn't just waste. It's a sign that your stack was built by accident. Someone signed up for a trial two years ago and auto-renewed. Someone tried a tool for one project. Marketing bought something for one campaign. And none of it got audited.
Cheapest first move: I built a Tech Stack Cost Auditor specifically for this. Takes ten minutes. It flags overlap and estimates what consolidation saves. Most SMBs I've run it with find $500–$3,000/month in redundant subscriptions.
5. You're making business decisions on data you don't trust
This one is subtle. It sounds like:
"I think last quarter was better than this one, but I'd have to pull the report to be sure."
"Our conversion rate is... somewhere around 12%? Maybe 15%? It depends on which platform you look at."
"I don't know what our monthly recurring revenue actually is. The bookkeeper has one number and Stripe has another."
If your reports require a human to "pull" them, and different reports tell different stories, you're running the business on vibes. That's fine for a while. It stops being fine around $500K/yr in revenue, and it's a catastrophe around $2M.
Cheapest first move: pick ONE metric. Just one. Make it dead-simple to see, in real time, without anyone pulling a report. Usually I start clients with pipeline-in-dollars for this quarter. Once they trust that number, we add the next one.
A real dashboard that pulls from your actual source of truth costs $500 – $2,500 to build. Looker Studio + a few APIs is a cheap, ugly, reliable starting point.
The pattern
Notice what all five signs have in common: they're not technology problems. They're organizational problems that technology can fix if — and only if — you admit the underlying truth.
The underlying truth is usually one of these:
- You grew past the size where one person holding it all in their head still works
- You adopted tools faster than you had processes to use them
- You've been "meaning to get around to fixing this" for six months
Automation isn't about being futuristic or having a cool stack. It's about the boring work of making your business repeatable so it doesn't depend on any one person's heroics to function.
Where to start this week
If you read this and recognized your business in more than two of these five, here's what I'd do this week, in order:
- Hour one: open a doc called "Things only one person knows how to do." Dump everything you can think of.
- Hour two: pull your last 90 days of SaaS charges. Highlight the ones you can't explain.
- Hour three: pick the single most painful manual process on your list. Write down the steps.
- Day two: decide if that process is worth automating. Rule of thumb — if it's done more than once a week and takes more than 15 minutes, yes.
- Day three: either build a v1 yourself with Zapier / Make, or get a quote from someone to build it properly.
You don't need to automate everything. You need to automate the top 20% of pain, and you need to do it before your best person realizes how much they're carrying.
If you want a second opinion on which automation pays back fastest for your specific business, that's exactly what my free Discovery Call is for. 15 minutes, no pitch, just a look at where the biggest leverage is.
Next step
If this hit a nerve, let's fix it.
15-minute Discovery Call. No pitch. Just a quick look at what you're dealing with.
Book a Discovery Call