How We Automate at Mindsora (And What We Learned the Hard Way)

There's an old saying in construction: the plumber's house always has leaky taps.

At Mindsora, we wanted to avoid that. If you sell automation, your own house needs to be in order. Not just for credibility — but because you learn from it. Every automation we build for ourselves makes us better at building for clients.

This is an honest account of what we automated, what it delivered, and — perhaps most useful — what failed.

Our setup

Mindsora is a small company. No office with twenty people. So our automations don't need to be enterprise-grade. They do need to be reliable and actually save us time.

Here's our current toolset:

Process	Tool	Status
CRM and pipeline	GoHighLevel	Fully automated
Lead intake	Website form to GHL	Fully automated
Follow-up sequences	GHL email + WhatsApp	Fully automated
Appointment scheduling	GHL Calendar	Fully automated
Invoicing	Moneybird	Semi-automated
Social media scheduling	Buffer + manual	Partial
Website	Custom React (Vite)	Manual deployment
Content creation	Manual + AI-assisted	Manual

You can see: it's not all-or-nothing. Some things are fully automated. Others are deliberately manual. There's a story behind each choice.

What delivered the most value

1. Lead follow-up (saving: 6 hours per week)

This was our first automation and by far the most impactful.

Before: someone filled in our contact form. I saw it whenever I happened to check. Sometimes after an hour, sometimes half a day later. Then I'd send an email. Sometimes I forgot the follow-up.

Now: the form submission lands in GoHighLevel. Within 2 minutes, an automatic confirmation goes out — via email and WhatsApp. Then a sequence of 5 touchpoints over 14 days kicks in. Each tailored to where the lead is in their decision process.

The results spoke for themselves:

• Response time: from an average of 3 hours to 2 minutes
• Follow-up rate: from 65% to 100%
• Conversion to call: from 22% to 41%

That 41% conversion isn't just a number. Those are real conversations with real business owners who would otherwise have dropped off.

2. Appointment scheduling (saving: 3 hours per week)

The back-and-forth about availability was one of my biggest frustrations. "Can you do Tuesday?" "No, Wednesday works better." "Morning or afternoon?" Four messages for one appointment.

Now, the automated follow-up includes a booking link. The lead picks their own time slot. Confirmations and reminders go out automatically. No-shows dropped from 18% to 4%.

Sounds simple? It is. And that's exactly the point: the most effective automations are often the simplest ones.

3. Pipeline management (saving: 2 hours per week)

In GoHighLevel, we have a pipeline with five stages:

1. New contact
2. Qualification
3. Call scheduled
4. Proposal sent
5. Client

Leads move through the pipeline automatically based on their actions. Form submission triggers stage 1. Booking link clicked triggers stage 3. Proposal accepted triggers stage 5.

I no longer need to manually track where everyone is. The system does it. I open my dashboard in the morning and immediately see who needs attention.

What didn't work

This is where it gets interesting. Because we also built things we had to tear down.

AI chatbot on the website (rolled back)

We built an AI chatbot for our own website. Trained on our services, pricing, FAQs. Technically, it worked perfectly.

But it didn't feel right. Our target audience — SMB owners considering automation — wants personal contact. They want to know who they're talking to. A chatbot on a website that promises personal service feels contradictory.

We removed it after three weeks. Instead, there's now a simple contact form with the promise: "Response within 2 minutes." That's what clients actually want. Not a chatbot — speed and a human.

Lesson learned: Not everything you can automate should be automated. Sometimes the human element is your unique selling point.

Fully automated content (rolled back)

We experimented with having blog content fully generated by AI. The output was factually correct, but something was missing. No personality. No opinion. No "here's how we actually do it."

Now our content process is hybrid. I write the structure and the core message. AI helps with research, suggestions, and first drafts. But the final text comes from me. You're reading the result right now.

Lesson learned: AI is an excellent co-author but a mediocre solo author. It lacks the nuance that builds trust.

Automatic proposal generation (simplified)

We built a system that automatically generated a proposal after every introductory call. Based on conversation notes and predefined templates.

The problem: every client is different. The automatic proposals were technically correct but lacked context. A business owner hesitating about the investment needs a different proposal than someone ready to start tomorrow.

Now we generate a draft proposal that I then customise manually. That takes 15 minutes instead of the 45 a fully manual proposal required. Not fully automated, but a 65% time saving.

Lesson learned: Semi-automation is often the sweet spot. Let the machine do the heavy lifting, but keep the human touch for the finishing details.

What we advise our clients based on our own experience

After two years of automating our own business, these are the insights we share:

1. Start with the pain point, not the technology. Don't pick a tool and then look for a problem to solve with it. Pick your biggest frustration and find the solution for that.

2. Measure before and after. You think you know how much time something takes. You're almost certainly wrong. Track it for a week. Measure again after automation. The difference is always bigger than you expect.

3. Don't be afraid to roll back. Not every automation is an improvement. If it doesn't work, or if it degrades the client experience, undo it. No shame in that. We've done it three times.

4. Never fully automate the first client contact. First impressions matter. An automatic confirmation is fine. But the first real contact — the call, the proposal, the kick-off — that has to be human.

5. Maintenance is not optional. An automation you build and forget will break. Tools update, APIs change, your offering evolves. Schedule 30 minutes monthly to check your automations.

The numbers after two years

Metric	Before automation	After automation	Change
Admin hours per week	18	7	-61%
Lead response time	3 hours	2 minutes	-99%
Lead to call conversion	22%	41%	+86%
No-show rate	18%	4%	-78%
Clients per month	3-4	6-7	+75%

Are these numbers representative for every business? No. Every business is different. But the direction — significantly less admin, higher conversion, better client experience — is consistent across every business we work with.

What's still on our roadmap

We're not done. Two things are planned:

Automated onboarding. When a client signs on, we want to automate the onboarding process: welcome email, intake questionnaire, first planning session, access to the project dashboard. Right now, that's still largely manual.

Smarter content distribution. Automatically converting blog posts into social media snippets, newsletter items, and LinkedIn posts. The content already exists — the distribution can be smarter.

To be continued.

---

Curious what automation could deliver for your business? Schedule a free consultation. We'll look at your processes together and show you where the quickest wins are — based on what we've learned from our own business and dozens of other entrepreneurs.