Automation First – What does it mean?
- Pius Schmid
- Aug 13
- 6 min read
When we talk about Automation First, we don’t mean one or two shiny tech projects tucked away in a corner of your company. We’re talking about an organizational transformation - one that creates a sustainable competitive advantage through cost savings AND generating more revenue. We don’t own automation-first.com for fun ;)
This shift is built on three core principles:
1. Team enablement
Everyone on the team should be able to automate their core processes. No waiting around for IT. No throwing tickets into the void. Just doing it. That doesn’t mean everyone becomes a developer overnight, but they should have the tools (and confidence) to build stuff that works for them. Whether that’s automating a Monday morning report or connecting the tools they use daily, it should feel doable.
2. Continuous improvement
Processes change, and they should be improved constantly. And yeah, automated too – so it’s scalable, understandable, and doesn’t fall apart if one person leaves. We take cues from agile here: iterate, get feedback, and don’t aim for perfect right away.
Also: don't be afraid of chaos. Early stages of automation (and AI) can feel messy. That’s okay. You learn by trying. Embrace the mess.
3. Rule-based first, AI second
80–90% of the value we create for our customers comes from purely rule-based automation — and where it makes sense, we integrate AI.
Why? Because a solid rule-based foundation creates clarity and structure. It allows you to define, control, and monitor processes precisely — which is exactly what you need when scaling AI effectively. With clearly defined inputs and outputs, you can steer AI much more reliably and avoid unpredictable results.
Use AI where it truly adds value — reading invoices, summarizing text, making smart suggestions. But don’t plug in ChatGPT just to tick the AI box. Keep it smart, safe, and grounded in real needs.
And stay flexible. AI is evolving fast. Avoid lock-in to a single vendor or model. Tools like OpenRouter help keep your setup future-proof.
Now let us show you what this means in detail - and how your company can start becoming Automation First.
The new default is automated
Automation First means: if something can be automated, let’s automate! Some companies even say, "If we can’t automate it, we won’t do it." A bit extreme but we get the vibe.
Of course, not everything should be automated 100%. Sometimes, you need a human eye. And that’s fine. But if you’re still doing a task manually for the third time, it’s probably time to ask: “Can we automate this?”
The upsides?
Things get done faster
Fewer errors
Better data
Lower costs
Teams stop doing the boring stuff
Everyone has time for the real work
No one dreams of copy-pasting data all day.
Will, Skill, and Environment
You need all three to make automation work:
Will
It starts at the top. Leadership needs to believe in this – not just talk about it. Make space to try new things. Be okay with experiments. Celebrate progress, not just the big wins.
Also, founders and execs need to walk the talk. Use AI tools yourself. Share wins. Show what’s possible. Don’t just delegate this to some innovation task force.
Signs your company has the will:
You talk about automation in all-hands, not just tech meetings
Teams are encouraged to bring automation ideas
Nobody gets roasted for breaking something on their first try
Skill
You still need people who can build the thing. Either in-house or with external help (like NOA ;)).
You need:
People who can and want to work with tools like Make, Zapier, n8n, Retool…
People who can map and rethink processes
Some guardrails: error handling, docs, backups
In the beginning, skill is what you buy. But make sure your team learns too. Otherwise, you’re just renting automation.
Environment
Even the best people can’t automate if the setup sucks.
Use tools that are easy to pick up (low-code)
Make sure people can access the data they need (safely)
Create space to try stuff, even if it fails
We’ve seen teams where IT locks everything down so tight, nothing gets built. That kills momentum. You need openness with smart boundaries.
And yes, people will break things. It’s part of learning. Don’t panic. Build a culture where it’s okay to say, “Oops. I’ll fix it.”
Building the Technical Foundation
Start with your data. Ask:
Where is it now?
Where should it live?
Is it clean? Is it accessible?
Shit in, shit out. That’s true for data science AND automation.
You’re probably starting from Excel files buried in inboxes. Step by step:
Get out of Excel chaos
Move to something better (Airtable, NocoDB, etc.)
If needed, migrate to Postgres or another scalable setup
It all depends on your business and team. There’s no one-size-fits-all here.
Tools & Tactics
Here’s a quick rundown of tools we like:
Zapier – super simple but might get you in the spirit
Make – more flexible and powerful
Retool – great for building apps and dashboards
LangGraph – if your team is quite technical
Custom Code – when nothing else works
Which one you choose depends on:
How complex the task is
Your team’s skills
Budget
Compliance needs
Low-code is non-negotiable if you want everyone involved. But that doesn’t mean avoiding code. Use both and mix it up. Whatever works.
Rethinking the Culture
Let’s get one thing straight: automation isn’t about cutting jobs. It’s about using the people you have in smarter ways.
We’ve had clients say:
“We want to do more without hiring five more people.”
That’s the vibe. You’ve got talented people – let them do the valuable work.
Imagine your:
Support team spends less time digging for info
Finance team closes books in hours not weeks
Sales team never touches the CRM manually again
Automation clears the path for the things you never had time or energy for before. That being said, people that don’t embrace the use of rule-based automation and AI will soon be left behind. So, learn how to use it!
We have seen that even if the company is young and tech-driven, integrating automation as a core component becomes harder the larger the organization. Starting early pays off later. Creating an Automation First culture needs investment. Money and effort. There is no shortcut. With money and effort you can automate first processes. The higher the ROI the better. People will come to love automation. Whenever we finished a project successfully, not once have we experienced that someone said “I want to do it the old way”. It is rather “Great, what’s next”. And not because people are lazy. But because they want to do more of the fulfilling work and less of the boring tasks.
Start small but think BIG
Don’t boil the ocean. Start with one tiny, annoying thing. Automate it, show it, let the buzz spread.
Then:
Find champions – people who love this stuff
Track what’s been automated (so you can calculate a ROI later)
Collect new ideas – everywhere
Share the wins visibly – in town halls, Slack updates, whatever works
Bonus: The AI Squad
Want to go faster? Build a small crew – 3 to 5 curious and hands-on people. They meet weekly, test tools, share results. If something works, scale it.
Soon enough, you’ve got a snowball effect.
Setting up long-term infrastructure
As automation grows, support it with solid infrastructure. You need both tech and org structure.
Tech Infrastructure
Central data storage
APIs everywhere
Error logs and alerts
Version control
Org Infrastructure
A central team guiding the way
Local champions in each department
The central squad sets strategy and helps teams succeed. Everyone else builds what they need.
It works because:
You keep quality high
You share learnings
You build a culture of ownership
Eventually, automation just happens. Quietly running in the background while everyone gets on with valuable work.
TL;DR
Automation First isn’t a nice-to-have. It’s a mindset shift - an organizational transformation that provides a sustainable competitive advantage. It means:
Thinking automation before manual
Giving people tools and trust
Building stuff that lasts
You need will, skill, and the right setup. Start small, scale smart.
And if you want a boost – well, that’s where we come in.
What is the biggest challenge for you?
Enabling the team to build automations themselves
Finding the right use cases
Building automations - technical challenges
