How I Automated My Entire Workflow (And You Can Too)

Six months ago, I was spending roughly 25 hours a week on tasks that didn’t require a single creative decision. Email triage, data entry, content scheduling, report generation, research compilation — the kind of work that keeps you busy without actually moving the needle.

Today, those same tasks take me about 4 hours a week. Not because I hired someone. Not because I stopped doing them. Because I automated them — systematically, methodically, and without writing a single line of code.

Here’s exactly how I did it, what tools I used, and how you can replicate the process.

The Starting Point: A Brutal Time Audit

Before automating anything, I spent one week tracking every task I did in 15-minute blocks. The results were humbling.

Email triage: 2 hours/day. Sorting, reading, flagging, responding.

Content scheduling: 1.5 hours/day. Writing posts, scheduling across platforms, finding images.

Data entry and reporting: 1 hour/day. Copying numbers between spreadsheets, generating weekly summaries.

Research and reading: 1 hour/day. Finding relevant articles, summarising key points, bookmarking.

Miscellaneous admin: 0.5 hours/day. Invoicing, filing, updating project trackers.

That’s roughly 6 hours a day of operational work. The problem? I was billing maybe 4 hours a day on actual client work. The rest was overhead.

So I decided to attack each category systematically, one week at a time.

Email Triage: From 2 Hours to 15 Minutes

Email was my biggest time sink, and the one I was most sceptical about automating. Turns out, most of my emails fell into predictable categories.

What I did:

I set up a system using Gmail filters and Claude. Every incoming email gets automatically categorised:

– Client urgent → flagged red, pushed to a priority inbox

– Newsletter/content → auto-labelled, batched for a 30-minute Friday read

– Admin/invoicing → auto-labelled, batched for Monday morning

– Spam/low priority → archived automatically

For the actual replies, I created a Claude project with templates for the 15 most common responses I send. Instead of typing each reply from scratch, I select a template, paste in the relevant context, and Claude drafts a personalised response in seconds.

Tools used: Gmail filters (built-in), Claude Projects, TextExpander for quick insertion

Time saved: ~1.75 hours/day

The honest bit: It took about two weeks of tweaking filters before the categorisation was accurate. And I still manually review anything flagged as “client urgent” — automation doesn’t mean abdication.

Content Scheduling: From 1.5 Hours to 20 Minutes

I post across four platforms: LinkedIn, Twitter/X, Instagram, and a blog. Creating and scheduling content used to consume my mornings.

What I did:

I batch-create all my content on Sunday evenings using a structured workflow:

1. I tell Claude about the week’s themes, events, and topics

2. It generates drafts for all four platforms

3. I review, edit, and personalise each one (this is where my voice matters)

4. Everything gets loaded into Buffer for scheduled posting

For images, I created a set of Canva templates with my brand colours and fonts. Creating a new graphic now takes 2 minutes — swap the text, export, done.

Tools used: Claude, Buffer, Canva templates, Notion for content calendar

Time saved: ~1.25 hours/day (across the week)

The honest bit: Batch creation means my content can feel less reactive to current events. I’ve balanced this by keeping a 30-minute “daily check” where I can add timely posts if something relevant happens.

Data Entry and Reporting: From 1 Hour to 15 Minutes

This was the task I most despised. Copying data from one spreadsheet to another, reformatting numbers, generating weekly reports with the same structure every time.

What I did:

First, I eliminated the worst offenders — any task where I was manually copying data between systems. Zapier now handles the data flow automatically.

For reporting, I built a simple Google Sheets dashboard that pulls data from multiple sources using their APIs. Every Friday at 9am, it generates a summary that I review in 15 minutes instead of spending an hour building it.

The key insight? Most reports follow the same structure week after week. If the structure is predictable, it can be automated.

Tools used: Zapier, Google Sheets + Apps Script, Claude for analysis interpretation

Time saved: ~45 minutes/day

The honest bit: Building the automated system took about 8 hours of initial setup. That’s roughly two weeks of “saved” time before it paid for itself. Worth it, but not instant.

Research and Reading: From 1 Hour to 20 Minutes

I need to stay current on AI trends, automation tools, and industry developments. Previously, this meant scrolling through dozens of articles, newsletters, and social feeds.

What I did:

I consolidated my reading into three curated sources instead of dozens:

1. RSS feeds filtered through Feedly, focused on 12 high-quality sources

2. A Perplexity daily briefing that summarises the top AI news in 5 minutes

3. Claude Projects where I upload interesting articles and ask for summaries and actionable insights

Instead of reading 20 articles, I read 5 curated summaries and dive deeper into the 2-3 that are most relevant.

Tools used: Feedly, Perplexity, Claude Projects

Time saved: ~40 minutes/day

The honest bit: I sometimes miss niche articles this way. But the trade-off is worth it — I’ve actually become better informed by reading fewer, deeper sources rather than skimming many shallow ones.

The Weekly Review: Where It All Comes Together

Every Friday afternoon, I spend 30 minutes reviewing my automated systems:

– Are the email filters still accurate?

– Did the content queue generate appropriate posts?

– Are the reports pulling correct data?

– Am I missing important research?

This is the non-negotiable part. Without regular review, automated systems drift, errors accumulate, and you end up spending more time fixing problems than you saved.

Building Your Own Automation Stack

You don’t need to automate everything at once. Here’s the approach I recommend:

Week 1: Email

Set up filters and templates. This delivers the biggest immediate impact and costs nothing.

Week 2: Content Scheduling

Choose one platform and batch-create a week’s worth of content. Notice how it frees up your daily routine.

Week 3: Data and Reporting

Identify your most repetitive data task and find a Zapier or Make workflow that handles it.

Week 4: Research and Reading

Cut your sources down to the 5-10 that genuinely matter. Set up AI summarisation for the rest.

After four weeks, you’ll have reclaimed roughly 2-3 hours a day. That’s 10-15 hours a week — enough to take on an extra client, learn a new skill, or simply reclaim your evenings.

What I Didn’t Automate (And Why)

Some tasks resist automation, and that’s perfectly fine:

Creative work. Blog posts, strategy sessions, and original thinking still need my brain. AI helps with drafts and brainstorming, but the ideas are mine.

Client conversations. Building relationships requires genuine human interaction. I automate the admin around meetings, not the meetings themselves.

Learning and skill development. I deliberately spend time each week learning new tools and techniques. This is an investment, not overhead.

The goal isn’t to automate everything. It’s to automate the repetitive, predictable, soul-crushing work — so you can spend your time on what actually matters.

The Numbers

After six months of systematic automation:

– Weekly time saved: 15-20 hours

– Client capacity increased: from 4 to 6 clients

– Revenue increased: roughly 40%

– Stress decreased: immeasurably (but genuinely significant)

The automation didn’t make me lazy. It made me more focused. I spend more time on creative work, client relationships, and strategic thinking — and less time on the operational plumbing that used to consume my days.

Where to Go From Here

If you want the exact templates, workflows, and tool configurations I use — including the email filter rules, content scheduling templates, Zapier workflows, and research system — I put everything together in a guide called The Complete Automation Stack. It’s the system I’d give to a friend who wanted to replicate what I’ve built. You can grab it at yellowchilli.gumroad.com/l/xvarrw.

Or, if you prefer to start from scratch: pick the single most time-consuming task in your day and automate just that one thing this week. Start small, prove the concept, then expand. That’s exactly how I built my stack — one automation at a time.