The Moment I Stopped Clicking Buttons

The session that changed how I think about client service happened on a Monday in March 2026. It wasn’t planned as an experiment. I had clients to get through, and somewhere between the second and third I realised I hadn’t opened a browser.

Not once — not for analytics, not for search console, not for keyword tracking, and not for email.

Every data source I needed was callable from the terminal. DataForSEO for keyword rankings. Google Search Console for traffic and queries. GA4 for analytics and conversions. Google Ads for campaign performance. Lighthouse for site audits. Shopify for store operations. Gmail for client communications. Brevo for newsletters. And Stripe for payments. I had every one of them wired into the same session, the same context, the same flow of thinking.

But the real shift wasn’t calling them one at a time. It was chaining them together. A Python pipeline that syncs keyword rankings, pulls search console data, cross-references the movement against last month’s baseline, layers in the analytics and the ad spend, and drafts the visibility report — in one pass. That’s the five-layer stack doing what it was built for. The thinking that used to happen between dashboards is encoded in the pipeline. I build it once, and it runs for every client.

And when you layer multiple data sources on top of each other, something happens that no single dashboard can show you. I’m working with a cleaning company running Google Ads. The Ads dashboard says four conversions — the campaign looks weak. But I’ve wired GA4 into the same session, and GA4 shows seventeen conversions from the same traffic. The real conversion rate isn’t two percent. It’s over ten.

The gap exists because Safari strips tracking cookies during a form redirect — a technical detail buried in the plumbing between the ad platform and the website. Checking the Ads dashboard alone, you’d never see it. Google itself recommends switching the bidding strategy based on the incomplete data. I dismiss it, because I can see both layers at once and the campaign is actually working.

The Context Switch Tax

I know what the old way looks like because I lived it for twenty-six years.

Open SEMrush. Wait for it to load. Navigate to the project. Export the CSV. Open Excel. Copy the numbers into the report template. Take a screenshot of the graph. Log into WordPress. Navigate to the page. Make the edit. Preview it. Publish it. Log into Mailchimp. Start the email. Paste the screenshot. Write the copy. Schedule it.

I’d spend fifteen minutes clicking before any thinking started. And that was one client.

The problem isn’t any single dashboard. The problem is what happens between them. Every login is a context switch. Every context switch is where momentum dies. The thinking that was building — the pattern I was seeing in the data, the connection between this keyword cluster and that content gap — evaporates the moment I switch tabs and wait for a loading spinner.

From the terminal, that friction doesn’t exist. I run the command, the data arrives, and the thinking continues. Same session, same context, and the same chain of thought.

What Actually Shipped

This is what a full service pass looked like that morning:

An industrial services company — I refreshed their keyword rankings across thirty-one keywords in parallel, wrote the visibility intelligence report, and composed an email to the account contact with the full analysis. The data showed twelve keyword improvements and three new page-one positions. I could see the trajectory because the previous report was in the same git history — one command showed me what had changed since the last service.

A solo manufacturer — I synced ninety days of search console data, spotted an opportunity in the query patterns, rewrote the service page to target the cluster, and pushed it live. The rankings data showed exactly where the growth was coming from and why.

A label printer — I refreshed the keyword rankings, tracked position movements against the previous baseline, and updated the visibility report. The deliverables are waiting in the client’s approval queue, but the compounding is visible in the data — page six to page two average across sixteen months.

I gave every client the same depth — which is harder to say about the old workflow, where the third client invariably got the version of me that was already tired of clicking. Each one documented in git with atomic commits, each one with an email drafted in the client’s specific context — not a template, not a status update, but a response to what the data actually showed.

The Map

In 1854, John Snow was trying to find the source of a cholera outbreak in Soho. He had two data sources: a register of deaths, and a street map. Neither told the story on its own. The death records were names and dates. The map was streets and buildings.

He put one on top of the other. The deaths clustered around a single water pump on Broad Street. The source of the epidemic — invisible in either data source alone — appeared the moment the layers combined. Snow didn’t have a better medical instrument. He had two ordinary sources of information, and he removed the wall between them.

That’s what happened with the cleaning company’s ad campaign. Google Ads said it was failing. GA4 said it was working. The truth only appeared in the layer between them. I didn’t have a better analytics platform. The insight came from the combination, not the tools.

I think that’s what this piece is actually about. Not that I stopped clicking buttons. Not that the terminal is faster. That when the walls between your data sources come down, you see things that no single tool was designed to show you. Snow saw the pump. I saw the campaign was working. The pattern is 172 years old.

Six SaaS subscriptions give you six slices of the same business. The SEMrush slice. The Google Analytics slice. And the Mailchimp slice. The whole picture isn’t in any of them — it’s in the space between. You can’t subscribe to it. You have to build it.

The Session Compounds

There’s a second effect that’s harder to see.

A dashboard resets to zero every time you open a new tab. The history is gone. The commands you ran last week, the patterns you spotted, and the queries you wrote — all of it evaporates when the browser closes.

The terminal compounds. The commands are in the history. The patterns are in muscle memory. The output from last week’s session feeds directly into this week’s analysis. I can run the same ranking check I ran a month ago and compare the results because both outputs exist in the same system.

The git log is the compounding layer. Every client service session produces commits. Every commit is searchable, diffable, and dated. When someone asks “what happened to our rankings in February?” I don’t open a dashboard and try to reconstruct it from graphs. I run a git log filtered to that client and that month, and the answer is there — every change, every report, and every deliverable, in order.

That’s the fond applied to client service. Faithful daily work leaves residue that becomes the foundation of the next session. The residue isn’t in a database behind a login. It’s in plain text files in a git repository, and the system that serves the clients can read its own history.

What This Actually Means

I think this matters for anyone running a service business — not just people who write code.

The question isn’t “should I learn the terminal?” The question is: how many context switches stand between your data and your decisions? Every SaaS dashboard is a context switch. Every CSV export is a context switch. Every login screen is a door between you and the work.

John Snow didn’t need a better microscope. He needed his data in the same room. That was 1854. The tools have changed. The physics hasn’t.

The whole picture lives in the space between your tools. And the only way to see it is to remove the walls.

Related: The CLI Is the Interface · The Five-Layer Stack · The Split Brain · Your Website Should Come With a Receipt · The Correction Loop · Content As Code · Ingeniculture · The Atomic Commit

If you want this kind of thinking applied to your business — here’s how I work with clients, or get in touch.