Google demoted my site in one week. The byline turned out to be the only distribution channel I actually own.
In the second week of July, my company website fell out of its Google rankings. Not a slow slide — a step. One of my main guide pages went from an average position of 10.6 to 20.5. Another went from 20.7 to 24. Page two of Google is where clicks go to die, and my pages moved there in about seven days.
I run a small automation studio in Israel. We build WhatsApp bots and business automations for clinics, law firms, retailers. The website was my storefront, and search was the street it sat on. I can measure exactly what the demotion cost me, because every WhatsApp conversation that starts from the site arrives through a tagged link: in late June and early July I was getting 14 to 17 site-attributed conversations a week. From mid-July onward: 2 to 4.
Seven days. That is how long it took to switch off the acquisition channel my business leaned on most.
The tempting move was to tear the site apart — rewrite every page, chase whatever the algorithm seemed to want that month. I mostly resisted, because the pages that survived were the ones that had been stable the longest, and because Google's own guidance keeps saying the same boring thing: don't churn content that works.
Instead I made a different decision. If one algorithm update could switch off my storefront in a week, then the storefront was never mine. It was rented. So I started publishing where the asset is the byline, not the domain: under my own name, on platforms that have their own readers.
Since August 1st I have published 16 articles under my own name — five on The Times of Israel's blog platform, seven on dev.to, four here on Medium. Roughly one every two days, next to a full client load. What follows is not a growth-hack thread. It's the three rules that made this sustainable, and the honest numbers of what it has and hasn't done.
Writing under a company logo, there's a template: quote a Gartner stat, add three tips, close with a call to action. Nobody believes it, including the person who wrote it.
Writing under my own name, I gave myself a harder rule: the core number in every article has to come out of a system I run, measured on the day I write. So one article audited 50,669 workflow runs from my own automation server and found that 85% of the failures alerted nobody. Another counted 89,479 real WhatsApp messages from one month to benchmark what the official API would have charged. A third came from grepping my server logs for 34 days, where I found that ChatGPT's crawler was fetching my pages more often than Googlebot — which, in a month where Google had just demoted me, felt like the punchline writing itself.
These articles take longer. They are also the only ones that get quoted back to me.
Twice in August I skipped a scheduled publication day. Once, the next two topics in my queue were case studies of client types I don't currently serve — the outline was sitting there, and inventing a composite "clinic client" would have taken twenty minutes. I skipped both topics instead. Another day, the queued topic was an architecture pattern I believe in but had never measured in production. Skipped.
A company blog under deadline pressure publishes the composite case study. It's just "content." Under my own name, a made-up anecdote is not content, it's my reputation with a fabricated line item in it. The byline enforces a standard the logo never did.
In late August I wrote about the response-time gap that opens up for Israeli businesses over Shabbat, when observant owners are offline. My draft headline said 15 hours. The pre-publish fact-check killed it: my median response times were contaminated by conversations the business itself had initiated. Measured correctly — customer-initiated conversations only — the number was 17 hours, and two supporting claims died entirely. The article shipped a day late with a different headline number.
Nobody outside would ever have caught the difference. That's exactly the point. The discipline isn't for the rare reader who checks; it's for all the ones who don't.
Now the part these essays usually skip.
My dev.to posts have collected a few hundred views in total. Not thousands — hundreds. On the attribution side, 89% of the WhatsApp conversations reaching my business in the last ten weeks can't be tied to any channel at all, because most people don't arrive through tagged links. Some of them may have read an article of mine. I cannot prove it, and I won't pretend otherwise.
So why keep going? Because the two failure modes are not symmetric. The company site's traffic was a rented spotlight: bright, and switched off remotely by someone else, in one week, with no appeal. The byline compounds slowly, but no algorithm update can demote my name off an article I wrote on someone else's high-authority platform. Meanwhile the studio itself keeps running — the client work on my automation studio's services page is what funds the patience this strategy requires.
And one early, measurable return: editors answer. A byline with published, verifiable numbers behind it gets treated as a source, not as marketing. That door was simply closed to "the company blog."
If your company's entire content presence lives on your own domain, you don't own a channel — you hold a lease, and the landlord rewrites the terms a few times a year. The hedge is not more SEO. The hedge is a name that readers and editors recognize, publishing things only you can publish: your systems, your failures, your measured numbers.
Here's my question for the founders reading this: open your company blog and look at the last three posts. If your personal name were on them instead of the logo — with your reputation attached — would you still publish all three unchanged? If the answer is no, what does that tell you about what you're publishing now?
About the author: Achiya Cohen builds WhatsApp bots and business automations for Israeli SMBs at achiya-automation.com.