The lab — Day 1 of 365 —
The site had two addresses
Before this day, DayBloq had a website the way a building has scaffolding. Something stood there. You could not live in it.
Google was keeping two houses for one site: daybloq.com and www.daybloq.com, each free to collect its own version of every page. That is not a penalty. It is worse than a penalty, because nothing tells you it is happening. You just quietly split whatever authority you have in half and wonder why neither half is enough. Everything now points at www.daybloq.com — canonical tags, the sitemap, the robots file, the Search Console property. One address. One house.
419 keywords in a meta tag
The old layout carried a keywords array with 419 entries in it. Not 419 words of writing. 419 comma-separated hopes, sitting in the head of every page on the site.
Google stopped using the keywords meta tag in 2009 and has said so out loud since. So the tag was not hurting rankings. It was doing something slightly sadder: it was a written record that nobody had decided what this site was about. 419 keywords is the same statement as zero keywords, typed at greater length. The FAQ and contact pages had their own versions of the same file — 528 and 457 lines of it.
The whole day removed more code than it added: 1,318 lines in, 1,643 out. That ratio is the honest summary of day one. Most of the work was deletion.
Labels a machine can read
What went in instead was boring on purpose — the structured data that tells a crawler what it is looking at rather than hoping it guesses. An Organization node so DayBloq is a thing with a name and a support address. A WebSite node. A SoftwareApplication node saying this is a free Android app on Google Play. A FAQPage on the questions people actually send us.
None of that makes anything rank. It removes the excuse for getting us wrong. There is a real difference between a site that is unconvincing and a site that is unreadable, and only one of those is fixable in an afternoon.
Rooms instead of a hallway
The other half of the day was giving the product somewhere to be. A page for Year in Pixels. A page for the digital bullet journal. A page for the mood tracker. About, privacy, terms — the pages a stranger checks before they trust an app with a year of their days.
We also added /llms.txt, because it costs nothing and takes five minutes. We are not going to pretend it is a strategy. Ahrefs looked at the sites that publish one and found roughly 97% of them get zero requests for the file. It is a business card left on a table in an empty room.
Search Console on the www property, sitemap submitted, pages indexed. That was the whole job: make the house exist, and make it exist once.