umershaukat67

Joined 1 month ago

Karma
2
Submissions
5
Member since
Sep 21
Founder of CMO Owls (www.cmoowls.com)
View on Hacker News

Recent submissions

Comment · 8 days ago
Try after 90 days atleast. That will be an accurate result.
Comment · 8 days ago
Haha, I would make verified crawlers free but the unverified crawlers I would make them pay. So, blocked by default, but if it verifies itself then free. And one time verification cost.
Comment · 8 days ago
Yes it is. I would actually say, building has been extremely easy, but the problem always comes with the distribution. With AI, you can build anything, but as a solo foudner and working with only AI will get you struggling with where to put my posts and how ot find the threads. This is somehting that needs to be strictly checked. And for marketing its not one person's job at all!!
Comment · 8 days ago
It is better to differentiate between the crawls, like which ones are training and which ones are search bots. You can disallow the training bots explicitly in the robots.txt and still show up in ChatGPT for example, when someone asks a question.

In my case, my logs were flooded with training crawlers, not the search bots.

Comment · 9 days ago
nice work man in bringing a raw page to 99 pagespeed. As you are letting bloggers pace their own ads, this will kill your scores. Today my own website cmoowls.com hit an LCP of 5.8s with a score of 70. But the LCP element was just a line of text, not an image.

The proble, was the element render delay. The first frame actually arrived after the load even, so our analytics already had started loading. Lighthouse counted this in the LCP and it was bad. Then. I fixed this by loading analytics after first paint instead of on loading. And then the next three runs gave me 99, 99, and a 100 (LCP 1.8, 1.8, 1.7).

But one small point here: CWV usually comes from Chrome's 28 days of visitors. 99 is just a lab score.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton