Update 5-23-16: The Official Google Webmaster Blog just announced a new feature rolling out that solves the issue of multiple properties for the same site!
For the last two months, I’ve watched a relaunched client’s Search Console charts spiral downwards. Their indexing dropped, their SERP rankings were meh, and external SEO report crawls showed numerous errors that the Search Console just didn’t echo. Anyway, I had a great big “D’OH” moment this morning upon realizing that Google sees http and https URLs as two separate sites. No amount of proper URL-change redirecting will affect this view, and Google’s Change of Address tool does not apply to a protocol change. Once I began tracking the newly-secure site as its own Google property, BOOM–all the data was there, along with all the mysterious errors other crawls had produced.
I’d assumed that Google, having spearheaded the “use HTTPS or suffer our wrath” initiative last year, would gracefully accept redirects and other change of address techniques when a site changes protocol, but no. Make no assumptions, friends–even Google doesn’t know what it’s doing.
Changing your domain protocol (http to https) makes Google view your domain as an entirely new site and it must be tracked in Google’s webmaster tools as a new site.
Have you ever told your kids (or were told as a kid), “I am not making different food for everyone. I spent a long time cooking, so this is what’s for dinner and we’re all eating it”? If so, congratulations! You understand the basics of programming logic.
No one likes to do the same thing over, and over, and over. A great deal of innovation in manufacturing began because someone like Eli Whitney said “screw this, if the same motion has to be made over and over again, why can’t I build a contraption to do the motion for me?” Computers are much the same way, and constantly being optimized–that is, they can do the same thing over and over but faster this go-around.
Why HTML is not a programming language
The cotton gin didn’t do much if there wasn’t someone loading cotton into it, and manufacturing today doesn’t get far if the materials aren’t loaded correctly. Websites are the same way–you can write all the fancy executable code and styles you want, but until you’ve aligned your materials correctly, no computer knows what to do.
Last month, I explained how hotlinking works and why it’s essentially stealing. I also explained webservers and configuring Nginx webservers to refuse hotlinking requests. It’s a wonderful, powerful little change! But sometimes for the sake of your readers or SEO purposes, you want to allow sites to use images straight from your server—like so:
I hate getting authentication errors in general. At the worst, authentication errors return as a “wrong password!” popup, becoming a catchall for “I don’t know the error, the server just didn’t like me.” At best, you have ruled out typos, permissions, or just plain “being wrong” and begin the onerous task of sifting through millions of tech forum and stackoverflow questions from people who haven’t ruled out user error, hoping to find the one post from a person whose error code and situation matches yours–and contains a solution.
Here’s hoping my solution helps someone, too.
I’m not good at making things pretty. I’ll just lay that out to begin with, in case you thought I’d be able to share trade secrets of beautifying the web. But I’m good at figuring things out and not afraid to make mistakes in order to reach my goals, and you’d be shocked at how far those two things will get you in life.
While checking the spread of my writing and tracking site engagement via Google Analytics, I found disappointing trends in behavior I believe can be fixed.
- Old blog posts don’t get read as often. Even though some of my old articles had a great deal of activity around them at the time they were written, once they’re off the front page of my site they fade into obscurity.
- Even with a good number of pageviews and long session time on articles, there’s a high dropoff rate from the home page. The flow of activity on my site usually starts with a shared link, with the next two pages most often clicked on being the home page and my “About Aimee” page (in that order.)
- Google analytics shows me very few people visit the category pages. When they do, only a handful click through to an article.
- A lot of people check my contact page but I don’t have actionable content there. In other words, it’s a dead-end page. Even if they find what they’re looking for (info about me) the only links lead them off my site.
As I’ve said, I’m not great at making things pretty but I do work in ecommerce, and I have a theory that the same sales techniques we employ can be used to sell my words here. I’m going to be marking goals to track in Google Analytics of lower bounce rate (the amount of people who leave after reading one page), more activity per visit (increasing the number of pages someone reads in a visit), and increasing engagement on old posts. Not all of these will be fixed with a site redesign, but it’s a solid first step to address these issues.