Simplistic case study - applied SEO recommended practices, result?

[Lagonda]

I like to give feedback, indulge me.

I have a site that has been neglected for years.
It's a forum, it was full of spam, no moderation, no active participation, no editing of titles, no mobile ready, no curation, no https, nothing, completely ignored.
Yet, it did get some random valid new posts, every other day.

In a constant search for procrastinating my other work, and as a kind of an experience and challenge, I decided to give it some care:

- made it responsive
- removed ads
- cleaned up and made the url structure more friendly
- updated code to give hard 404 (it was giving soft 404)
- updated code for every page to have a unique url through 301 redirects (no duplicated content)
- added some meta information
- implemented https
- moved it to a SSD dedicated server with Varnish (made it fast)
- removed all non essential files to reduce requests and page size (average 13 requests page size 70KB)
- cleaned up the content (removing duplicates, redundant, merged, spam, etc.)
- edited most recent titles, to something readable and describable (removing ALL CAPS, "help me" and such)
- implemented new more specific categories
- implemented tags for a "same topic" funnel
- implemented schema.org
- removed several pages (search results pages, profile pages, login, etc.) from Google index, keeping only the "juice"
- improved significantly the internal search pages
- did some curation
- updated the design to the millenial trend where there's nothing to click on
- implemented AdSense's matched content
- got an eye on search console everyday

And... the result after 3 whole months?

Nothing.
Some traffic flutuations, but nothing worth noting.

That's it. That's my feedback.
In a sentence: DON'T BOTHER. :D

My conclusion: if the content is good, unless you're doing something really wrong, Google will take care of it.
To me, there is no amount of SEO good practices that will pay your stress/work, just have good content, keep it simple and working.

Will I see a change meanwhile? I'll keep you posted. :)

[goodroi]

Why did you try to resurrect this dead site? You admit it was neglected for years and had a very long list of problems. I'll be honest not every site is worthy of being saved. What is special about this damaged site that made you want to resurrect it?

Sometimes it is wiser and more efficient to start with a clean slate. If a website has completely cratered, it can take more effort to repair it than it does to launch a new site.

I see you mentioned a bunch of improvements but you didn't mention anything about backlinks. Backlinks are still very important. I would suggest you develop a promotion strategy. How are you going to attract new users to discover your improvements? Making changings and expecting Google to make your site popular again is rarely going to be successful. Google generally prefers to rank popular sites instead of making old sites popular through their rankings. Promoting your site outside of Google can really help your Google rankings. For example getting some social media love might be useful for your site. Getting reviewed by popular people in your industry can also help. Good luck.

[aristotle]

Thanks for reporting this interesting experiment. I agree with goodroi about the importance of backlinks, although I would have still expected all the improvements and fixes to have some effect.

updated the design to the millenial trend where there's nothing to click on

This "trend" is new to me. Can you explain it more fully?

[martinibuster]

edited most recent titles, to something readable and describable (removing ALL CAPS, "help me" and such)

1. That's a start. Did you hack the software so that discussion titles become Title Tags and H1 headings?

2. Links are still an important part of ranking. The previous version of the forum may have discouraged linking. So the new version is a step in the right direction for rehabilitating the site and bringing it back to life.

Good luck and keep us posted.

[Lagonda]

It's a neglected site, not a dead one. ;)
It still gets +1.5k visits/sessions everyday, *all* search engine traffic... my objective was to achieve 2k.

I implemented those measures just for the fun of the challenge and to see if it would improve the overall performance - search engine wise and visitor retention/usage.

I didn't want to get more backlinks or exposure exactly because I wanted to know if SEO so called best practices were really that important.

"millenial trend where there's nothing to click on" - look at any modern app, there's one action, two the most, nothing else. You even get apps where you just swipe, nothing else.
Look at "vintage" sites. Lots of actions, lots of things to do, to click on.
Watching my 14 year old daughter using several apps, I noticed this - they don't care about anything, how it's done, how can I use this or that, how can I do this or that.
They are told: this does this, nothing else. And they're ok with it.

I'll be providing further updates, if any necessary.
I, as an internet "oldie", like to have control, they don't, they just shrug their shoulders and get on with it.
So, "millenial trend where there's nothing to click on" = removing everything except what can't really be removed.

The HTML semantics are ok - everything is semantically correct.

[TorontoBoy]

I, too, believe that content is king. Sure, take care of the basic SEO stuff, like ensure Googlebot can index you, but by and large I leave it alone. Your experiment is an interesting one. Please keep us updated.

Smartphone apps need to be simpler than web sites of old because there's just not the accuracy of the mouse vs finger. There's no hover as well.

[Lagonda]

Adding some context, after a private exchange with goodroi.

This site doesn't have an active "link building" strategy.
At some point it did but, as soon as that work surpassed what I was earning from it, I dropped it.

So, FYI, it only has onsite optimization, nothing more.

[Lagonda]

Found out today that I was sending 302 instead of 301.

Why?

This was the code:
header('Location: '.$Url.'', 301);
Where it should have been:
header('Location: '.$Url.'', TRUE, 301);

[TorontoBoy]

Good for you for finding your bug! That probably was quite painful,s so I share your frustration.

[phranque]

Crawling your site with Screaming Frog or Xenu Link Sleuth is always a good idea.

[System]

The following messages were cut out to new thread at: Does Google really treat 302 & 301 redirects differently? [webmasterworld.com]

Let's stay on topic and if we want to explore in depth any side topics to start a new thread :)

[edited by: goodroi at 7:16 pm (utc) on Nov 3, 2017]

[Lagonda]

Ok, six months after, traffic wise, nothing really improved.

Going to try another experiment with a different subject same platform soon.

Same treatment as the one above but, this will get a even more radical approach.
It's a dead site as it's offline for almost a year. It had 10% of the visits of the above.

New design, ultra minimalist, one single "call to action" per page, no fuzz, no buzz, just plain content delivered as soon as possible to the reader, you want it you get it.
Betting on curation - as it is a somewhat question-answer format, single comment threads don't go public, only threads with useful answers.

A few hours work, let's see how it goes...

[TravisDGarrett]

I can't find a the link again, but once I read an interview, of Google guy; and he or she was explaining that it was taking a whole year, to be able to measure the SEO benefits of such or such change.

Also, no change, does not mean it was useless. It's not possible to tell, but may be that if you hadn't made these changes, the traffic would have decreased over this period.

[Robert Charlton]

I can't find a the link again, but once I read an interview, of Google guy; and he or she was explaining that it was taking a whole year, to be able to measure the SEO benefits of such or such change.

TravisDGarrett, perhaps this is the link you're thinking about. It's via John Mueller, not Matt Cutts (aka "GoogleGuy"). I post it occasionally, as it's one of the best discussions we have on the topic....

Search engines need time & other signals to confirm a site is "fantastic"
June 20, 2012
https://www.webmasterworld.com/google/4467831.htm [webmasterworld.com]

[TravisDGarrett]

Hi Robert !

By Google Guy, I meant "someone from Google", not THE GoogleGuy, ah, ah.

This was this one I was talking about :

Google: Give SEOs 4-12 Months To See Positive Ranking Benefits
[seroundtable.com...]

[Robert Charlton]

Travis, the thread I cited is our discussion about the seroundtable article you mention, and I consider them as a unit.

[TravisDGarrett]

Travis, the thread I cited is our discussion about the seroundtable article you mention, and I consider them as a unit.

I am not sure, the topic at WW is from 2012, the article I was mentioning is talking about a video posted by Google on February 2017 : [youtube.com...]

[Robert Charlton]

Travis, you're right. I read the SER headline too quickly and just assumed it was the same article.

That said, now that I've looked more carefully, I think you should look at both threads. Also, Maile Ohye's video is typically straightforward and goes down easily as all of her videos do.

The John Mueller discussion is more nuanced and gets into the kinds of techniques that can hurt or help, and looks at what happens when an SEO tried to use shoddy shortcuts.

[TravisDGarrett]

I think you should look at both threads.

Oh, yes, of course, I didn't mean anything else. My remark was just to complete what could have already been said, with more recent information. The Topic at WW has not been updated since 2012, so, this video from last year, shows that it still requires time.

[robzilla]

I know you're focusing on SEO and rankings, but I'd be curious to know what all those changes meant for user experience metrics like bounce rate, time-on-site, pageviews per visit, etc. Did any of those things improve following the makeover?

My conclusion: if the content is good, unless you're doing something really wrong, Google will take care of it.

That's their ultimate goal, so perhaps they're succeeding. They want to reward good content and user experiences, not good optimization.

[advancedBeginner]

This is an extremely interesting thread. Thank you Lagonda for posting all of this information in your experiment.

[Lagonda]

As I said above, it's a forum but, a niche one, that doesn't match the usual hobby/passion/interest forum profile.

It's a health forum, you search for a symptom/condition, you land, you see other user's testimony, you say it's this or you say it isn't this what I'm looking for and, 99,99% of the times you fly away forever anyway... 0,01% of the time, the user registers, posts his concerns or experience and, flies away to never be seen again.
There are no frequent users because it isn't a hobby/passion - it's a momentary situation and, when it's not, there's no incentive to keep users returning.
There are other limitations and details but, to keep this message short, that's the main picture.
Yes, I know, it's kind of a flawed business model, but, it's a pet site of mine.

So, as you can see, bounce rate, pages/session and such aren't the best metrics in this case.

Concerning the wait time for the results to appear, I was aware but, it's been almost a year since most changes were implemented...
I'll keep you posted, as promised.

[robzilla]

So, as you can see, bounce rate, pages/session and such aren't the best metrics in this case.

Perhaps not, but did you look into it anyway? All those things should feed back into that 0.01% sign-up rate.

[Lagonda]

They are all steady, flatlined, no changes.

This was a neglected site because of this exact diagnostic - no engagement.

It was supposed to be a self support plataform where you could share experiences with people in the same health related afflictive situations.
Yes, it has some "success" (for the country audience size) but, nothing like what I was expecting.

To single out the country small size factor, I even thought about a franchising deal to use the same platform/design on other countries and see if that works but, never bothered...