web
statistics

What I learned (painfully) about SEO during 2014

After being hammered by Google’s algorithm changes in 2013, I was determined to be “squeaky clean” with search engine optimisation during 2014.

Nice idea. But I still had an awful lot to learn. And more pain to go through.

My mantra was “quality content” upon more quality content. Of course, the fundamentals never change – title tags, heading tags and image tags. So they would still be the foundation of my SEO strategy.

And then it would be content all the way. Well, a year down the track, here’s what I’ve learned about content.

Not all content is created equal

One of my key content strategies last year was blogging. On client’s sites, I wrote blog posts on a regular basis. A nice steady stream of content. And I waited for the websites to climb up the Google rankings.

Well, they didn’t. In one case, after a blitz of new blog posts, the client’s site dropped three pages overnight. In another case, I added just one new page of content and a couple of days later the site had dropped more than 20 spots.

So, what the heck was going on? Doesn’t Google want content? I’m providing content. So why doesn’t Google love me?

I thought, if one new page of content can cause a site to drop, what will happen if I remove that page? I removed it – and sure enough, the site came back – higher than before.

So, what was wrong with my content? It wasn’t cheap, rehashed garbage. It was well written and added value to the website. But Google still penalised it.

Finally, the penny dropped

When I looked carefully at my content, I was still thinking in terms of keywords. Old habits die hard. Back around 2004, I read a book called “The Nitty Gritty of Writing for the Search Engines” which was all about using keywords in a “natural” way in web content. I practised this method and it became ingrained in my thinking.

This keyword-focused approach used to work. But not any more. If you think keywords first, you are likely to over-optimise your content and end up getting penalised as a result.

Remember, you are creating content for humans, not search engines. Google is so smart now, it looks beyond keywords. It understands the words that are used naturally in conversation around a common theme. This is known as semantics. So, you don’t need to stuff with keywords any more. Just write naturally about the topic and you are more likely to end up in Google’s good books.

So, now it’s starting to make sense

This explains why we now see sites ranked number one on Google that are not clearly optimised for specific search queries, or don’t have traditionally strong SEO signals, such as link profiles.

So the number one lesson I’ve learned about content – be careful not to over optimise for keywords. You can use your keywords in the title and maybe in a heading tag. And that’s enough!

On that note, it seems Google does still rely heavily on the title tag. I’ve seen pages rank highly based on the title tag alone, with no other attempt to use the keywords in the content. But a word of warning on title tags – don’t make them too long or try to over-stuff them with keywords. Just one phrase is enough.

And finally, let’s look at links. While spammy links can hurt your rankings, as I discovered painfully in 2013, we should not discount links altogether. I saw one site rise to the number one ranking in Google last year just by the addition of one link from another site with a high page rank.

The more competitive the market, the more quality links you need. But if the links are quality, you will often need surprisingly few to make a big difference.

SEO in 2015 and beyond

The fundamentals of SEO are really quite simple now. Use the title tag to tell Google what the page is about. Use your “keywords” in the title (but only one keyword phrase). And then write content that is relevant to that theme but doesn’t over-emphasise the keywords.

SEO gurus might claim this is too simplistic. And if you’re an SEO guru, I’m sure it is. But for the ordinary web marketer who wants to see maximum results with minimum input, this is the way to go.

No comments yet.

Leave a Reply