09 Jul SEO Test
How to Run SEO Tests Responsibly
Every marketer worth their salt knows about testing.
You test landing pages to see which one drives a higher conversion rate. You test offers to see which ones result in more leads. You test headlines to see which brings in more readership. You test your button colours. (Even though they don’t do anything in the long run.
3 simple steps to follow
- Experiment vs. control: You’d never, ever, ever change every single headline on all landing pages for paid campaigns. So don’t do it organically, either. You tweak a single element and run it against the control group to limit your risk.
- Segmentation: Similarly, you’d never throw up a new landing page for each paid campaign. Instead, you’d pick one keyword. One campaign. Or 10% of the traffic. Again, you’d use a much smaller-than-usual segment to control variations.
- Repeatability: Got some decent results? Good. Do it again. One-time blips won’t pay the bills.
Point is, proceed with caution on this stuff. You don’t want to do something you can’t undo. You don’t want to de-index your site if you don’t know exactly how to roll back those changes.
3 SEO Tests You Need to Try
- Remove bold tags
Keyword density used to be a thing. You wanted to place oh-so-many keywords into a piece to hit the 1-2% that guaranteed nirvana.
Keyword stuffing quickly became a thing, too. (In fact, it still works on YouTube.)
The theory is, if a little of something works, a lot of something will kill.
Fast forward a few years, and we’re still going for the same tricks. For example, bolding keywords.
You know, try to work them into the H2 if you can. Then slap at least a few bolds on before they go out the door. This sounds silly. It can’t work… can it? Because it starts to look ridiculous, too.
Turns out, one SEO experiment showed that, unsurprisingly, being overzealous with <strong> tags can backfire.
- Strip dates from URLs
Chances are, you created your blog years ago. Before you knew what you were doing – or maybe your business started it before you were even around.
The bummer is that decisions made back then can (and often do) come back to haunt you today.
Take permalinks. It’s normal to assign a custom permalink structure in WordPress when you’re first getting started.
For example, select one of the following, and you’re stuck with dates in your URLs for good:
- Optimize for dwell time
Originally introduced by Duane Forrester, previously of Bing, dwell time refers to the length of time a visitor spends on a page before heading back to the search engine that sent them there. (We all know it was Google. Sorry Bing.)
Ideally, the longer the dwell time, the better.
It makes perfect sense when you think about it.
SEO isn’t about rankings, keywords, etc., contrary to popular belief. It’s about answering search queries. It’s about being the best at giving people what they’re looking for.
Your goal is to match search intent.
Someone hitting your site and then the back button seconds later would result in a low dwell time. And it’s a bad sign that you haven’t been able to give them what they were looking for.
So it’s kind of like the Bounce Rate or Time on Page you’re already used to. But not really. A little more nuance is involved.
Dwell time is an important concept because it dictates how we should design pages and what should go on them. It’s why long-form posts tend to rank better than short-form ones – not because people like reading (they don’t), but because it helps keep people glued to your screen a little longer.