Your SEO Guide

Why Twitter is Critical to your SEO, PPC and Email Campaigns

I’m sorry, but business owners are blind. They see Twitter as a way to reach an audience of consumers, but they completely miss the real value of Twitter for business – A/B testing. That’s right, I’m talking about using Twitter as a true business tool to improve your open rates and click through rates for every digital marketing channel.

Let’s dive in.
You’re doing some keyword research for your website’s SEO, and you need to choose between “Keyphrase A” and “Keyphrase B” for the homepage. Keyphrase A is searched for more than Keyphrase B, but Keyphrase B would still bring a lot of valuable traffic to your site. Most SEOs will either stop here and choose Keyphase A, or they may do a little extra diligence and check out which keyprase has less competition. All competition being equal, Keyphase A will is crowned king. What’s missing? What’s missing is the questions, “What kepyrase will users be more likely to click on?”.

The Misconception
AOL leaked a study back in 2006 that showed the click through rates for their organic listings. Everyone still references that study when they say the #1 organic spot will have a 43% CTR. Even if that CTR didn’t change since 2006 (which it has), it would still be WAYYY off when applied to a specific industry and even more so query. Where am I going with this? My point is that if the #1 result has a terrible title and description it is not going to get clicked on nearly as much as if it had a compelling title and description. Obvious, right? Yeah, except SEO’s don’t seem to care. When’s the last time you saw “Estimated CTR” in a metric in a keyword research report? I’ve been in the industry since 2002, and I’ve actually NEVER seen a research report containing CTR data. I think a lot of that is because CTRs are hard to measure. Apart from Google Webmaster Tools data, which is almost useless for this type of thing, you’re out of luck. Or are you?

Meet Twitter – You CTR testing tool
Twitter is a gift to SEO’s.  Your Twitter audience is willing and ready audience to provide real time CTR feedback without even knowing it. They can help you  test multiple versions of title tags and descriptions to see which will perform the best even before Google’s bots get a sniff of your new title.

How do I do it?
Take Keyphrase A and tweet it out on Monday at 10am and 2pm. Make sure you’ve got a shortener like bit.ly to track performance. Then wait a week, and next Monday test Keyphrase B at 10am and 2pm. If you’d like some additional data points you can schedule your tweets to send at 11pm and 2am for folks in different time zones. HootSuite or Bufferapp.com can help you with that.

While you’re waiting for the results of your Twitter test, why not use Adwords to test optional title tag and meta description combinations for each keyphrase. Setup two ads, each with slight different titles and descriptions, then run both ads on Keyphase A and Keyphrase B. Now your testing which ad performs best and which keyphrase sends the most traffic.

The Results
Use Twitter as your first data point for keyphrase. Take your best performing phrases and create ad copy for PPC testing. Take your best performing ad copy combinations and use it for email campaign open rate testing – you can even test meta descriptions by doing a/b tests on the main email body copy, looking at the email CTR’s.  After testing all these channels you should have a good feel for which title and descriptions will perform the best for each keyword.

This my friends is a scientific approach to search engine optimization. It effectively squeezes more traffic out of #1 rankings, and it’s what makes good SEO’s great SEO’s.

  • http://diyblogger.net/about Dino Dogan

    Dont think I didnt notice how you accounted for variables as much as practical by using the same day and same time frame :-)

    So here is my question…do SEO who do SEO for a living have the luxury of taking the time to run these splits over the course of a week or more?

    • http://twitter.com/dancristo Dan Cristo

      That’s a great question.
      When I did SEO in-house there was plenty of opportunity for me to run tests like this. Partially because it was truly an on-going optimization and re-optimization of the handful of sites I worked on, and partially because I also managed the PPC, email and what would have been the social had Twitter and Facebook been around in 2005.

      At an agency, there is less time to run these sort of tests, and it’s hard to coordinate with the various other agencies who manage their email and social accounts. That being said, there is no reason why you can’t find a week or so to run a test for the homepage and top category pages. In fact, agencies can even hire Twitter Power users to run a series of tests for them. Not only will that drive direct traffic to their website, but they’ll get all the SEO learnings from it as well.

    • http://www.freelinereport.com/ Brad Fallon

      HI Dino,
      Well i am doing seo :)

  • http://twitter.com/LeoWid Leon Widrich

    Dan, I have to say, I am quite clueless about SEO. Your mentioning of Twitter to be used as SEO makes a hell of a lot of sense to me though. Even to me.
    I am actually trying to use this in a lighter version by trying different post titles. So I schedule the same post with Buffer at one point and then at another point again with a different title, to see whether the title has an impact. Do you think that is a viable thing to do? Or would I blur my results somehow if I then deduce title A is better than B?
    Oh and of course, many thanks for mentioning Buffer here, it is much appreciated. Have you been using yet? Would love your views on it.

    • http://twitter.com/dancristo Dan Cristo

      Hey Leon,
      I’ve signed up for Buffer and have explored the site’s functionality, but I haven’t gotten a chance to really use it yet. I did however install the chrome extension. I’ll be happy to give some feedback once I start using it a bit. btw. I love the site look and feel. Very pleasing on the eye and easy to use.

      As for your question on a/b testing… If you run your a/b split at different times of the day you introduce a very big variable, especially on Twitter. If you don’t have time to wait a whole week to do a same time & same day test, I would run the a/b on different days, but the same time.

      For example: Run tweet A on Monday and Wednesday at 2pm, and tweet B on Tuesday and Thursday at 2pm.

  • http://twitter.com/nuttynupur nuttynupur

    Like the way you’ve married traditional search with modern Twitter’s advantages of real time feedback! Thanks. Am keen to try this out now:)

    • http://twitter.com/dancristo Dan Cristo

      Thanks so much. Yeah, I think we’re just tapping the tip of the iceberg when it comes to real time feedback. I’ve been posting a bit more frequently on my blog, so I appreciate the encouragement.

  • http://www.techwork.dk Thomas

    Hi
    That is really clever what you are doing here. I have never thought about using Twitter like that. Thanks for the tip.

    • http://twitter.com/dancristo Dan Cristo

      Thanks Thomas. “Necessity is the mother of invention”. There’s a real need for real time feedback in marketing, and Twitter fits the bill. I appreciate you taking the time to leave some comment love. Hope you’ll frequent the blog.

  • http://twitter.com/ShawnWalsh Unify Interactive

    Nice post Dan,  Thanks for the tip. I like the way you think!

  • Gregory Brian Moore

    Do you know a way to find highest CRT tweets in the history of your Twitter account?