Potential Keyphrase Ranking with Twitter

By Michael Dorausch, D.C.

I set out to perform an experiment this morning related to longtail keyphrases (in this case focused on local search). Using Twitter, the popular micro blogging service, I sent a tweet to followers, asking them to share information about their location, and local weather.

An investigation like this one works best when you have active followers. Fortunately I had replies within minutes, which makes an experiment like this much easier to blog about.

OK, the question asked was… Weather Near You. Tweet me City you’re in and Current Weather info. (experiment and blog post) Thnx!

And the replies were as follows…

From Virginia Beach Al Doss (abdoss on twitter) messaged it was 34 degrees and partly sunny. Pamela Lund (Pamela_Lund on twitter) messaged from San Diego with a report that it was 51° and skies were clear.

Audrey Seiberling (ShirleyTipsy on twitter) tweeted from Rancho Cucamonga (that’s in Southern California) that skies were clear and it was 56° out (almost same whether as San Diego).

The coldest weather report came from Dan Perry in Chicago (danperry on twitter) with news that it was only 5° out. I guess that’s not too cold for Chicago, but not the kind of weather my Los Angeles blood could handle for long.

I wasn’t expecting such a nice weather report from the Portland, Oregon area but Aaron Hockley (ahockley on twitter) tweeted that it was sunny and around 40° in the Portland, OR area.

Steven Crick (stevencrick on twitter) replied via TwitterFon saying that in Orlando, Florida it was just simply cold. Funny, I just realized I didn’t ask anyone to give me city/state information, but several did (still works for this experiment though). I wonder if Florida cold is like Venice Beach cold (not really cold but cooler than normal).

The tweet from Boston, Massachusetts I received from ljd on twitter did not include weather information, but he sent a second message saying it was clear and cold in the Waltham, MA 02451 area. The city state and zip code information is good enough for this experiment, so I included his tweet.

Another tweet came from the Portland, Oregon area. They must love their twitter in the northern Western states. Emarketing Strategist Elge Premeau (elgepremeau on twitter) said the weather in the Portland area was Sunny and warm for a change. Woo hoo!

So far, we had two localized responses from the Portland Oregon area and now we have two from San Diego. Michael Buonomo (SearchBuzz on twitter), SEO for geeks.com, tweeted this message regarding weather in the coastal San Diego area… Sunny SD Baby — 67 degrees getting up to 74 today. (coastal)

I like that Michael had such a positive outlook on the days weather, not only giving us the current temperature, but what the afternoon coastal highs were expected to be. If this were a contest, it looks like Oregon would be the winner. Another tweet came from that state, this one from Todd Mintz (toddmintz on twitter) saying that the weather in Beaverton, Oregon was Sunny & 33 degrees.

OK, so that’s 10 responses so far to my question about location and the weather. What does this all mean? It may not mean anything to you but here’s where I’m going with it. This next screenshot is a bit larger and shows the Twitter page title tag appearing at the top of an Internet Explorer browser.

In the screenshot, John Andrews (johnandrews on twitter) tweet to me regarding the weather in Seattle (@chiropractic Seattle, WA steady cool breeze, overcast, 39 degrees) shows that the @ reply to my user profile @chiropractic followed by city and state information [Seattle, WA] creates a title tagged page with a high value keyword phrase (local search related to the chiropractic industry).

I noticed I missed somebody. Brian Chappell (brianchappell on twitter) tweeted that it was cold and sunny in Cary, NC (35-40 degrees) but I didn’t grab a screenshot. Next time Brian.

Now, this is SEO theory, and I’m not suggesting anything secret or making any claims, but for keyword phrases that are not too competitive, would someone theoretically be able to tweet @ a profile that included a keyword, and complete the phrase they were seeking to rank for? This may seem like a waste of time and possibly rightfully so for competitive phrases. I’m thinking about the long tail here, and also just sharing my thoughts on the experiment out loud.

Let’s say someone tweets the following information…

That’s a nice locally specific  keyword rich description snippet in less than 140 characters. There is even room for a short URL. Theoretically speaking, any chiropractor could create an account on twitter, ad a shortened URL, and tweet address information to @chiropractic. We know there are tons of bots tweeting local information (and all kinds of other keyword rich data) constantly, but in the cases I’ve shown above I’m thinking about the keyword specifically appearing in the title tag as a factor.

Questions I have include:

• would the domain trust of twitter help to rank a long tail key phrase (versus the same information on a new site)?
• would the retweeting of the initial tweet URL have any effect on rankings?
• for noncompetitive keyphrases, how many links to the tweet URL would it take before a page 1 ranking took place?

The weather examples don’t necessarily apply but I figured it would be an easy way to get localized data to help me with this experiment. What if the data was keyword rich and specific to the keyword profile? My initial thinking is that with little effort pages could rank fairly well, although I don’t have any screenshot queries to show in examples (yet). That will come in another post.

Thanks to everyone that helped out, you guys are great!