Measuring Up: Testing the Science of Our Digital Marketing Approach

Some time back, I laid out a social media experiment that I would be running over the next several months. Without rehashing the entire article, I outlined a plan that called for broad scale, targeted messaging on Twitter. The argument I made was that mass scale deployment of tweets would enable EdgeTheory to own a steadily growing percentage of relevant keywords — first on Twitter and then eventually on Google as well. Owning keywords, I noted, is key to steering traffic from a search engine to a specific landing page. If you want to claim the traffic, you’ve got to own the signposts.

The time has come to perform our first assessment. If you remember, we’re attempting to drive readers to a company blog that focuses on the conversational aspects of digital marketing. (In case you missed the original post, you can read it here.) The keywords we targeted all revolve around conversation. Some examples include ‘dialogue,’ ‘chatter,’ and ‘communication.’

To implement the study, we started out with a handful of conversation-themed blog writings, which we then began to promote across Twitter using 30 different handles. We originally scheduled five tweets to go out per day on each handle, but later upped that number to 20. We also turned on Kudzu, our in-house metrical tool that scours Twitter for volume of keyword usage and reports back on the percentage we (or our clients) happen to own. We call this percentage our ‘conversation share,’ and for this experiment we gave Kudzu our list of keywords associated with conversation.


So how are we doing? First, there’s already been some major success, a few things we hadn’t even anticipated. But there are also some areas where we need to adjust. Bear in mind that digital marketing is a calculated science but never a precise one. The minute something succeeds it’s time to readjust your strategy, because what works today might fail tomorrow.

Now to the measurement. Taking a look at our ten most successful keywords, the first thing that jumps out is that all of them are preceded by a hashtag. We practically owned “#assumptions” over the past 30 days, but across all of Twitter it only appeared 88 times. In the scheme of things, then, #assumptions isn’t used very much. The same goes for #discourse, for which we hold nearly 80%. It’s nice to dominate in a few chosen keywords, but if few other people are using them you might not be getting much traction.

I was pleased to see how we did with regards to ‘#discussion’ and ‘#dialogue,’ both of which enjoy a little broader usage, even if not to the extent I originally thought they would. Our highest ranking term without a hashtag was ‘chatter,’ of which we owned only one percent. As for ‘conversation’ itself, we aren’t yet in the running. Even though we deployed 1,257 separate tweets containing a reference to ‘conversation,’ its total usage across all of Twitter was in excess of 43,500,000.

What was enlightening, though, was the frequency with which the word ‘talk’ was tweeted. In fact, it’s only used about ten percent as often as ‘conversation’ — a little more than 4,000,000 times. With more targeted messaging, EdgeTheory can definitely gain ground here.

Why is this important? Because Twitter is an open platform that Google and other search engines are repeatedly scouring and indexing. The fact that ‘talk’ was used so many times on Twitter means that it likely gets frequently used in search engines as well. Using a downloadable plug-in called Keywords Everywhere, I see that it gets nearly half a million searches a month but isn’t considered very competitive. The term ‘conversation,’ however, only gets used about half that amount on Google and is considered even less competitive. Good news for us! What this means is that EdgeTheory will probably succeed more easily with ‘conversation’ on Google than we will on Twitter. But this is actually what we want since we’re using Twitter as our doorway to Google anyway.

We still have a ways to go, however. A search on ‘conversation’ in Google does not even place us in the first 20 pages of results. The way to tackle this problem is to ensure that a greater amount of tweets containing the keyword ‘conversation’ goes out. We also need to increase its frequency within the text of the blog posts as well.

One really surprising result, however, is that a Google search for “conversation share” (using quotation marks) landed the blog on the second page of returned results. Granted, it’s not page one yet, but we only need to move up a handful of slots in the rankings to get there.

Something else of interest is happening as well. It just so happened that the EdgeTheory landing page came in just beneath the blog post in the search results. This makes sense when you think about it. As people discover the blog, they’re likely to click around and explore the rest of the site as well. As our blog continues rising in Google’s ranking, will the EdgeTheory landing page travel upwards with it? Let’s hope!


In examining the keyword list associated with this experiment — the one we put into our Kudzu engine — I noticed that some words and expressions were never deployed at all. This means it’s time to revisit the list and either remove certain phrases or making a consolidated effort to include them in the outgoing tweets. Your keyword list should be dynamic and reflect a lot of activity — not read like a pauper’s bank account.

There are also some key phrases I need to add to the list in order to measure their performance on Twitter. For instance, ‘conversation share’ was left off the original list and is, therefore, not being tracked in Kudzu. Adding it — along with ‘conversation media’ and ‘conversation technology’ — will allow EdgeTheory to keep track of these phrases going forward.

Finally, we need to double down on our primary keyword — ‘conversation’ — to ensure it is being deployed as often as possible. We may never crack a dent in the 45,000,000 or so times it gets used every month, but we still may be able to impact our Google rankings. I’m guessing that the majority of people who use ‘conversation’ are using it casually and sporadically. That is, they don’t repeatedly tweet about it. So it’s still possible for EdgeTheory to have more conversation share around ‘conversation’ than people who use it occassionally. This will still suit our purposes when the Google crawlers come along.

It appears I have my work cut out for me, then. I’m off to take care of these action items over the next few days — and to create even more blog posts around digital conversations as well. Doing so will give EdgeTheory even more credibility with Google. Be sure to keep up with the blog here and to check back in with us in a bit to see how we are progressing.

Until then,

G —

Our technology combines human and artificial intelligence so organizations can understand and participate in the conversations that shape their world.