If I know one thing about strategic marketing, it’s that relevance is crucial. Think about it—Google has built an entire empire by delivering highly relevant ads to the right people at the right time.
In September, I crafted an experiment to test how powerful marketing relevance could be in social media, particularly Twitter. My small experiment lasted only 24 days, but resulted in 913 click-throughs on links I sent to my articles! It cost me nothing to perform, but those results equate to about $4,100 that I would have had to spend if I did this through a PPC campaign. As a bonus, I also gained 102 highly engaged Twitter followers in the process.
Read on to see how I did it.
The idea was to contact people Tweeting about a certain subject and provide them a link to a relevant article on my site.
For example, if someone Tweeted, “Email marketing today is so tough,” I would send them a Tweet saying, “@username Saw you Tweeting about email marketing, thought you might like an article I just wrote about it http://bit.ly/RaTeoB.”
The trick to the experiment was making it scalable. I needed a way to avoid manually searching Twitter and sending out each Tweet. I also needed to automatically keep track of the number of Tweets sent and the amount of clicks.
I first used IFTTT which stands for “If This Than That.” IFTTT, if you’re not familiar, is a free online service that lets you sync online tools like Twitter, Facebook, Pusher, SMS, RSS and many more to do cool stuff. You could, for instance, push a Tweet out anytime you post a new YouTube video, or archive all your Tweets in a spreadsheet on Google Docs.
For me, I wanted to use IFTTT to automatically search Twitter for a specific key phrase and then automatically Tweet to the person who Tweeted it. There was a catch though, Twitter didn’t allow this. Luckily, I figured out a workaround. I used Buffer as a middleman in the equation. If you don’t know Buffer, check out the short video I created, How I Automatically Tweet Great Content Throughout the Day. To be brief, Buffer is a queue system to space out your Tweets according to a schedule you create.
So, I made a “recipe” on IFTTT to search for a key phrase like “email marketing,’” and then when someone Tweeted using that term, my script would create a Tweet to that person and add it to my Buffer queue. Awesome.
I then wanted to be sure I captured all the Tweets I was sending, so I added a recipe to store each Tweet I sent in a spreadsheet file on Google Drive, Google’s free cloud storage.
For the key phrase, “email marketing,” the whole process went like this on IFTTT:
- Search Twitter for “email marketing.” If found, go to 2.
- Add this Tweet to my Buffer: “@username Saw you Tweeted about email marketing, though you might like this article I just wrote http://bit.ly/RaTeoB“
- If I push a new Tweet, then add that Tweet to a spreadsheet on Google Drive
Tracking clicks was pretty easy. I use bit.ly as a link shortener which was already built into Buffer. Bit.ly automatically tracks any clicks on any links you make with it.
I used this process for three good articles I wrote:
- Why Your QR Code Marketing Sucks
- The 40/40/20 Rule of Marketing
- Email Marketing — 3 Counterintuitive Tips to Boost Results
The key phrases I used for the articles were, respectively:
- “QR code marketing”
- “Marketing tips”
- “Email marketing”
The technique I used for selecting the keywords was to choose a phrase that enough people were Tweeting about, so I didn’t hit the same audience over and over. I also wanted the phrase to be as specific as possible so it would be highly relevant. It’s very similar to a PPC marketing strategy.
The results were better than I could have imagined.
The leverage of social media marketing is incredible.
In 24 days, I sent out 593 targeted Tweets pointing people to my three articles. The crazy part was that I got 913 clicks! Because social media is, well, social, people were quick to share my articles after they clicked the link.
Here’s a quick recap of the results:
To estimate the amount I saved on advertising, I used Spy Fu, a free and fairly accurate way to estimate the cost per click for each key phrase I used. Spy Fu gives you a min and max range, so as you can see, this campaign could have costed me $4,524 if I had used a PPC marketing strategy.
During the experiment, I tried not to Tweet much in addition to the targeted Tweets. I wanted to see what all the other effects would be in a vacuum. Here’s what else happened during the 24 days:
- I was Retweeted 181 times – This clearly had an affect on the amount of clicks I tracked.
- I gained 102 high quality followers – I say, “high quality,” because these people were engaged. Most of the time they would Tweet back actually thanking me for sharing the article with them! Here’s the graph from TwitterCounter:
- My Klout score increased from 54 to 59 – Not a massive jump nor my favorite way to measure impact, but if you use Klout than you know how difficult it can be to move the needle in such a short amount of time.
Strategic Marketing Lessons
The experiment shows that relevance plays a huge role in marketing success. Even though the people I pushed targeted Tweets to might have not been searching for answers, they were definitely interested in the topic, which drove them to click.
But at the end of the day, content was indeed king. You’ll notice that my article about 3 Counterintuitive Tips to Boost Email Marketing Results had the highest click rate. This doesn’t surprise me since I spent the most time researching that and creating the article and title. It definitely had the best content and the click numbers showed it.
If I were to perform the experiment again or amend it in the future, I would add a control group where I tweeted irrelevant information to people so I could measure the lift. I didn’t do this because I didn’t want to tarnish my reputation of my personal Twitter account.
Unfortunately, Twitter has since put a stop to services like IFTTT using Twitter as trigger for other events on the Internet. They claim it’s to reduce the load on their servers, but I think the other reason had to have been to stop spammers. My technique, if it fell into the wrong hands, would really hurt the Twitter experience. I can see how it would be addicting to get those kind of results. It would just be a matter of time until people started to push crap and ruin Twitter for all of us marketers.
As I mentioned before, I tried to limit my Tweeting during the experiment to get the best read on the results, but the additional Tweets I did send were to the many people who responded to me. I didn’t want to be a spammer and it was a lot of fun connecting with the like-minded people out in the Twittersphere. I made sure to personally followup with every person who responded to one of my auto-Tweets.
Here are just a few responses I received:
All in all, I think the experiment was a success. It was nice to see first hand the quantifiable validation behind some of the ol’marketing rules we all know and love.
Are you surprised by these results?
What would you have done differently?
Leave a comment below!
Twitter has put a stop to services like IFTTT using Twitter to trigger things on the Internet. Apparently they wanted to reduce the load on their servers which was being hit hard by online tools that constantly searched their site. Unfortunately, this means that the technique I used for this experiment cannot be repeated.