Like any process, the first step is one of the most important and can often make or break the success of your efforts. This is especially the case in deciding on the topics you are going to write about on your blog. After all, if you choose a topic that your audience isn't interested in, it is pretty obvious that they probably aren't going to read it.
Generating Relevant Content Topics Using Twitter & Email
Like most brands, we have a content strategy that both resonates with our readers and aligns with our overall marketing goals. Unfortunately, it is very easy to lose focus and create disparity between what we think our readers want and what they actually want. Yes, we can check to see how our blog posts are doing on our social channels or check out the various engagement metrics on Google Analytics, but that's not always enough. Sometimes we need more than just the numbers we rely on daily to make educated decisions. So, that is why we decided to explore the same avenue only in a different direction using two of our main promotion channels.
I'm going to start with a platform that most, if not all, bloggers and publishers are using - Twitter. One good way to find out what kind of content your followers value is to export your click data for your tweets and see if there are any trends. Although this does work, there are also many factors that may impact your click rate like time of day, day of week, the amount of characters, and the use of hashtags. For that reason, we decided to analyze the data that nearly every one of our Twitter followers provides that is not affected by any of the previously mentioned factors - the bio.
Step 1: Export the Data
There are many ways to do this, but we used Simply Measured's Free Twitter Follower Report. I recommend this tool because it is free, it provides you with all the data that you need in order to complete this analysis, and it provides you with a lot of other data that you may not already have.
Step 2: Compile the Bios
After you have exported the Excel file, go to the tab titled "Follower List". This is all of the raw data that supports the other worksheets in the report. You can explore all of the data that it offers you, but for the purpose of this analysis, we only need the bios. Copy all of that data and put it in a separate spreadsheet.
Step 3: Clean Up the Data
This is a tedious and, depending on how many followers you have, time-consuming part. You need to separate each word of every bio into its own cell. We used Excel's "Text to Columns" feature (Data -> Text to Columns...) with a space selected as the delimiter. From there, we copied all of this data into one column, and then removed any unnecessary symbols. After you have done all of this, you should have a fairly extensive list (our list was a little over 65,000) of keywords all in a single column.
Step 4: Aggregate the Data
Make a copy of the worksheet with all of the keywords and remove all duplicates so that you have a list of the unique keywords. Then, drawing the data from the full list of keywords, use a "=countif" equation to get the number of times that each keyword appeared. From here, order the list by count from highest to lowest and voila! You now have a list of the most popular words that appear in your followers' bios.
Step 5: Clean Some More & Visualize
I consider this last step optional, but am sharing because I think it is a cool way to further visualize the results of your data. We removed any stop words or other words that have no meaning (a, an, of, is, at, the, etc.) and got a list of the top 100 words. From there, we copied these into Word It Out and formed a word cloud which you can find above.
What do you think? If you follow Zemanta on Twitter, or are a regular reader of our blog, do you think this is an accurate summary of content that you would be interested in?
This approach is probably not applicable to many of you, but I decided I would include it anyway in hopes that I would either inspire you or spark some productive brainstorming. As many of our readers know, we have a pretty awesome newsletter that we send out every week. This newsletter promotes not only our content that was created during the week, but also content that we find around the web that we think our subscribers will find interesting. By doing this, not only are we able to see which content to feature in our newsletter, but we also gain an insight as to what content we should produce that will resonate well with our subscribers.
Step 1: Export the Data
First, we exported all of the newsletter data from the past 6 months and put that into a spreadsheet. This could be pretty tedious depending on what email client you use. We use MailChimp and as far as I know, you have to export a separate report for each email blast you send out.
Step 2: Aggregate (and clean) the Data
After getting all of the raw data into a spreadsheet, we combined it and organized it so that we had 6 primary columns: URL, Domain, Blog Post Title, Content Focus, Clicks and % Total of Clicks. This allowed us to see the URL of the link that was provided in the newsletter, the domain of that blog post so that we could see which external domains were providing the best results, the title of the blog post, the content focus (I will get to this in the next step), the total amount of clicks that URL received, and the percentage of the total amount of clicks that issue of the newsletter received (for comparisons sake).
Step 3: Label by Content Focus
Now you need to get the only data that isn't provided during the export: the "Content Focus". In order to get this, we went through each of the URLs and attempted to label it with a content focus - basically, what is the main category of the content. This was a bit tedious, and a bit difficult to come up with the labels and apply only one per blog post, but definitely worth it at the end.
Step 4: Combine and Analyze the Data
You can now aggregate this data to see which categories received the highest share of clicks from each newsletter. This may not be as effective for someone that only has a few emails worth of data, but we had six months' worth so it provided a pretty good look as to what content our subscribers were clicking on the most.
Now that you have a good idea of what both your Twitter followers and your email subscribers find valuable, you can begin to make decisions. For us, it is quite obvious that our Twitter followers value marketing, social media, blogging, technology, and SEO the most and that our email subscribers value PR, blogging, tools, content marketing and social media the most. Combine insights from both of these analyses and you have a good idea of what topics we should focus on in order to drive engagement and provide value to our audience.
So far it has worked fairly well. After seeing the results from this analysis, we decided to do a post on the best tools for content marketing and blogging. What happened? We saw the largest spike in traffic and engagement for that blog post than we have seen for any other in months.