Identifying Your Audience: a Data-Driven Approach to Content Planning

Recently I was working on a piece of content planning for a new client. This client operated in a space that I had not worked with before and I was faced with the most terrifying thing.

A blank slate.

Now my colleagues have written extensively about content, ideas and planning (We even have a tool for generating ideas). I, however, love data. I also love having an excuse to open a giant spreadsheet and start diving into statistics when someone asks “why?”.

Aims:

By the end of this article you will know how to:

• Determine who is influencing your customers


• Determine what sites your customers are likely to be on


• Determine trends in the content that your customers are likely to read and share

Tools we will be using:

Followerwonk

Gwittr

Buzzsumo

Excel skillz

Warning

Some of these steps are very manual and time-heavy. Hopefully in the future utopia tools will exist to do them for you, but for the moment exercise those CTRL, C and P buttons.

Depending on your client and how much time you are prepared to spend on this, there is the potential to gather huge amounts of data. Save your work and save often. If you have version control, even better.

We were easily approaching 300,000 lines of data at some points.

Overview

The process is neatly summed up by this handy image.

Sounds straightforward right?

Right.

Step 1: People

Get most influential twitter followers

Get a list of followers from Followerwonk

Order by Influence

Collect twitter handle for the top 10 (trust me, you wont want to do much more than that). These are our level 1 influencers (from here on: 1i’s).

Get their most influential followees

For your FIPs, use followerwonk again to get a list of all the people that they follow. Put each of these lists into its own table in it’s own tab.

Pull out the most influential 10 of that whole set. These are your level 2 influencers (from here on: 2i’s)

Work out the intersect

In new tab create a table to check for each 1i whether they are following each 2i. It will look something like this:

 

From here you can discount anyone who is followed by less than a handful of your 1i’s.

Those are the people who you want to share your stuff

The highest scoring accounts here are the ones we want sharing our content. They are followed by our most influential followers, so the chance of more people seeing anything that is shared is higher.

Warning:

There is a risk that the results of this step can be very generic or obviously generally highly followed accounts. Celebrities with huge followings, news sites, etc.

If this is the case, you can filter it down by using, for your 1i’s, your top 10 influencers who have a specific keyword in their bio. This is a bit of a blunt instrument, but if your data set is large enough, should be fine.

Step 2: Websites

Get the influencers top shared sites

Once you have a list of 2i’s, put each of them into gwittr to get a list of top shared domains.

Collate. Shared by more accounts increases visibility.

Put all these domains into a list and just count how many times each one appears. The higher the better. These are the sites that are shared more often, so people are more likely to visit.

Step 3: Content

For top sites get most popular articles

For each of your top sites, put them into buzzsumo to get a list of the top pages by number of social shares. Write down the titles of the articles. you’ll want in the region of 10 per site unless the shares drop off dramatically, you can use your judgement here.

determine trends

Look through all of the titles, what patterns can you see? Are there, for example, a lot of “How To” style articles, or videos? Whatever it is, that’s what you want to be producing.

Conclusion

Deliverables

The 1i’s and 2i’s. A list of people to target in outreach or for expert advice in your content.

The top websites. A list of websites to target in your outreach.

The common themes in highly shared content. Suggestions for topic, style, format and tone of your content.

limitations

Because many of the tools only allow you to do one query at a time, there are several points where we are taking the top X number of results, where ideally we would want all of them. This means you are looking at a small section of your available dataset.

next steps

Come up with your ideas, approach your people and websites with them to see if they like them.

make it

profit

The post Identifying Your Audience: a Data-Driven Approach to Content Planning appeared first on Builtvisible - A Creative Digital Agency.

Identifying Your Audience: a Data-Driven Approach to Content Planning