One of the most important steps in getting a content strategy up and running is the content audit. It tracks and records content and hopefully uncovers all your skeletons in the (virtual) closet. In the second installment of our content series (check part one here) we talk about some of the do’s and don’ts in content audits with Emily Warn, co-founder of Two Pens.
Emily serves as an online and content strategist for high-profile websites and is also a poet and essayist. She is the founding editor of Webby Award-winning poetryfoundation.org and was the editor-in-chief of Microsoft.com. Most recently, she was the managing editor of the Microsoft Office Blogs , helping develop a blogging platform for the launch of the new Office 2013. At Two Pens, she trains business people in blogging and social media writing.
Flip the Media: Can you still remember your first content audit?
Emily Warn: The first one I ever did was at Microsoft, when we upgraded from Windows 95 to Windows 98. In those old days we called it “content migration” and we just used a spreadsheet. As sites got more complicated and were around a lot longer they had more irrelevant legacy content.
What is different today compared to those early days?
The major difference is that there are new tools that can automate content auditing. You don’t really need it for a small site but for a large site it is invaluable. But you still have to set up the structure. Especially for people entering the field it is important to remember that when you do a content audit – and it’s a given that sooner or later you’ll have to do one – you have to set up the directory structure before crawling the site. Content audit tools basically index every page of your site like a search engine and then produce a list of content based on your structure. But if you don’t structure it in the first place you have to do a lot of manual work later on.
So how do you go about setting about your spreadsheet or content audit tool?
You want to look backward and forward at the same time. You want to look backward and check how your content was classified which may or may not be the same as the navigation. Then you look forward and based on your research and new business goals you create a taxonomy. You can’t really do an audit without doing a taxonomy of some kind. A ‘taxonomy’ is just a fancy word for classifying content into categories.
With the taxonomy, you try to find all the content that fits under those categories. So for instance, let’s say you are at Apple and a gung-ho marketing team created a whole page about all the new improvements to Safari. It’s about new features but somebody put it under “How-to”. In your content audit it shows up under “How-to” but as a human being and not a tool you are going to go “That is not ‘How-to’; it is not going to help anyone use the product. Let’s move it over into product marketing descriptions.”
In a content audit you have to evaluate content. How do you come up with the evaluation criteria?
Basically, you want to align the content with the business goal. In order for content to achieve that goal it has to talk to the right people. It has to deliver the information they need in the context of a task they are trying to complete. So let’s say the task is evaluating a product in order to buy it – so then they need the features to consider and maybe compare against another product. You are always evaluating over task and business goal.
We talked about the difference between a small and a large website. With a large website, do you always look at every single page or do you sample if it is too large?
The Poetry Foundation has an archive of 10,000 poems. We knew we didn’t have the resources to check the quality of each poem so we sampled a percentage. Our brand was “high quality” so if someone found a poem or article with typos then our whole brand would be compromised. We found a lot of errors, so we hired people to do a complete audit.
So they haven’t yet developed a tool that evaluates the quality of a poem.
(Laughs) No, sadly no. Even most editors can not evaluate the quality of a poem.
A lot seems to depend on the “quality” of the content. How do you make that judgement?
Basically it is the content strategists who do that. If you need help you go to subject matter experts. But for instance, if it doesn’t meet brand requirements, that’s an easy one. If it’s outdated, that’s an easy one. It’s when you get into the writing and the design – that’s when it becomes subjective and you need to have some quality standards.
Content audits always pass judgement on somebody else’s work. What is a good way to tell people that maybe their content is not up to standard or no longer needed?
I actually don’t think most people mind. There is some conflict sometime but if you give them a reason, they’ll understand it.
Is there an example of a particularly hard content audit you did?
Let’s see, I did one for a Microsoft product. The issue there was that the content was duplicated – I found the same content in three or four different areas. I had to read through each piece of content to find which was the latest one. It was a small site but it took me about a month to figure it out.
Ok, let’s summarize. You’ve done all the collecting, the big spreadsheet is ready. How to go about the analysis of all that data?
I think before that point you already have your basic content strategy. People are using that word in all different ways but to me content strategy is how to use content to meet your business goals. So you analyze it in the context of that basic principle and decide from there.
The last step of a content audit is to present the results to the client. What do you need to keep in mind when you do this presentation?
It’s a sales job. You are selling them on how many editors you might need, whether it can meet the goals of the business, or whether it can engage the audience. But it really depends on your audience. If you are selling to someone who manages all the content in the organization, then you want to present in greater detail. So you want to say for example, X amount of pages have no metadata whatsoever. If it is a VP or director you are not going to care about the ID of the pages or how many of H1 headers. They want to know: Is metadata part of the problem? -yes or no.
A final question about content production. How do you get experts with something to say to regularly submit content?
It’s difficult. Everybody thinks they have a book in them. And everybody thinks they have blog posts in them and all a blog is is sitting down and writing in the first person “I” and you’ve got it. But it takes time and people are busy. So they end up not doing it. So you have to choose between a) your blog is going to suck because you don’t publish often, or b) find people who are good and go back to them over and over. Within in every organization you have to look for the people who actually like to blog. They don’t even have to be that great a writer, they just have want to talk to people. They build a following and once they do, other people can fill every once in a while.