Did you know you can try GatherContent free for 30 days? 🤑 Free Trial →

Content Strategy

6 steps to using experiments to create more valuable content

Jamie Griffiths • 7 minutes

Anyone who’s been tasked with devising or executing a content strategy for a new brand, product or campaign knows the problem well: how do we come up with something that we know will resonate with our audience and speed them towards their goals?

In this post, I’ll outline an experimental approach to content production which has its roots in Lean/Agile methodologies. It will help you find messages that strike a chord with your target audience quicker, so you can produce better quality content that meets your business goals, using fewer resources and with less risk.

Overcoming the fear

The worst outcome for any content strategy is spending ages designing, researching and building well-written, great-looking content which falls completely flat. I have a theory about why we experience this: because we’re afraid to fail. We’re aware that we have gaps in our understanding about our audiences and their needs, so we delay putting content in front of them by overcooking it. Once we accept that we don’t know everything there is to know about our audience, and that to a degree, failure is guaranteed, we can get on with making our mistakes, learning from them, and building better, more successful content.

How much do you really know about your audience?

I hate to be the one to break it you, but you know those audience personas you spent ages brainstorming, researching and writing? The ones that are supposed to help you produce killer content which leads to huge engagement and tons of goal completions? They’re wrong. Useful maybe, but wrong.

How do I know this? Because for a start, they’re just personas. They try to represent an entire audience segment with a single, fictional character. Unless your audience is incredibly niche, there’s no way you were able to poll every single member of that audience on what their needs, interests and pain points are. And even if you spent many years holding focus groups with every single person who falls into a particular audience segment, the information you extracted isn’t necessarily accurate. People are way too complex and changeable. They often don’t know their own minds or are unable to express themselves clearly and that’s before we get to your ability to accurately interpret what they’re saying.

Once we accept that we don’t ‘know’ anything about our audiences we can recognise what we do have for what it is: a set of assumptions about our audience which needs to be tested in real-world conditions.

We need to design some experiments, with the goal of validating or invalidating the assumptions we’ve made about our audience. Note that proving ourselves wrong is a valuable outcome here. Whether our test content succeeds or fails, we move closer to a set of assumptions about our audience that will help us produce more effective content.

Let’s explore how taking a lean/agile approach to content production can help you produce better content, faster and with less resources.

Step 1: decide what you’re testing

Our first task is to define clearly the assumption that we’re testing. Which assumption should we choose to test first? The one that will lead to the most value for the business. This will vary from business to business, depending on where you are in your journey. If you need to build an audience then developing your knowledge around the higher aspirations, dreams and interests of your audience will be valuable. If you’re already getting lots of traffic, then assumptions about preferences and needs which affect conversion might be the place to start.

Let’s make this concrete. I recently helped a startup ecommerce analytics platform devise their first content strategy, the goal of which was to drive interest and awareness in the, as yet unreleased, product. One of the key audiences the client wanted to engage was accountants in private practice. We had three assumptions about this audience:

  • that they had spare capacity which they wanted to sell to existing clients
  • that they wanted to become all round business advisers rather than mere compliance specialists
  • that they wanted to make better use of their clients’ data to generate more valuable insights

We didn’t have the time and money to go out and survey a load of accountants so, instead, we came up with the following messages, each of which tested a different assumption, to test via a LinkedIn sponsored posts campaign:

  • Accountants: redeploy those unsold hours to higher-margin services
  • Accountants: learn how to become an all-round business adviser to your clients
  • Accountants: make better use of your clients’ data to generate valuable business insights

You get the idea. We need a set of key messages around which we can quickly build content to test our assumptions about the audience’s drivers.

Step 2: set meaningful, valuable goals

How will you know if your key messages are effective or not? Any experimental content needs to drive audience members towards some kind of goal, even if that goal is as simple as clicking on an ad.

To minimise wasted effort, and start generating value as quickly as possible, the goals should reflect your actual business goals. If your aim is building an audience then clicks and shares might be enough, but if your aim is to generate sales leads then try to get your audience to fill in a form.

The journey you come up with for taking the audience from the content to the goal should be as simple as possible to avoid muddying our results. Remember, we’re testing our key messages, so we want to eliminate other factors as potential drivers of goal completion. Make sure any and all touch points along the journey (social media ads, landing pages etc) are built around the key messages that we’re testing.

Our LinkedIn ads targeting accountants all pointed to the same short blog post which spoke to all three assumptions. Because we were primarily concerned with testing our key messages, the goal was simply clicking on one of our sponsored posts.

Step 3: find your audience

We need to put our experimental content in front of as many pairs of eyeballs as possible. Ideally, eyeballs which closely match the characteristics of our audience segment. Social media advertising, with its wealth of targeting options, provides a quick and easy way to find relevant people, but with a price. Accurately targeted search or display ads are another way to buy an audience, which might be more suitable to your customer journey.

Publishing on your website or via organic social media is cheaper but it might be a while before you have enough data to draw meaningful conclusions. The question to ask yourself is ‘how do I generate the most value using the least resources?’ If you’re trying to formulate a content strategy for a new flagship product then the ad spend may well be worth it. If you’re much further down the road, trying to make incremental gains, then it may not.

With the option to target by job role, company size and industry, LinkedIn was great for isolating an audience of accountants in private practice. We didn’t need to spend much to work out which of our key messages were effective at engaging the audience, and which needed to be discarded.

Step 4: run a controlled experiment

Controlled is the watchword here. We have an assumption about our audience’s needs, pain points, interests or preferences that we think we can utilise to drive them towards a goal. Don’t muddy the waters by introducing messages which are designed to address other assumptions you’ve made about your audience.

Think Minimum Viable Content, where the two key criteria for viability are:

  • It delivers the key message(s) which address the assumptions we’re testing;
  • It drives the audience towards completing the goal.

Think quick and easy. Think blog post rather than whitepaper. Think Facebook post rather than blog post. Think image rather than infographic. Think 150-word unstyled landing page rather than a complex one which will take your developers half a day to build.

Until we’ve validated the assumptions we’ve made about our audience, why waste time and resources on content which is built around those assumptions? Why take the risk?

It took only a few hours to write the blog post and LinkedIn sponsored posts for our audience of accountants. From that we learned that one of our key messages was effective (‘redeploy those unsold hours to higher-margin services’). Imagine we’d spent weeks putting together a beautiful whitepaper, which, because it was based on the same limited understanding of our target audience’s motivators (only 33 per cent correct), would have been doomed to failure.

Step 5: analyse the results

At this stage it’s important to stick to the measurement of success that you defined in step 2. Our understandable fear of failure can cause us to spin a failure as a success in order to justify the time and effort we spent on making the content. But if we remember that failures generate value too, by deepening our understanding of our target audience, then we can get over our hangups and call a dud a dud. Also, if you’ve been following this advice closely, the time and effort you spent on producing the failure shouldn’t be worth crying over.

So, if driving newsletter sign-ups was your goal, don’t spin a piece of content that generated zero sign-ups as a success because it attracted a load of clicks. Instead, subject the assumption you made about your audience to intense scrutiny, then come up with a new version, and design an experiment and some new content to test it.

Step 6 : back the winners

It might take quite a few experimental pieces of content before you come up with a successful recipe. (This is another reason to make your experiments as targeted, and as low-risk as possible.) But your successes will point the way to a lot more success, because you’re starting to build a recipe for what works with your target audience. The key message that worked so well as a Facebook post can now be adapted for use in a variety of other content types and channels. You can now start weaving it through calls to action, tying it to the concrete benefits of your brand, product or service.

Out of the key messages we tested with our accountancy audience, one was a clear winner, attracting click-through rates of over 3%. That was the key message that we would take forward to build a content portfolio around – producing valuable content that would speak directly to that concern. Content which would show accountants how to redeploy their unsold hours to higher-margin services, and explaining how the service my client offered could help them do it. Meanwhile, we’d continue running small-scale experiments to find other effective messages.

Once you’ve validated two or more assumptions about your audience, your possibilities for generating successful content ideas starts to increase exponentially. And once you’ve formed a pretty good understanding of what content is effective at driving your audience towards goals, you can start using experiments to find the most effective content types, lengths, channels and content promotion tactics.

Start embracing your (small) failures

Putting lots of time and resources into producing content that’s based on a hazy understanding of your target audience is a high-risk move. Use small content experiments to validate (or invalidate) the key assumptions you’ve made about your audience’s needs, interests and motivations. Ditch the messages that fail to help you achieve your goals and double down on your successes, using the effective messages to build out higher-value content that is more effective still.

While the approach outlined above is most obviously relevant to content marketers, there’s no reason it can’t also be applied to website content. Tools like Google Optimize make it cheap and easy to test out the effectiveness of different headlines, calls to action, images etc. As long as you remember to keep the costs of the experiment proportional to the value that can be created, and that you remain unafraid to call out a failed experiment for what it is, then a high-performance content strategy is only ever just a few small failures away.

6 steps to using experiments to create more valuable content

Free Online Masterclass

Content Strategy & Delivery

Kristina Halvorson

Anyone who struggles with content consistency and efficiency—and who doesn't?—will greatly benefit from this course. If you want to get smarter about content planning, creation and delivery, start here.

Kristina Halvorson

CEO & Founder, Brain Traffic

Reserve my place

Brand new book

Download your free Kindle, ePub and PDF copy

iPad with Content Delivery cover

Easily organise and produce content for website projects

Start your trial now
Mark McClendon
Our content processes used to involve an ineffective mix of spreadsheets, documents, and a slew of file storage solutions. With GatherContent we've solved all of those problems and more. Mark McClendon — Partner & Executive Director, VML

Late content from clients will delay projects and website launches

Content challenges and inefficient processes result in bad outcomes for your agency’s projects, letting down clients, delaying projects and losing you money.

See how agencies use GatherContent

Get the most from GatherContent

Book a free 30 minute call with one of our team to help you get set up with GatherContent and answer any of your questions.

Book a 30 minute call

About the Author

Jamie Griffiths

Content strategist and copywriter

You might also like these posts…

© GatherContent 2018 WeWork, 2nd Floor, 115 Mare Street, London E8 4RU, United Kingdom. VAT No.: GB140105279. Company No.: SC400199