1. Who are you reaching and engaging?

For some projects the aim is to reach the broadest and largest possible audience. In this case, measuring the total number of people who engage with your work might be all the insight you need. For other projects – perhaps the majority – the aim is to engage with specific groups of people. In this case, a well-designed evaluation can help you understand the types of audience engaging and therefore whether your marketing is well targeted. You might also see whether different types of audience engage with your work differently.  

There’s a range of ways you could classify your target audiences to help you plan this. These might include: 

• Where they live – e.g. are they local to you, further afield, perhaps even overseas? 

• Are they new audiences or people who have engaged with your work before? 

• Do they have certain demographic characteristics – e.g. age, gender, socio-economic background?  

• Might they have interests or needs that draw them to your work – e.g. a hobby or passion for a particular artist? 

Sometimes it can be helpful to combine these characteristics into audience “segments” and then to bring these segments to life by imagining them as fictional “personas”. For example, “Grace is a 20-something student who lives nearby but has never visited our venue and is really passionate about jazz”. Then your next set of questions become: “How can we make sure this digital work appeals to people like Grace?”, “Where might we focus our project’s marketing to reach them?” and “How will we know if we’ve been successful in engaging them?” 

The art of the possible

So far so good, but this is where evaluation theory meets digital reality. Most of the platforms we commonly use to share work online provide limited information about audience types. Also, changes in data protection laws in recent years, while understandable in terms of protecting peoples’ privacy, have the side effect of making this type of data harder to collect. Platforms are providing more privacy controls, and more people are opting out of allowing themselves to be tracked online. 

So, you will probably find that what you can glean from the data on your social media channels, or your website analytics, is limited to the broad location and a guesstimated age range and gender for your audiences.  

But don’t let this put you off being specific about who you’re trying to reach. It’s just that you may need to look slightly sideways for ways to define and measure this. For example, you might be able to tell from topics of interest what people’s reasons are for engaging with your content. Or perhaps you can classify them based on how they reach your content or where they go to next (see How are audiences finding and interacting with your work? for more about this). This type of information can help you to shape your audience segments and personas around what it is possible to track online.  

When diversity meets privacy 

Reaching diverse and under-represented audiences is often a goal with digital work, in terms of protected characteristics such as age, race, religion, or disability. Given the sensitivity of this information and the data protection constraints we’ve just mentioned, directly and anonymously surveying your audience is usually the most reliable way to measure diversity. But a survey may not be practical for your project and, even if it is, response rates are often low.   

So, how can you develop a better understanding of audience diversity when direct surveying isn’t feasible? A useful approach is to look at what you can understand from the online routes you use to reach your target audiences. For example, you might work with social media influencers who actively engage with a particular under-served community. You might secure media coverage on websites that already have a loyal following among the audiences you want to reach. Or maybe grassroots community engagement in an offline setting is a useful way to reach a specific geographical/interest group.  

In all these situations, look at how you can track your engagement with influencers and partners. And then consider if you might be able to track which of their communities then come to your work (see How are audiences finding and interacting with your work? for tips on this). This data then becomes your measure of success in reaching the hard-to-reach. 

Are you new here?

What if your aim is to grow the number of new people you are reaching? This is where it pays to benchmark your activity: put simply, you can only know whether you’ve successfully increased the size of your audience if you know what size it was at the point your project started. 

Here, it’s important to consider how you are framing audience growth. There are two basic approaches: the first means thinking about growth broadly at a brand level across all the content you publish. If this is your priority, you might track overall subscriber/follower numbers on your social channels or total visitors to your website, using a free tool like Google Analytics 4 (Digital Culture Network has a useful Introduction to Google Analytics 4). This approach is most relevant if you run frequent digital activities that broadly target the same audiences each time.  

But what if your digital activities are less frequent or each one tends to target a different audience? Then your framing of ‘growth’ might focus on the numbers of people you reach with each project or each piece of content. Growth then becomes more about increasing the average reach of your marketing activities each time, so you know you are improving your ability to engage people. It becomes less about building brand-level followers.  

This is not to say you must make an either/or decision. Both types of growth measurement can be valuable to monitor. It is just about considering in advance what you want to learn. This will determine what is important to benchmark and then periodically monitor.  

And what about if you are starting from scratch with no prior activity to benchmark? Then in many ways your job is easier. Simply track your follower or visitor numbers on a weekly or monthly basis from a starting point of zero. Combine this with looking at the reach of each piece of content and this will give you a sense of how your audience is growing and your approach to production and marketing is improving.  

Are we over-reaching ourselves?

To avoid getting bogged down in detail, we’ve allowed ourselves to get quite far into this article without defining what we mean by ‘reach’. However, particularly with online work, there are lots of different ways of understanding and measuring reach. So it’s important to be specific.  

Broadly, when we talk about ‘reach’ we mean the number of people who have experienced our work, however superficially. A big complication here is that different platforms have different metrics for reach. This means you are almost never comparing like with like. For example, to register a ‘view’ on YouTube your video must be playing for at least 30 seconds, but on social media platforms a view can be registered the second your video starts playing (e.g. TikTok, Instagram stories) or after three seconds (Facebook, Instagram). That doesn’t mean that one is better than the other. It’s important to bear in mind the individual culture of each platform and the different experiences users go there to have. 

What it does show is that it is meaningless or even actively misleading to add up reach measures across different platforms or to see these numbers as a proxy for genuine engagement with your content. Are you really interested in counting every single encounter with your work, such as a video that auto plays while someone scrolls past? Or is it more important to understand whether people have engaged for a particular length of time or in a meaningful way? 

Of course, we understand the lure of adding up and quoting those reach numbers. They are often the biggest numbers in our data dashboards and they make us feel good about the success of our projects. They might also seem like positive evidence to cite in our reports to funders or other stakeholders. However, as funders themselves will tell you they much prefer a more considered approach to digital evaluation than simply quoting large and perhaps largely meaningless numbers. You can mention headline reach numbers but then it’s good to move on to more revealing insights. 

Let’s get engaged then 

So, what do we mean by ‘engagement’? This is the extent to which people interact with or experience the work you have created. Ways to measure engagement can include: 

• Duration measures: e.g. for video or audio content – the average view/listen times or percentage of a piece of content played; for interactive works such as websites or games – average visit duration.  

• Interaction measures: e.g. for interactive experiences – the amount of content that audiences navigated through or interacted with; for co-creation or participatory experiences – the number of users that contributed content or the amount of content contributed.  

• Reaction measures: e.g. the proportion of audiences that liked, shared or commented on an experience. 

You can see that these measures give different ways of understanding the extent of audience engagement. Picking which measures to use means considering which aspects of audience behaviour are most important for your evaluation aims. You need to balance this with the usual practical constraints of what you can measure with the time and tools at your disposal.  

Finally, remember that context is key when you are interpreting your engagement data. If people are spending a long time on your website, that might not mean they are deeply engaged – they could be struggling to find what they are looking for. And if the comments on social media are negative in tone or practical (e.g. asking for help), then talking about total comments received may not be a good measure of engagement.  

An action is not a person  

As a final piece of advice in this section, remember to think carefully about how the data you are using does or doesn’t correspond to numbers of people. If the same person likes, comments and shares a post on social media, that is three different engagements or actions but it is not three different people. Types of data that most closely correspond to individuals usually have ‘unique’ in the title: unique users, unique views, unique listens etc. Unique metrics more accurately correspond to the number of people who took an action on the same platform, rather than the number of actions.