Most marketing teams in the arts aren't short of numbers. Between your website analytics, your CRM, various reports and a spreadsheet or two, there's usually more data than anyone has time to read. And the question many marketers come up against is "which of this is worth acting on?"
If you've ever pulled a monthly report and struggled to make a decision off of back of it, these four steps might help.
They’re developed from a talk that our Digital Marketing and Content Specialist, Ella Bailey, gave at the AMA Conference 2026, based on the process she followed to land on an up-to-date and relevant set of KPIs.
We had some great chats with conference attendees about data-specific challenges they're facing and how our talk might help them tackle them, so we hope lots of you will find this helpful. None of it needs a new platform to begin, and you can adapt each step to the way your team already works.
Before you commit to tracking a metric, ask three questions of it:
Track the answers in a way that makes it easy to compare the results - e.g. a filterable spreadsheet. It's a fast way to spot the metrics that are actually helpful, work out which are not, find any gaps in the data you’re tracking and reframe numbers that are “almost there but not quite”.
You might be able to gather context from a number alone, but just in case you can’t, there are a few tools you can use for this purpose.
The first are leading and lagging indicators. These are basically a remedy to the fact that often, the initial result of marketing activity and the final desired result are tracked by different metrics, at different moments in time. Leading indicators measure early activity you can still influence, like click-throughs to an event page. Lagging indicators measure the result after it's happened, like sign-ups. Line them up as a chain, say sessions, then engaged sessions, then conversions, and you can see where people drop off and begin to work out why.
When you're trying something new, try using hypothesis-led metrics - these help work out what to track and what decision it might help you make. Write a short hypothesis: “if we do this, we expect that, measured by this”. Track a guardrail metric alongside it to check the experiment isn't harming something else, like keeping an eye on engagement time and lead quality while you try and improve your click rate, for example. Tracking both allows you to make decisions as to whether something was working or not - because the hypothesis-led metric suggests whether the result matches your predicted change, and the guardrail metric confirms your results don’t come at the expense of something else.
And when a number shows scale but not depth, pair it with a metric that shows quality - for example, engagement as well as views. It can be helpful to place these together in visualisations (e.g., charts) or next to each other in spreadsheets - the two numbers, seen together, will give you a much better idea of where you actually need to take action.
Once you’re confident in your data set and have built context around it, the next challenge is acting on it. A good place to start is to frame your reviews or reporting around three kinds of question: diagnostic (what happened and why), forward-looking (what you do next) and reflective (whether you're tracking the right things at all). This avoids falling into a pattern of explaining changes in numbers without actually agreeing what to do next.
It’s also a good idea to agree the action that you will take if a number changes, as well as the person who will take that action, in advance, so a number that moves prompts a response rather than a debate.
It helps to be realistic here too. A small team can't act on every insight, so weigh up whether you've the time and resources to act on something, how long results will take and whether there's a quicker win elsewhere.
And get all your numbers into one place while you're at it, with no maths to do or additional reports to pull before it's readable - use automatic updates or visualisation platforms where possible.
The data set you build won't be the data set you keep. For any metric you're thinking of dropping, ask whether you can explain why it's tracked, whether it's ever triggered an action and whether it has a clear owner. If the answer to any of those is no, it's time to change it, hand it over or retire it.
That willingness matters more than it might sound. There's no final form for this: what you track and how you track it will keep changing as your work does, and being ready to change it is what ensures that you keep the gap between numbers and action to a minimum.
Our conclusion after going through this process is that the real payoff is confidence, rather than a perfect dashboard. Confidence in the numbers that we’re tracking and sharing, and in the everyday decisions behind them. It’s a project that buys back time at every review once it's in place.
We're still gathering notes on what venue teams actually do with their data, and we'd love to hear yours. If you've got a process that works, or challenges you’re facing, do get in touch and share it.