Make sense of audience feedback, survey responses, attendance figures and website analytics in a new way.  

We all collect lots of data on our audiences and visitors but don’t have the time to turn that information into something that can help our decisions. Artificial intelligence offers a way to work with data that is safe and secure. Many of us may be apprehensive about using AI but using it for data analysis is one of the biggest wins.  

This is a practical course designed for anyone working with audience or visitor data. Data analysis skills are not required. You will use realistic sample datasets to discover how AI can help you in a safe supported environment. 

Understanding Your Data comprises of four two-hour live online sessions on Zoom. 

Each session combines taught content with hands-on exercises. You will build and test practical solutions throughout the course with sample data provided. 

If you have any access requirements, please contact sophie@a-m-a.co.uk

  • Session 1: Making sense of what people say 
    10am–12pm, Wednesday 10 March 2027 on Zoom
  • Session 2: Making sense of what people do 
    10am–12pm, Wednesday 17 March 2027 on Zoom
  • Easter Break: No sessions
  • Session 3: Bridging the gap
    10am–12pm, Wednesday 7 April 2027 on Zoom 
  • Session 4: From insight to action
    10am–12pm, Wednesday 14 April 2027 on Zoom

Each session includes a 10-minute comfort break about halfway through. 

You may be collecting plenty of audience data but just can’t find the time to answer the questions that matter. Survey responses and audience feedback takes hours to read and process. Different data sources may be telling different stories. 

AI rapidly speeds up this type of work as it can quickly find patterns in large datasets. However, AI can go wrong – without clear instructions it can make claims across data sources. So we will spend time testing and evaluating outputs so you know how to protect against this. 

This course is for people whose work involves marketing, audience development, communications, digital, visitor experience, fundraising or research across arts, cultural and heritage organisations. 

It is particularly useful if you work with audience data but are not a data analyst.

Cost

Standard Member rate: £250 + VAT 

Small Organisation / Between Jobs / Freelance Member rate: £150 + VAT 

Non-member rate: £395 + VAT 

Book now
A person holding a clipboard looks up

10am–12pm on Wednesday 10 March 2027, Zoom 

In this first session we will look at how large amounts of unstructured audience feedback in surveys, comments and reviews can be turned into structured insights. 

This session covers:

  • Sources of audience feedback and how to anonymise them before any data goes near AI 
  • Data safety and a review of AI tools  
  • How AI can support clustering, theming and insight extraction 
  • Checking AI’s claims against the original data 

Exercises include:

  • Clustering feedback comments into topics 
  • Extracting themes with built-in ‘uncertainty markers’ using the double-pass method 
  • Building insight cards with testable claims and confidence ratings 
  • Catching AI making things up by tracing claims 

By the end of this session you will have: a repeatable workflow for turning audience open-text feedback into evidence-based insights and a data safety and security checklist. 

10am–12pm on Wednesday 17 March 2027, Zoom 

This session gives you a practical, non-technical way to work with numbers, including attendance data, demographics and engagement metrics. You will use AI to spot patterns, ask better questions and recognise when the evidence is too limited to support a conclusion. 

This session covers:

  • What quantitative data is and what you probably already have in ticketing, CRM, email, web analytics and donations systems 
  • Working with attendance figures, demographics and engagement data to find useful signals 
  • Understanding sample size, confidence and when your data is too thin to draw conclusions 
  • Where AI helps with numbers and where it misleads: the difference between pattern and proof 

Exercises include:

  • Giving AI a sample dataset and interrogating what comes back: what is useful, unsupported or invented? 
  • Interpreting a web analytics export to identify what matters 
  • Spotting gaps by identifying what is missing from the data and why it matters 

By the end of this session you will have: a practical workflow for analysing quantitative audience data with AI and greater confidence working with basic audience datasets and web analytics exports. 

10am–12pm on Wednesday 7 April 2027, Zoom 

This session brings qualitative and quantitative data together. You will compare what people say with what they actually do. We will find missing opportunities and questions worth testing. 

This session covers:

  • Why qualitative and quantitative data often tell different stories and why you need both 
  • The Say vs Do comparison: pairing insight cards with behavioural data 
  • Checking whether qualitative and behavioural data relate to sufficiently similar audiences, activities and time periods to be compared 
  • Reading the results: what agreement, disagreement and missing information each tell you 
  • Using relevant sector benchmarks where they are available, without treating comparison as proof 
  • Moving from insight to hypothesis: deciding what you would test next 

Exercises include:

  • Running a Say vs Do comparison with an insight card and a behavioural dataset 
  • Comparing one aspect of digital performance with an appropriate benchmark 
  • Identifying one mismatch and designing a low-cost test to investigate it 

By the end of this session you will have: a Say vs Do comparison template, a clearly framed hypothesis and a practical test you can take back to your team. 

10am–12pm on Wednesday 14 April 2027, Zoom 

The final session brings together the methods from Sessions 1 to 3 and turns them into a repeatable way of working. You will follow a prepared audience question from raw data through to an evidence-based insight, then adapt the complete process, or the parts most relevant to your own organisation. 

By this point you can produce an analysis. The harder part is now deciding what question you are answering, what evidence it needs, where AI can help and what you need to check. That is what this session is for. 

This session covers:

  • Taking an audience question from raw data to a supported conclusion 
  • Connecting qualitative feedback, quantitative data and relevant benchmarks without overstating what they prove 
  • Turning findings into a concise insight with its evidence, its limitations and a clear next step 
  • Bringing the course methods together into a flexible audience insight process 
  • Identifying the data, instructions, checks and decisions each stage needs 
  • Where automation or agents might help and where your judgement remains essential 
  • Adapting the process for your own role or organisation 

Exercises include:

  • Working through an audience insight challenge using a prepared dataset 
  • Producing a three-line insight summary: the finding, the evidence behind it, its limitations and what to do next 
  • Bringing together four reusable workflows: an open-text feedback analysis, a quantitative data check, a Say vs Do comparison, and an audience insight report 
  • Creating a simple implementation plan covering your question, data sources, AI tasks, verification steps, responsibilities and first test 

By the end of this session you will have: a completed audience insight report, templates for four reusable workflows and a practical plan for putting the process to work for yourself.  


  • Ensure your data is anonymised and clean for AI analysis 
  • Turn survey responses and social media feedback into evidence-based themes  
  • Compare what audiences say with patterns in their behaviour 
  • Use AI to investigate quantitative data without needing to become a data analyst 
  • Conduct trends analysis over historical data 
  • Ensure you trace any AI-generated claims back to the evidence 
  • Turn potential insights into practical next steps