My notes from BCS SIGiST Summer Conference

My personal notes from BCS SIGiST Summer Conference 2025.

Here are my notes from BCS SIGiST Summer Conference 2025. A mix of talk summaries, reflections, and key takeaways from the sessions that stood out to me.

If you're after my overall impressions of the conference, head to: Top 3 Takeaways from British Computing Society, Software Testing Conference 2025

This post includes write-ups on:

• You are not your customer - Emily O'Connor

• Testing: a solved problem in engineering? - Callum Akehurst-Ryan

• Failure Is Not an Option: Testing for Resilient Space Systems - Beth Probert

And highlights from:

• How Testing Makes Retail Greener! - Sivaprasad Pillai

• Beyond Code - A Tester’s Approach to Pull Request Reviews - Andrea Jensen

• Testers of the Future - Gerie Owen

• How Do You Test an AI-Based System? - Bryan Jones

---

This is a lengthy post, so it may be best viewed online, as email clients can sometimes truncate the content.

---

You are not your customer - Emily O'Connor

Key Takeaways

• Assumptions can help uncover new use cases, but they can also hide bias and blind spots

• Average users are rarely represented in development teams, so we must actively challenge our defaults

• Biases around age, gender, education, and income affect how systems are experienced

• Not everyone thinks or perceives the same way. Understanding this can help us design and test more inclusively

• Constraints can be strengths. Emily’s unique perspective showed how asking more questions leads to deeper insight

• AI might support testers in spotting hidden assumptions and improving how we question our systems

Summary

As software testers, assumptions are almost our lifeblood. It’s how we uncover use cases we may have missed or overlooked. Emily’s talk highlighted how powerful it can be to notice where assumptions are being made, then challenge them. This opens up new ways of exploring and understanding systems.

With AI becoming more integrated, I think this could be an exciting opportunity for testers to deepen their skills. It becomes less about the testing itself and more about detecting assumptions, questioning them, and building systems that can check those assumptions and their outcomes.

It also reminded me how our own biases can influence decision-making. This starts to feel more like a psychological skill than a purely technical one. I wonder whether AI tools could help us identify hidden assumptions and support this way of thinking.

Notes

Topic: You Are Not Your Customer

Started off as a developer but liked doing things differently and moved into testing.

Who is the average Jo?

• More women than men in the US and UK

• Jessica is the most common name

• This raised a point about who we assume we’re designing for

What about the dev team?

• Be careful of assumptions. Averages hide nuance

• In software, this could mean we miss important use cases for real users

Test plans

• Often stuck in default formats that don’t reflect real users

• This includes things like accessibility, but also different OS versions and setups

Biases and hidden assumptions

• Age

  • Younger users tend to have higher IT literacy

  • Older generations may not

• Gender

  • Women often have smaller hands

  • Are clickable areas usable by all on mobile?

• Education

  • Higher levels may make it easier to comprehend language

  • Are we designing for a range of socio-economic backgrounds?

• User journeys

  • Can users recover or interpret steps, or must they start over?

• Income

  • Higher earners can afford more services and tools

  • Is this reflected in what we build?

As testers, we need to call these things out.

• What are the different ways your system can be used?

• What are the socio-technical factors influencing your users?

Looking at things from her point of view

• She can’t see images in her mind - Aphantasia

• For example: imagine a tree

  • Most people see a visual image

  • She doesn’t

  • Instead, she sees metadata: facts and concepts

• This is measured on a five-point scale

  • Ranges from metadata-only to detailed imagery

Developing a test plan

• Her constraint is a kind of superpower

• Aphantasia means she has to ask more questions

• She can’t make as many assumptions as others, which helps in testing

Emily took us through an example of language apps and all the assumptions we might make

Too many to list, but the main takeaways:

• Can you empathise with your average user?

• The people building systems aren’t often the ones using them

• Her inability to picture solutions helps her ask better questions

• Is your system truly accessible?

• Don’t rush to build on your assumptions

---

Testing: a solved problem in engineering? - Callum Akehurst-Ryan