Playing back our research at Salesforce World Tour London

• It was Salesforce World Tour on Thursday and Stephen Wood and I presented findings from our research into consumer attitudes to AI in financial services. The things I personally found most interesting in the research were:
  • There is a huge opportunity to use generative AI to adapt to customers' mental models. According to Mintel, only 14% of UK consumers use budgeting apps or tools in their banking apps. A big reason for this is the existing tools are opinionated and prescriptive. Spreadsheets let people model their money in a way that makes sense to them, but that might leave them with blindspots. This is a gap generative AI can address.
  • Be careful what you automate. Lots of people get their financial confidence from spending time on it, researching decisions, writing out their transactions etc. Speed and ease doesn't equal better.
  • I came away from the research with broader questions about trust and how it is formed. Most people would check an AI's work, but happily follow the guidance of an Instagram finfluencer. So to build trust with an AI powered finance tool, does it have to be a chat bot so you can form a relationship? (This also plays into some questions I have around building trust with remote teams. Something to dig into further...)
• Team Health Check Survey Earlier in the year I subjected a project team to the ~100 question Team Diagnostic survey. They all gallantly completed it but had some strong feedback on its length and how that overshadowed any value we got out of it. There was also feedback that much of the survey felt geared towards "permanent, functional" teams rather than project teams. This week I spent a bit of time creating a revised version. I drew from the Team Diagnostic survey but also the Five Dysfunctions of a Team Diagnostic tool, Amy Edmonson's psychological safety assessment, and filled in some gaps around feedback, decision making, and belonging. It's ~25 questions and the project team I've tested it with this week had no complaints about its length. And so far it's done what it's meant to: pulling out the different experiences that people can have of the same team ahead of discussion in a mid-point check-in. You can take a look here and use it if you like. I would love to hear your feedback.
• This week was the UN's Behavioural Science week. There was an interesting session on bringing together behavioural science and foresight (a reductive summary: use foresight to think through the consequences of behavioural interventions and how the system might evolve; use behavioural science to think through more accurately how people in various roles may respond in future scenarios) and behavioural science and organisational behaviour.
  • I'm probably simplifying massively, but often when I look at things about bringing behavioural science into other areas, success seems to come from running a good design process, and in particular strong ethnographic research, more than the application of any behavioural science framework. This is important because, as came up last week, context is everything. Another good reminder of that from Rory Sutherland here)
  • Quick aside: the UN has identified five skills–the Quintet of Change–that its organisations need to adopt to support their efforts towards the Sustainable Development Goals. They are innovation, data, digital, foresight and behavioural science.
• I was briefly and hopelessly addicted to Vinted this week. It isn't necessarily obvious, but there are some powerful game mechanics at play in there.