Thread by @RohanIrvine

Back from the break, Stephen Cox from Westpac is talking 'The good, the bad, and the empathy' a journey of data driven personas.
#designresearch

The good: How our team has been able to connect the business to the customer.
The bad: How I stuffed up the research methodology and cost the company time and money
The empathy: how we've managed to create it.
#designresearch

There are 4 researchers to 50+ designers at Westpac who deal with generative and evaluative research.
#designresearch

Their generative research tips into the strategic design team which eventually trickles into the day to day design work.
#designresearch

The research teams vision is to: Brings the business closer to our customers.
#designresearch

You would think it would be easy to align that vision with Westpac. People buy into the fact they need to understand the customer, but they don't necessarily buy into understanding the customer at an individual level.
#designresearch

At the bank we believe in numbers. They're characterised as Truth.
The bank has a positivism approach to the world. If we can put numbers to people we can prove out what the truth is.
#designresearch

Stephen sits at the other end as an interpretism.
#designresearch

He asks, how can we connect our business to the truth of the customer. How can we use those data to do this?
#designresearch

They started with Personas—but they've fallen out of favour. People have stopped coming to them about behaviours and intents and requesting personas based on demographics and segments. In the last 2 years, people have stopped asking for personas.
#designresearch

When people stop asking for the people they're designing for—they've begun designing for themselves.
#designresearch

Stephen started a project called Data Linked Personas.
#designresearch

At the end of the financial year they secured some unspent money to build these data driven personas.
#designresearch

They spent a month talking to stakeholders around the business.
"what do you think the attributes are that make a difference to how people do online banking"
#designresearch

We came up with a list of 120 different things that may affect the way you bank online.
They hired @u1group to do an attitudinal survey that makes people bank differently.
#designresearch

These are the 3 parts they came up with.
#designresearch

They came up with 6 personas that could be linked to their internal data.
#designresearch

They had some people who didn't fit and they just thought they weren't being honest so they cut them out.

When the put the archetypes into the data they started finding that no-one really fit into them.
#designresearch

So Stephen got a data scientist to run models on people's behaviour and came up with 6 behaviour groups.

Unfortunately, completely different to their personas.

He had a view that people's attitudes correlated with their behaviours.
#designresearch

#designresearch

He went back to the drawing board after scrapping the project, but decided to look at the behavioural and attitudinal data. He found there was no link.
#designresearch

Online banking is a high function experience.

People use digital banking to get their banking done.
#designresearch

Pivoting: I've taken some money, and we haven't seen the results we wanted.

#designresearch

Outcome driven innovation.
#designresearch

People are buying your service for an outside. Not for the service itself.
#designresearch

The outcomes are ultimately measurable. They connect back to numbers.
#designresearch

What is the most useful outcome you want to focus on?

#designresearch

If it's really important, and the satisfaction is low, you can do work on it.

It helps us link numbers to customer needs.

#designresearch

How does it apply to personas?

We had to distilled our exisiting research and conduct more interviews to understand what jobs exist in internet banking.

#designresearch

We tagged all our transcripts with the jobs that people were doing.
#designresearch

They found 5 jobs and 44 subjobs that people are doing.
#designresearch

Core jobs + Stories = CoreStories
#designresearch

They're linked to our data, and we can answer questions about what our customers are trying to do.
#designresearch

We broke a generic persona, we have real stories that are unique.

The jobs can be common across these people but the people can be different.

#designresearch

Stephens team built a website that links the transcripts to the job so you can read the verbatim conversation.
#designresearch

It frees us from the tyranny of one persona for one task for one project.

Our stories can be as diverse as our customers.

We can be more inclusive.

#designresearch

Our customers can speak how they want to speak.

#designresearch

We can get people reading these stories and the context of their stories.

#designresearch

#designresearch

What has Stephen learned?

#designresearch

Now with the level of humanness that's in the data, there an opportunity to actually turn this data into personas, but should I?
#designresearch

Q from the audience:
With the database how much investment in time does it take to maintain?
#designresearch

A: Not much, it took several months to create the jobs.
#designresearch

Q: What's changed since having this information?
A: Its only been live for a small amount of time but feedback has been 'This is excellent, it's made the numbers mean something to me'
#designresearch

Thanks @S_Cox!
What was the software you used to tag the transcripts: asking for @uxresearchguy
#designresearch