De-risking a £300k project with a Design Sprint

This case study is for a real paying client at thestartupfactory.tech, but the names of the business and the people are changed for anonymity.

ClosedBox is a training and development consultancy, teaching "soft" skills to employees such as leadership, delegation, or critical thinking. A pair of work colleagues, one managed by the other, sought to productise part of their offering as a tech product.

After positive initial conversations, development had begun in earnst, but I quickly found the project struggling with a consistent direction and fact based decision making. What's more, the project was geared up to spend it's entire £300k budget before involving a single end user.

For me, that prompted the question:

HMW gain confidence in and influence our product direction and decision making, before spending the entire budget?
Snapshots from the workshops and user testing sessions I conducted.

Overview

First an overview of the roles I assumed, the skills I used, and the team I assembled for the job.

Roles

  • UX Consultant

Timeline

  • 2 weeks

Skills and techniques

  • Design Sprint

Approach

I arranged and performed a Design Sprint to:

  • Engage key stakeholders
  • Gain everyone’s understanding
  • Make our thinking tangible
  • Perform timeboxed interviews with users

Running Design Sprint 2.0

I used the Sprint 2.0 methodology from AJ&Smart. There were a few reasons I opted for a Design Sprint:

  1. Our target users had up to 14 hour work days; we needed to be efficient.
  2. We needed clarity as quickly as possible, to determine if we were digging a rabbit hole.
  3. Gathering buy-in was easier using a tried and tested methodology.

In addition, I made some some modifications for my use case.

Modification #1: Workshop format

Since the ClosedBox were based over 1 hour away, and the co-founders could only spend one day per week with us, I broke things down into a series of bite-sized workshops.

Modification #2: System modelling

We had different understandings, and different ways of expressing that knowledge.

Everyone seemed to have different understandings. I asked everyone to draw how they understood the problem domain and share it amongst ourselves, before consolidating it into a single view. See ‘drawtoast.com’ for more.

Modification #3: Internal ideation

I made a judgement call to rely on the clients’ strengths and include them during idea selection, whilst educating my team further during ideation.

Modification #4: Note-taking delegation

In hindsight this was the most powerful. I convinced one of the co-founders to take notes during test day, resulting in them gaining a deeper appreciation for our target users.

Seeing is believing.

This was the digitised map of what happens after a course was delivered.
The final storyboard, once we picked which idea we wanted to learn from.
The final prototype, assembled in Figma.

Challenges

I was only able to involve one of the co-founders directly in the user testing phase. Ultimately, they saw the need to pivot while the other co-founder did not.

A debrief is no substitute for experiencing the learning first hand.

Outcomes

After debriefing the client and our internal team we made a pivot of the project, and launched a pilot test at 50% of the budget spend. Previously this had been 90% or higher.

In addition, I found out one of the co-founders left some time afterwards, and a year later started their own version based off the learnings from our Design Sprint together.

At the end of the day, it's about calculated risks, and I wish them both success.