The innovation path of Vaudit, a new video evaluation tool being developed by Fuse and The MISSION, has taken a few twists and turns since we first blogged about the project.

Following a slightly winding route is to be expected when launching a new concept. A team planning any new product has to keep an open mind and frequently question their assumptions. Significant amounts of time and money are almost always at stake and it is important not to take commitment to building new products lightly.

One of the main challenges faced in designing Vaudit has been the translation of a very human process involving a degree of judgement and opinion, into something numerical, scalable and digital. For the product concept to work, an algorithm needs data to act upon, which means collecting consistent information. The question is, which are the right data points and which are red herrings?

A close working team of MISSION individuals from Fuse, Ethology, krow and Bray Leino –  ably supported by our innovation partners at Subsector –  initially focussed on which problems the product needs to solve. The next step was to identify what features and functionality are needed to best serve those problems. Now the challenge is to turn that wish-list into a cohesive digital product, which is not without its own complexity.

Removing a road block

Identifying the correct data points initially had the team scratching their heads; getting to a solution in an efficient way felt like a considerable barrier. Working out how to make sense of the right data sources and the best way to use them was complex and potentially very time consuming. Fortunately, within MISSION we have some fantastic brains to make this conundrum somewhat simpler! One of the roles of Fuse is help bring the right people together for projects like Vaudit.

For any innovation process it is vital to have a cross-functional team with the right skills involved from the very start. This allows for sharing of knowledge and cognitive diversity which helps to solve problems as efficiently as possible.

Drawing on Group expertise, we have been fortunate to involve Brad Stacey (Technical Director – Bray Leino) who has an excellent working knowledge of the capabilities and design of learning algorithms.

During an exploratory workshop in the salubrious surroundings of krow’s Goswell Road canteen,  Brad used lots of scribbles and pointing to explain how a background data study could solve the problem faster, and in a more methodical, evidence-driven way.

Testing Testing

Most product design projects follow a different path than initially expected, as new information and evidence is uncovered along the way. It’s important to seek out and embrace this insight and feed it back into your plan. As we write, the Vaudit concept is due to go into testing with prospective customers, using a basic prototype and a set of simple questions.

When testing a product it is vital to avoid collecting opinion and to instead try and gather evidence that the concept will serve critical needs. Asking ‘do you think this is this a good idea?’ should be avoided at all costs, to reduce the number of false positives.

In the case of Vaudit, mapping out every feature would be extremely complex and long-winded, so our test will consist of a simple visual of the dashboard and a set of descriptive pages. A companion survey will ask people to assess how well the product serves the customer needs identified using the N2D Method in stage one of the design process.

For the test, we are employing the ‘Service Scorecard’, the second tool in the N2D Method. This tool is designed to give a team an overview of the strengths of a particular approach, as well as identifying weaknesses and opportunities.

The results from this test will help us to map out the next phase of the project and then proceed with confidence. We’ll write again with the outcomes and let you know what path we take next!