Clouse listed five steps he deemed essential to succeed when implementing AI, starting with the first step:
1: Assembling a Team:
As with any project, the quality of the team leading it is essential. The quality of the team depends on various factors, according to Clouse. It is essential that the team consists of well-rounded data generalists that have a good knowledge about the topic they are working with.
In addition to this, it is important that there is a good chemistry between the team members, and that they can communicate and cooperate well. Good cooperation and communication leads to great results.
Clouse also mentions the importance of the team being able to communicate to a broad audience in order to spread their message across many platforms.
The next step in the process is:
2: Choosing a Business Domain
When choosing a business domain is it important to avoid ‘premature centralisation’, Clouse said. The reason being is to ensure that the team is better positioned to learn specifics and build area expertise.
This is also helpful in giving the company a better chance at showing value, Clayton said.
Step three in the process is concentrated around finding the right manager with the right skills and mindset.
3: The Right Manager
Clouse said that what he considers the right manager, is someone who has a flexible approach to creativity. He or she needs to be a “champion of data-driven solutions”, and also believing that problems are solvable.
A good manager is someone who is able to think in a “bigger picture”, Clouse said.
In the fourth step, Clouse explains what he believes is essential to set his team out for success.
4: Set your Team for Success
Clouse explains how the key to maintaining the motivation internally in a team, lies with having quick wins. Everyone who has ever worked on a project knows that actually seeing good results is key to motivating people to work harder to maintain the success.
“It is also essential to focus on fostering excitement among your employees”, Clouse said, which links back to his argument about maintaining a good and communicative working environment.
In order to avoid failure, it is important to focus on areas where traditional analytics have failed previously, in order to avoid making these mistakes in the future.
The fifth, and last point that Clouse highlighted during his speech, is teaching.
Providing your team with knowledge is essential, Clouse says, highlighting the importance of promoting how different tools can be applied to provide an insight to the various use areas.
It is also essential to allow time and freedom to design training sessions, implementing one-on-one conversations and mentorship, Clouse said.
“So how do we get there?” Clouse asked at the end of his presentation. “First, we need lots and lots of data, because the more data we get, the more insight we gain”.
Then it is important to get access and integration, and ensuring that this access is across domains, developing intuitive toolsets and set goals for wide-scale integration.