8 Steps to Make AI Profit Small Businesses & Beyond

Share on facebook
Facebook
Share on google
Google+
Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on reddit
Reddit

With one new AI related company launched every seven days, the UK is a hotbed of Artificial Intelligence innovation with a market that is forecast to grow significantly in the coming years, to roughly £650 billion by 2025

AI startups are ubiquitous across the world. However, like any gold rush, only a few will find gold and many others will be disappointed. How can businesses be more successful with AI? They need to be able to talk business, rather than stick an AI label on a weak business proposition and expect success. 

Jen Stirrup of Data Relish has been presenting on this topic at a number of events, and you can find her presentations over at her SlideShare account.

How can you get started with AI? Make it boring.

Data rich industries can often be the boring industries, such as finance or insurance. AI relies on data, and these are the areas where AI can make a real difference. 

According to Angel List, the following startups are being successful:

On October 16, Quantemplate raised a $12 million Series B. They are a London-based startup that uses machine learning to help insurers process data. This round comes on the heels of several others at insurance-focused machine learning startups:
  • Ethos, a data-driven life insurance issuer, raised a $60 million Series C in late August.
  • Clearcover, a platform that uses AI to sell auto insurance, raised $43 million in January.
  • Flyreel, an AI product for underwriters, raised nearly $4 million from Google in April.
In 2019, the value of insurance premiums underwritten through machine learning is estimated to be $1.3 billion. By 2024, that number is projected to be $20 billion. 

Prove your success with small steps

It’s good to start small, and use pre-baked technology.

Scope does not creep. We should call it Scope Gallop instead because scope can run wild. With AI projects in particular, scope runs out of the building, slams the door behind it, and runs down the street cackling. Manage expectations and start small to help you to be successful with AI.

The 8 'C's of Artificial Intelligence

When you brainstorm your AI Projects, consider the following 8 ‘C’s of Artificial Intelligence:

Brainstorming these items will help you to focus on important areas, but be mindful of scope. You can see the first part of one of Jen Stirrup’s presentations on the topic here.

AI Maturity Matrix

Readiness assessments don’t really tell organizations much, other than they are not ready to get to their ambitions with AI.

At Data Relish, we focus on your maturity, and we try to help you to get to a good place, fast. You need your mission statement and objective; thinking about the 8 ‘C’s will help, but it has to tie in with the company mission statement. It has to say something about you, and what you are going to do.

Here is a sample AI Maturity Matrix to help you on your way. Based on the 8 ‘C’s mentioned previously, you can start to map our your maturity matrix in AI. Then, you can dial down to the detail of how you are going to get there.

 

Credit: Jen Stirrup, Data Relish. Contact us to request permission.

It can help to visualize where you are in your AI journey, and this gives you steps to improve upon. Here is a sample AI Heat Map using Power BI and Excel, so it gives you a sample:

Credit: Jen Stirrup, Data Relish. Contact us to request permission.

There is more to it than that, obviously. Contact us at Data Relish to book some time, and we can help.

Ethics isn't a sales tool. It should be part of who you are.

Talking about ethics and AI seems to be optional, but a good moral compass needs to be at the heart of any AI conversation.

There can be a callousness at the heart of conversations on AI, as people see it as a sales tool that will help them to sell solutions or a service in some way. 

If an organization is talking about ethics, but the organization, or its leaders, are not behaving well, then this undoes any ethical efforts. It seems a simple thing, but people lose trust very quickly and you don’t get it back again. This is particularly important when we look at data; if you are not transparent and honest about the intentions of gathering data, then people assume the worst and they will not trust you. Transparency and clarity is absolutely key.

If you’re not sure what that looks like, try Wheaton’s Law

Next Steps

Get in touch with us and we can help you as your trusted advisors on your AI journey.

Keep Exploring

2 Responses

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.