Skip to content

How to Look for AI Use Cases in Your Business


Sep 7, 2020 - 8 minute read

How To Look For AI Use Cases In Your Business Blog 416X300
Michał Zgrzywa AI Director

AI Director at Objectivity, experienced manager, software developer at heart. 

See all Michał's posts

Finops Stickad Blog430x300


Artificial Intelligence (AI) is widely recognised as the technology that will significantly change the way companies work. The research findings published by Gartner, Forrester, and other leading consulting houses, continues to validate the increasing value of data. Our world is changing. Organisations are experiencing tremendous success as a result of new technologies—image and video analysis, natural language processing, recommendation engines able to draw valuable insights from data, etc. However, many companies are still trying to figure out how to get started on their AI journey and how to apply this technology in a manner that will bring value to their organisation.

One of the reasons for this is that the technology is just one aspect of a successful AI solution. The second, and more important one, concerns having a proper business case.

In this article, I describe Objectivity’s proven approach, which could help you discover how to look for AI use cases in your own organisation.

Set the Scene & Fight Against Misconceptions

There’s no way you can find a successful business case without the people responsible for the business side of things, i.e. the business stakeholders. This may seem obvious but it’s really not that uncommon for companies to centralise the responsibility for innovation in Digital Transformation or IT teams. And this is absolutely fine, but in such cases, it’s crucial to include the people who are responsible for the part of the business that’s being analysed.

This can sometimes prove difficult as such people always have their calendars full and lists of priorities longer than the Chinese Wall. Nevertheless, it’s important to include them.

First of all, they understand what the real pain points of their business are. Secondly, they will be able to easily recognise naïve solutions that won’t be able to stand the test of daily operations. And finally, if they’ll have the opportunity to contribute to the solution’s concept, there’s a higher chance that they and their teams will be more willing to adopt it in their work.

However, when the group is extended and involves business stakeholders, there’s a need to make the technology understandable for everyone involved. It’s quite common (and expected) for business stakeholders to not to be up-to-date with what AI technology is offering, and which part of the offer is mature enough to be implemented.

Therefore, it’s crucial to first try to remove some of the commonly held misconceptions about AI.

So, no, AI isn’t about an army of Terminators out to get our lives, or at least our jobs. It’s about assisting humans in making decisions and relieving us from having to perform mundane repetitive tasks, thus allowing us to focus on the more rewarding parts of our jobs. And no, AI isn’t something that will take place in some distant future and only at super high-tech companies. AI is already delivering great value to regular businesses from all types of industries.

As such, now is the time to change people’s existing negative misconceptions (e.g. ones derived from SciFi film representations) into positive connotations.

For instance, the AI adoption trend can be compared to the industrial revolution, where the invention of the steam engine enabled machines to perform simple physical tasks at a low cost. If someone was running a factory back then and was able to imagine another way of performing simple tasks like pushing, moving, or lifting the steam engine, this would have allowed them to innovate and disrupt their market.

The AI revolution is similar in terms of how it could impact businesses and their market strategies.

Today’s technology has given us the possibility to perform simple tasks—ones which were previously only reserved for humans, such as image recognition, understanding the message in a text, spotting a trend or correlation in a series of numbers—in an automated manner.

Nowadays, these types of tasks can be performed at a much lower cost. Companies that can re-imagine their businesses in a way that incorporates taking advantage of the opportunities presented by AI could potentially disrupt their industries and dominate their markets.


Once you have a team in place with whom you can innovate, there are at least three approaches you can take: to analogise, to start with decision-making, or to start with leveraging your data. Our advice? Don’t choose one, apply them all.

The first approach requires a bit of preparation. The idea is to show your team as many examples of cases in which AI has brought value as possible. These examples can be from any business domain, not necessarily only yours. They’re meant to spark the imagination of your team and help them to realise that AI was able to solve similar challenges to the ones your organisation is facing and can be leveraged to do the same for you.

So, start with the research, find as many interesting examples as you can. Group the examples into similar packs (e.g. recommendation engines or automated speech recognition). Describe them to your group and give them time to reflect, discuss, and generate similar ideas that would bring value in their line of business. When we work with our clients in a similar workshop format, we’re always amazed at how many good ideas are generated.

The second approach involves taking a closer look at the decisions being made in your organisation on a day-to-day basis. Ask yourselves what types of decisions are being made every day that impact company results in the most significant manner. Next, have your team list the data sets they have in their business (by data, we also mean films, photographs, graphics, documents, audio streams, etc). Then, let the team imagine and discuss what would be possible if they had access to an unlimited number of team members, who could look through these data sets, search, compare, and prepare answers and suggestions. Again, give them time to reflect and generate ideas. Some of them will go back to the analogies generated during the first approach—this is a good thing.

You may also decide to progress with the third approach whose focus is primarily on the data. To explore this area, ask your team to think about the data sets you have as an asset. Discuss each data source. What information does it possess? What kind of insights could it generate? Would it make sense to look for trends in it? Would it make sense to look for cases that are different from current trends (i.e. anomalies)? Would identifying these cases help to increase quality, decrease costs, or realise unspotted opportunities? Could data itself become a product that other businesses would like to buy? Why? etc. This approach is difficult and explorative but could help you generate a range of innovative ideas.


The approaches described above should generate a lot of ideas, some brilliant and some impossible. The next step is to prioritise the list and choose a few good candidates for the Proof of Concepts (PoCs).

Prioritisation should focus on at least two aspects. The first is business (or social) value. If the idea successfully turns into a reality, how much value will this generate for your company? Think about a realistic scenario. If possible, try to assign real values (e.g. money saved or additional profit). If not, don’t worry, you can simply organise the ideas you have on your list—compare them in a relative way and rank them from the most to the least valuable.

The second aspect is the technology risk and, ideally, you will need to involve someone who has at least high-level technical competencies. Try to assess how feasible the idea is, whether the technology is already mature enough, and how costly it will be to verify the idea using a PoC? Finally, how expensive would the final solution be? Again, you don’t have to generate exact estimates—simply organise your ideas from the easiest to the most complex.

ai use cases matrix with value and technical risk

Having analysed these two dimensions, you’ll be able to clearly see which ideas seem to be the best to start with. Ideally, there will be many high-value, low-risk initiatives, which should be tested with well-defined PoCs. And, once your business sees its first promising results, it will be the time to also try the highest value, high-risk ideas; however, it’s probably not the best idea to start with them.


Successful PoCs may transform into projects that have the potential to change your business. And the unsuccessful ones will teach you valuable lessons. This is why it’s important to iterate during this process. Your team will learn from successes and failures. New ideas will pop up, which were not present initially. Everyone involved will become more experienced in innovation. This is because iterating allows you to include more people, which brings various new perspectives and ideas.

Here is how the iteration scheme could look like:

iteration scheme

To learn from PoCs in the most efficient manner, it is crucial to define them properly. They should be as small as possible, testing only the feasibility of the idea. The most common mistake with AI PoCs is to implement too much too early, which increases the cost and makes it more difficult to abandon unsuccessful ideas quickly and fairly painlessly. What’s more, the verification should be defined in business metrics, not technical or mathematical ones—for easier understanding and learning. Finally, a PoC must have a business owner; it shouldn’t be “something done by the technical people” that business stakeholders aren’t very interested in. Without their support, it could prove difficult to implement the solution, even if the PoC is successful.


I hope you will be able to find many AI use cases in your organisation, which will help you move forward on your journey towards generating business value from AI initiatives. There are plenty of exciting opportunities out there and, hopefully, the described approach will help you to start or structure your current approach towards a higher chance of success.

If you’d like to see what the start of such a journey looks like in practice, read our Leonard Cheshire case study. Leonard Cheshire is an organisation helping people with disabilities around the world to achieve economic independence—we helped them to identify AI value opportunities by organising an “Art of the Possible in AI” workshop. People responsible for operations generated many excellent ideas on how to bring social value through employing AI. One of these ideas— i.e. a recommendation engine that helps people choose their career paths—ended up receiving funding from the fantastic “Microsoft AI for Accessibility” programme.

For more information on how to find data & AI use cases in your business, download our latest complimentary eBook: "How to Build a Data-Driven Organisation".

Finops Stickad Blog430x300
Michał Zgrzywa AI Director

AI Director at Objectivity, experienced manager, software developer at heart. 

See all Michał's posts

Related posts

You might be also interested in


Start your project with Objectivity

CTA Pattern - Contact - Middle

We use necessary cookies for the functionality of our website, as well as optional cookies for analytic, performance and/or marketing purposes. Collecting and reporting information via optional cookies helps us improve our website and reach out to you with information regarding our organisaton or offer. To read more or decline the use of some cookies please see our Cookie Settings.