Is your organization planning on enhancing your services with Artificial Intelligence (AI)? According to Forbes. Before the pandemic, 80% of enterprises were considering or exploring the opportunity. That is an extremely high number, unless 'considering' also includes just discussing the possibilities of AI on a Friday afternoon.
80% of enterprises
The 80% mentioned in the Forbes article is indeed a bit misleading. This number comes from research executed by Vanson Bourne. They selected 260 IT and business decision-makers VP-level or higher from organizations with a global revenue of more than $50M. Clearly, those are not just any enterprises.
So if you are just starting to have this conversation on a Friday afternoon,you are probably on par with many other (service) organizations.
Considering AI solutions
Considering AI solutions seems a logical first step, but be aware that it could put you in the wrong direction.
Take for instance the following list of possible applications:
- Manufacturing robots
- Smart assistants
- Proactive healthcare management
- Disease mapping
- Automated financial investing
- Virtual travel booking agent
- Social media monitoring
- Inter-team chat tool
- Conversational marketing bot
- Natural Language Processing (NLP) tools
What do the things on this list have in common?
Well, these are all opportunities that describe or confirm what the technology could do. However, this list says nothing about what users actually want or need.
A better conversation starter therefore is, "What problem can AI solve for you and/or your clients?".
Developing a successful new service offering that makes use of AI requires you to start with the need of your clients. If your starting point is the technology, your project is doomed to fail.
AI vs humans
Another debate that often comes up in these "consideration", is the quality of current AI solutions.
Advocates of the technology like to point out what is already possible. Indeed, there are some fantastic examples out there. Such as AI solutions that can create inventions on their own (source)!
Those still on the fence will say that AI solutions are not perfect yet. They argue that driverless cars cause deadly accidents. Indeed, no AI solution is perfect.
However, it matters what you want to compare the AI solution against. As in the case of driverless cars, if you want them to be 100% safe, that is a rather unfair standard, as we humans are by no means perfect drivers either.
Interestingly, both arguments are rather irrelevant, as research has shown that the best of both worlds comes from combining AI with human interaction (source). It is not AI versus humans, but AI with humans that makes the difference (source).
How to make AI work for you?
Below are five things that you may want to address while discussing the adoption of AI.
1. Start with the problem in mind, where is the biggest opportunity?
First, start with a problem in mind. For example, are there currently underserved markets? That is, are their clients out there that really would benefit from your expertise, but cannot afford your hourly fee?
What would it take to serve these new markets with a smart solution and how would that impact your current customer services?
2. What are the implications for your current clients?
Second, what could AI do for your current clients? Could it help them to make better decisions? To sift through data faster? Could it make your services more accessible? Will it perhaps give you better insights into all the ongoing issues and enable you to proactively address things that your clients are willing to pay a lot for? Because you are preventing things from happening instead of reacting when it happens?
And again, what are the implications for your current service offerings?
3. What off the shelf solutions are available?
Third, as a service provider, you probably don't have the scale to support the development of an AI solution in-house. So, what software can you purchase off the shelve, that would help you to get started?
Or are there perhaps startups that you could collaborate with? Who would be willing to develop the solution together with you? As a side note, to be an attractive partner in such collaborations, there must also be an upside for the startup.
4. Are there options to pool data with others?
Fourth, AI is as smart as the quality of the data it gets fed. How much data do you have to feed the beast to make it work for you? If you only handle a few cases a year, will you have sufficient data to support the development of proprietary algorithms? Or can you perhaps pool with others, so that there will be enough data to create a powerful AI application? What are the implications of sharing the data and algorithms you develop with others?
5. The technology is just one component
Fifth, bringing AI into a service organization will take more than making the technology work. You will have to create a new service offering around it, to be able to deliver your clients the value that you envision. Creating a new service offering won't happen overnight or be the result of a conversation at a bar on a Friday evening.
It takes deliberate planning and experimenting to get from idea to a valuable new service offering.
It starts with considering the options
So, when you are considering AI and discussing the opportunity during a happy hour, the things listed above are what you should be talking about.
Having these conversations is important. It is the start of a meaningful discussion about what AI can do for you and your clients. And who knows what can come from that!
However, don't wait too long with acting upon these conversations. As others in and outside your field will be eager to jump on these opportunities if they are promising. The last thing you want is that while you are still discussing what could be possible, others are already stealing away your clients with superior AI-driven new service offerings.