The ITSM.tools 2021 IT service management (ITSM) hot topics and trends poll identified artificial intelligence (AI) as a key area of focus this year. To help, as one of the many articles we have planned, we asked ten ITSM industry authorities for their view on the guidance and help organizations need for succeeding with the adoption of AI-enabled capabilities. These tips for AI success are presented below in author-based alphabetical order.@SophieDanby asked ten #ITSM industry authorities for their view on the guidance and help organizations need for succeeding with the adoption of AI-enabled capabilities. This is what they said. #AI #ArtificialIntelligence Click To Tweet
Ian Aitchison, Independent
Firstly, do plan to adopt AI-enabled capabilities.
Organizations that don’t start to engage with new concepts will fall behind in skills and competitiveness. These tools and features are not yet very mature. Expect maybe 50% failure where the value doesn’t fit your organizational need, but with that will come 50% that will. So do get involved.
So what sort of AI capabilities are there?
Broadly speaking, the AI landscape can be broken into solutions and features that:
- Analyze data and make recommendations. This tends to be the easiest to implement.
- Take action, either on request or independently. Often this is closely connected with automation tools
- Communicate. Here the natural language/conversational/AI pieces come to play.
Of course, sometimes you can also combine all three of the above.
Next, where are you getting those AI capabilities from?
Consider Built-In AI vs. Add-On AI vs. DIY AI.
Built-in AI is the easiest to adopt. However, remember that an expert vendor in, say, service management, may not be an expert vendor in AI and ML. There is a higher chance that their AI features are light or lower-value.
Then there are add-on AI vendors that – for example – may provide a chatbot that plugs into your automation or service management tools. These add-on AI tools are more expert and are more likely to carry greater feature power and value. But they’ll almost certainly cost you considerably more.
Third, there is the DIY AI build-it-yourself model. The value here is that your organization can learn how to use machine learning models and really take unique benefit, but this is a very significant investment and carries very heavy overheads.
Finally, implement in small steps for AI success
Identify the best use cases and target these only. Phase introduction. Be aware of the culture of your organization.
Go point by point, use case by use case, and control target audiences until you’re sure it’ll work for you and your culture.
But don’t hold back. This innovation tide is already pulling ahead of many of us, and the water is rising fast. Better get in there and start swimming.
For further insights from Ian on AI success, please read his full article on how to succeed with AI here.When it comes to #AI, you should implement it in small steps. Identify the best use cases and target those only. Phase introduction. Be aware of the culture of your organization, says @IanAitchison. #ArtificialIntelligence Click To Tweet
Alan Berkson, Global Director of Analyst Relations, Freshworks
As with most projects, starting with AI-enabled capabilities requires a good foundation. For AI, that foundation is data — training data. Successful AI deployments require good training data to power supervised and unsupervised learning. Ironically, the best implementations will apply human judgment to the data used in the machine learning models.Successful AI deployments require good training data to power supervised and unsupervised learning – @berkson0 #AI #ArtificialIntelligence Click To Tweet
When it comes to prioritizing AI-enabled capabilities, organizations need to be super-targeted on the highest value that can be delivered quickly with the existing data quality and data types the organization has available. The learning curve is steep and the journey can be very expensive, but certainly worthwhile. Starting this journey in 2021 will also make it possible to learn a lot quicker and to become more graceful and capable over the next few years, which may provide a competitive advantage.
As with any significant investment of resources (financial, people, technical, and information), there needs to be a clear business case and it needs to include how AI is expected to impact people, activities, and roles in the organization. For example, will the AI-enabled capabilities replace existing human roles?
Organizations adopting AI cannot be successful without a set of core digital technologies. AI is also reliant on sufficient, relevant, clean, and well-managed data, so solid data governance and management is a prerequisite. Organizations should consider starting their journey with defining the data strategy and thinking about how the Data Catalog can facilitate the achievement of their AI objectives.
AI is transforming how we process and experience information, and it means that a lot more information is collected about us, which can then be used to provide customized products, services, and experiences. However, organizations need to consider the different customer expectations as well as legal and regulatory obligations for different parts of the world and different business contexts related to data capture, retention, use, and privacy.AI is reliant on sufficient, relevant, clean, and well-managed data, so solid data governance and management is a prerequisite – @Veridity #AI #ArtificialIntelligence Click To Tweet
The following is an extract from a longer AI-focused ITSM.tools article by John:
Stay focused on value
One of the challenges with any technology is that we get enamored with the technology and everything it can do and forget the basics of any project that releases a new service: what is the purpose? How will we know if we succeeded? How will it impact (increase value) for our stakeholders?
Understanding what is needed to support and improve AI is also critical to any initiative’s ongoing success. How will the use of AI technology align with your organization’s values, goals, objectives? Just because you can do something with AI doesn’t mean that you should.
The many opportunities of AI for ITSM organizations
Today, there are three significant opportunities with AI for ITSM organizations to plan for. The first is in providing a more interactive response, e.g., chatbots or virtual assistance, referred to as Level or Tier 0.5. The second area is using AI technology in assisting support representatives with more complex issues. The third area is the use of machine learning in analytics to identify correlations that humans cannot do cost-effectively.
Don’t overlook AI and 5G
One technology that will facilitate AI adoption and accelerate the value of these AI-driven services is 5G. Not only has computing power been a factor in the usability and acceptability of AI-driven systems, but network speed has been a factor too. The use of machine learning requires a significant amount of data, which often needs to be real-time, which can only be met with 5G speeds.
Address people and AI factors
One unanswered question is, should the users know that they interact with AI technology, not a person? Ultimately, the user is looking for the easiest path for a resolution, but should they understand how their data will be reused? Should they know that they’re not interacting with a person? How do you ensure no dead ends or endless loops which will quickly frustrate users, just as a 404 or “page not found” error does today for self-service?
For further insights from John on AI success, please read the full article How AI Will Impact ITSM in the Next Few Years.There are 3 significant opportunities with #AI for #ITSM orgs to plan for says @ITSMNinja. Virtual assistance, supporting complex issues, and machine learning in analytics. Click To Tweet
It’s true that AI elevates our service delivery, but if organizations want to succeed with AI adoption, first they must figure out who they are and their goals. So, I’d advise any company heading down this path to perfect their BI (business intelligence) before AI. Why? Well, AI is only as smart as the data we feed it. This means if we don’t have comprehensive data to teach AI how we work, it’ll never truly succeed. We gather this data through BI: a method of collating data from services and transforming it into insights and reports to help us better understand where we’re at. The key thing to remember is that if we want AI to help us, we must help it to get up and running first.I’d advise any company heading down the #AI path to perfect their BI first. Why? Well, AI is only as smart as the data we feed it – Sumit De, @TopDesk #ArtificialIntelligence Click To Tweet
One pitfall companies often encounter in the process of starting new AI initiatives is that the excitement around AI might lead to AI being viewed as a goal in and of itself. But executives should be cautious about developing a strategy specifically for AI, and instead focus on the role AI can play in supporting the broader strategy of the company. In short, view AI as a tool and find AI applications that are particularly well-matched with business strategy.
A related comment is that I wouldn’t recommend that companies pool all their AI resources into a single, large, moonshot project when they’re first getting started. Rather, I advocate taking a portfolio approach to AI projects that includes both quick wins and long-term projects. This approach will allow companies to gain experience with AI and build consensus internally, which can then support the success of larger, more strategic, and transformative projects later down the line.
A second area is reskilling. The skills needed for AI projects are unlikely to exist in sufficient numbers in most companies, making reskilling particularly important. Focusing on growing the talent base is crucial given that most engineers in a company would have been trained in computer science before the recent interest in machine learning. In addition to developing engineering talent, an equally important area is that of consuming AI technologies. Managers, in particular, need to have the skills to consult AI tools and act on recommendations or insights from these tools.Remember AI is not the goal. View AI as a tool and find applications that are particularly well-matched with business strategy – @KHosanagar #AI #ArtificialIntelligence Click To Tweet
Organizations can increase the odds of AI success thinking through the following interrelated dimensions of business, data, people, process, and technology.
- Business: Pick the right set of use cases based on the alignment between business needs, AI technology maturity, and organizational capabilities.
- Data: Having the right data is at the heart of AI success. Most AI projects fail because there is no data available or the available data is of low-quality.
- People: AI talent, particularly expertise in real-world at-scale AI deployments, is in short supply. Organizations need to build both an in-house AI Centre of Excellence and also partner with external specialists.
- Process: AI projects are different from IT projects and you need to accordingly adapt your process, methodology, and governance frameworks. Don’t forget to analyze and mitigate any unintentional harm caused by the AI solution.
- Technology: There’s no best AI platform but there’s the right AI platform for your organization’s needs. And don’t forget that you’ll need MLOps tools/capabilities too.
Forget the AI, and focus on value – if organizations start designing solutions with AI in mind, then they’re focused on the wrong thing. Instead, organizations need to simply focus on finding the best solution to business problems and if AI happens to be that, then so be it.
The best AI is invisible – organizations adopting AI-enabled capabilities needn’t be explicit about it. Their focus should only be on solving the problem. For example, Gmail users would remember the autocomplete capability. As you type, Gmail recommends the next few words and it’s pretty useful. But Gmail doesn’t push that solution as an “AI-enabled capability” by branding it with a bot. Instead, it slips it into the solution and the unsuspecting end user would never know it’s AI.
Market the solution, not the AI – your end-users don’t care if it’s AI-enabled or machine-learning-driven. All they care about is a solution to their problem. When you communicate to your end-users, be sure to focus on the solution and how it helps them rather than how you’ve solved it. Organizations should definitely talk about their technology stack but not to the end users.Market the solution, not the #AI, says @yenceesanjeev. End users don’t care if something is AI-enabled or machine-learning-driven. They only care that there’s a solution to their problem. #ArtificialIntelligence Click To Tweet
For executives looking to embrace AI, the first thing to do is make sure AI gets the resources and mindshare that other strategic initiatives have. And when we talk about resources, it’s not the technology – it’s the work that goes in beforehand, for example, making sure business data is the correct quantity and quality. And perhaps more importantly, building a team that can scale-up sustainable AI initiatives and AI success – with the training to spot significant issues like bias and misleading outputs. AI will have a massive impact on the modern world, and it’s down to leadership teams to make sure they’re approaching it with the resources and focus it needs to get it right.For execs looking to embrace #AI, the first thing to do is make sure AI gets the resources and mindshare that other strategic initiatives have. And when we talk resources, it's not the technology, says @Originollie… Click To Tweet
Artificial Intelligence is alluring. It looks like magic and can be Pandora’s box.
I suggest you go beyond treating AI as one more technology you can just immediately plug into your organization (it’s albeit more like a varied set of interrelated technologies). Don’t forget you also need to onboard the right people, information, and partners. Start with understanding AI, then find a realistic project for and with your organization.
First, educate your organization on what AI is and can do, so you know which kind of problems are better solved by the different AI approaches available (like machine learning, deep learning, natural language processing, etc.), and what it entails in terms of internal and external resources and capabilities. For instance, machine learning demands high-quality data, which is not always readily available within the organization. Also, remember the involved data may have constraints (think privacy), and your sector regulations can have auditing requirements limiting the use of AI.
A few years ago I did the excellent and free “Introduction to AI” course by Reaktor and the University of Helsinki. They now have two free courses you may find of interest: https://www.elementsofai.com/
Second, check why you want AI: Is it for solving a problem? To support an objective a line of business wants to address? Start small, with a concrete aspect on a current product/service which can benefit from AI: Can I automate a business process, gain further insight through advanced data analysis, and/or support engagement with employees and customers thus freeing up humans for a more meaningful experience? For these potential use cases, it may well be that initially they can actually be tackled without upfront high investment on advanced AI or machine learning, rather with relatively simpler (and cheaper) chatbots or Robotic Process Automation (RPA).
Finally, as internalized capabilities related to AI get better, then more ambitious initiatives – which may need investment in cloud and big data infrastructure, and adopting new partners expert in AI – can be tackled with the right balance of internal business knowledge and external AI experience.
With power comes responsibility. Hence, before investing in training your own people, developing or acquiring software or a cloud service, and double-checking what data will be processed and why… Start by understanding AI and thoughtfully picking a manageable real-world use case to solve that benefits your organization and sets your IT and innovation people up for AI success.Artificial Intelligence is alluring. It looks like magic, but it can be Pandora's box. You need to go beyond treating AI as one more technology you can just immediately plug into your organization – @rumagoso #AI #ArtificialIntelligence Click To Tweet
Hopefully, if you’ve read all of the above, you’ll see that there are many factors related to AI adoption that need to be considered for AI success. If you’ve already succeeded with AI projects in ITSM and ITSM tool use cases, then I’d love to hear the advice you’d give to others. Please let me, and them, know in the comments below.
Sophie is a freelance ITSM marketing consultant, helping ITSM solution vendors to develop and implement effective marketing strategies.
She covers both traditional areas of marketing (such as advertising, trade shows, and events) and digital marketing (such as video, social media, and email marketing). She is also a trained editor. Y