Like any shiny new technology, AI had some hype associated with it. Over-promises on the part of vendors combined with unrealistic expectations on the part of customers have created the atmosphere that currently hovers around AI: one that is unmistakably tinged with disappointment.
The best way forward is simply to inject more realism into the proceedings. We’ve already gone through the ‘AI will solve everything’ claims by vendors or overeager customers, as well as the executive-level ‘we need some AI, and we need it now’ fevers.
That’s actually a good thing: Now that the fever has broken, we can hit the reset button and focus AI efforts on specific real-world use cases where it can deliver value. These range from newer areas like contract intelligence – better understanding the risks and opportunities that lie buried within mountains of contracts – to longstanding business problems like knowledge management that can now be tackled more efficiently and with fewer people. More on these in a bit – but first, an examination of the barriers that can prevent organizations from having success with this technology within their organization.
Getting off on the right foot
Part of taking a more realistic approach to AI involves understanding that AI is best approached as a long game. These are long-term projects that require the things that all good projects require: change management, people management, and expectation management. It’s now becoming clear – in perhaps a way that it wasn’t before – that a ‘back to basics’ approach that encompasses people, processes, and technology is what makes for a successful project.
Additionally, you need to take the time to understand what problem you’re actually trying to solve – and then realize that the AI aspect is an enabler in that project, rather than the core of the project. The goal is to find opportunity and then capitalize on the opportunity – and AI is simply part of that wider undertaking.
A close understanding of the problem you're trying to solve helps put a ‘number’ on it to better understand how valuable that problem is. How much does the business problem currently cost the organization? How much could it save the organization if it were solved? Knowing that number will help organizations determine whether it's worth deploying AI against the problem, or whether the problem is better handled by a combination of AI plus other approaches.
For example, people often get obsessed with the idea of using big data to tackle a business problem. In reality, with things like contract intelligence, you only need a small set of data that’s highly repeatable in order to get started. When you get into inconsistent data and high variability, it becomes a bit more involved. That’s why you have to know the ROI of the business problem you’re tackling – in that circumstance, the opportunity has to be big to make particular AI investments worthwhile.
Recession creates necessity
In the age of Covid-19, when the world is on the brink of a global recession, many business problems that might not have seemed worthwhile before are taking on new importance.
Let’s revisit our earlier use case of contract intelligence. Organizations are looking at their contract portfolios and searching for opportunities that might be hiding in there. Is there money that is currently being left on the table? Maybe there’s some fee that you’re contractually allowed to charge that you wouldn't have enforced when the economy was full steam ahead, that you would during leaner times.
Likewise, knowledge management, on the face of it, might not seem like the ‘sexiest’ project. However, that can quickly change in an uncertain economic environment that is only going to become more competitive as customers carefully scrutinize every expense and determine who they’re going to spend money with. In this environment, insight into your own data, into your people, and into your clients and the various projects or matters you’ve tackled for them takes on outsize importance. Having a knowledge bank for your organization to tap into is a huge competitive advantage – and if you don’t have that knowledge identified, you’ll miss opportunities.
In both of these scenarios, AI enables these business problems to be tackled without requiring armies of bodies. With contract intelligence, AI – powered by human-trained models – can efficiently review heaps of contracts and understand what clauses and provisions they contain – which, in turn, informs what risks or opportunities lie buried within them. Similarly, AI can automate much of the curation aspect of knowledge management, automatically surfacing expertise and best practice documents.
Projects like this are where ‘the rubber hits the road.’ It’s where AI can realistically be deployed to deliver actual results. Measuring the success of an AI implementation in one of these areas should similarly be tethered to reality. For instance, if an organization is expecting to fix 98 percent of the revenue leakage in its contracts with AI, then a 10 percent return might seem rather measly. However, what if that 10 percent figure translates into millions of pounds a year? Suddenly, 10 percent is actually a very respectable return. That’s one of the advantages of having a realistic approach, rather than one freighted with inflated and unrealistic expectations.
Productizing and democratizing
For their part, vendors are also embracing realism by increasingly providing easy-to-use AI products for their end users. After all, is it realistic to expect your customer to know how to code just to take advantage of AI? No. They have business problems they want to tackle. They don’t need to know how to code; they don’t even necessarily need to know how the product works. They’re just interested in the outcomes that AI can deliver, and vendors are responding in kind.
This approachable user experience can be considered part of a larger trend of democratization around AI. Once the province solely of the larger, multi-region organizations that could afford it, today AI is accessible to organizations of all sizes, including the smaller ones that might not have been able to make a business case for investment in this capability just a few short years ago.
Now that people have accumulated some hands-on experience with AI products, they’re better understanding the reality around AI. They’re seeing for themselves the areas where it can deliver value, as well as the processes and change management that needs to accompany it to help ensure its success. There’s an understanding now that AI isn’t magic, and that it requires some effort.
Ultimately, this is a positive development. By getting more experience with AI, and approaching it more realistically, organizations are able to sweep aside the hype and approach their AI projects in a new way – allowing them to effectively derive the benefits that this technology can bring to the complex business challenges they face.
Alex Smith, Global Product Management Lead, iManage RAVN