Delivering the DATA Economy

AI: Expectations Versus Reality

Posted by Hans Godfrey on 8/13/19 3:53 PM


The bar for expected returns from AI projects is often high. Firms of all sizes have attempted to deploy AI-based solutions with the expectation that AI might ameliorate a host of business challenges. However, do AI projects live up to these hopes or is the reality a little more grounded?

As part of London Tech Week in June, Agorai conducted a survey of AI Summit attendees. Our goal was to gain insights regarding business expectations of AI and what benefits firms had actually received. Based on the responses, three core themes surfaced: 1.) Is AI delivering on business needs? 2.) Are AI projects being deployed as quickly as possible? 3.) Are companies of all sizes able to easily access AI technologies?

Is AI delivering on business needs?

The majority of respondents said that AI projects should, above all, automate processes and reduce business costs (63% and 51%, respectively). In reality, only 37% said processes were actually automated and only 26% said expenditures were reduced.

Considering the emphasis that most companies placed on automation, that 37% tally is surprising. There are several things we might be able to conclude from this gap in expectations:

  1. How companies deploy AI may not align with business expectations. This situation often occurs when businesses experiment with AI solutions, but don’t necessarily tie proofs of concept (“POCs”) to tangible business metrics. Before beginning any AI project, it’s important to ensure that the business use case is clear and that the expected key performance indicators are measurable and agreed. In order to identify those use cases, you should ask yourself where the biggest gaps exist in your business, where a small change would produce easily measurable results. You should plan deployments where you can start small and then expand outward from there.
  2. Business expectations of what AI can achieve are too high. Often, the expectations for AI projects exceed the capabilities of the technology applied. Typically, companies have multiple objectives they want to achieve from a given technology and this typically increases implementation and deployment timelines. The key to a successful rollout is to set clear objectives and metrics, start small, measure consistently, make adjustments and then scale the solution overtime. Businesses typically achieve success by picking one or two use cases to start then applying lessons learned to a larger rollout.
  3. Results may take longer than expected to materialize and projects are therefore abandoned. Finally, a third of participants responded that the last AI project their company executed was never completed. This could be the result of multiple issues including a gap in the capabilities required to deliver AI-based solutions. More often, we find that project failure directly correlates with points #1 and #2 above. AI solutions are not magic bullets. Businesses must set realistic expectations for what they want to achieve and implement POCs narrowly to accelerate deployment timeframes and improve chances of success. POCs can then be scaled to achieve increased value for the enterprise.

Are AI projects being deployed as quickly as possible? 

As previously stated, the failure rate of projects was surprisingly high, with a third of the respondents saying that their company’s AI project never finished. Despite this failure rate, respondents largely believed AI projects should be delivered quickly, with 64% of participants saying projects should be delivered in less than 12 months. This opinion seems to be supported in that just over half of those surveyed said that their firm’s AI project was finished in under a year (with some tangible business benefit).

These results align with Agorai’s delivery philosophy for AI-enabled solutions. Given the lack of business-awareness of AI tools, delivering results quickly is imperative to success. To do that, companies need to look holistically at solving problems – first inward  and then looking outward. Do you understand the potential barriers to success? What are the daily problems being confronted by both employees and customers and what kind of solution would make a difference? Have you sized your use cases to the smallest increment that still delivers tangible business value? How will you measure progress?

For additional insights into how to build a plan to successfully implement AI in your organization, check out our latest white paper, Six Steps to Deploy Artificial Intelligence Successfully in your Organization.

Are companies of all sizes able to easily access AI technologies?

Nearly half of respondents said that AI is either not so easy or difficult to access. From our perspective, this result is not surprising at all. There’s a huge gap between large enterprises and small to medium-sized businesses in capability for identifying, assessing, and deploying next generation technologies.  Generally, smaller companies lack the scale to make the necessary investments in technology, data, or in-house expertise. They may recognize the promise of AI enablement, but they don’t know what the most effective use case might be, what solutions to use, or how to deploy those solutions. Hiring in-house expertise or traditional consultants that could close knowledge gap is expensive.

Agorai addresses this gap by fusing the best of consulting with product. Providing insights in a cost-effective way, we guide companies through the process of leveraging AI solutions to realize meaningful benefit quickly. We support business domain knowledge of several industries, as well as deep understanding of the AI landscape. Our approach is purpose-built to work within existing business strategies and transformation initiatives to quickly identify, pilot and scale solutions to achieve business value.

At Agorai, we are making AI understandable, accessible, affordable and profitable so that companies of all sizes, can join the AI economy. 


Topics: Insider

Recent Posts