Busting the myths around intelligent enterprises

Getting from experiment to exponential

At a time of COVID-fuelled digital acceleration, deploying AI at scale is the Holy Grail. Everyone knows they need to do it. Few know how to get there.

Last year, Accenture’s global study of 1,500 C-suite respondents found 75% of executives believe they risk going out of business in five years if they don’t scale AI . Yet, the study also recorded 76% of executives admitting their organisations are struggling to do just that.

In our experience, 80-85% of businesses are stuck in the proof of concept (POC) factory, only 15-20% are strategically scaling across point solutions – and a very few, less than 5%, have industrialised AI for growth.

As our survey respondents know, there’s a very big prize here.

The tiny percentage of companies who’ve cracked the intelligent enterprise ‘secret sauce’ are outperforming their competitors dramatically. Even those scaling across point solutions are getting impressive results. For example, an intelligent cashflow forecaster can now predict cashflow six months in advance with 95% accuracy , despite the underlying complexities of global enterprises with ultra-volatile forex fluctuations, complex products and services and a highly varied supplier base. Using traditional methods, even today we see a lot of companies not able to get an accurate read on cashflow for the last month.

Imagine what happens when you create a culture of innovation and AI, with data and analytics democratised across your organisation:

•  30-60% of back office work can disappear as it becomes self-served, digitised or automated
• This relieves organisational talent to unlock trapped value by focusing on more strategic work

As a result, your internal and external customer experience goes through the roof during moments that matter. That’s year one.

By year two, you now have insights on your customer behaviour, in-the-moment interactions, near real-time transparency on how work is flowing through your organisation and specific bottlenecks directly impacting customer value and experience. Now you have the power of data to architect and, most importantly, intervene in-the-moment for future value, outpacing your competitors in your reimagined business.

Why aren’t more organisations doing it?

Despite the extraordinary potential upside, most AI experiments and pilots have low levels of success, with marginal ROI if any. Cutting costs is easy and delighting customers can also be achieved – but doing both is hard and requires an innovative approach.

Typically, we see organisations struggling to scale AI for three reasons:

• AI/automation anxiety – In some parts of organisations, AI is still seen as a threat to jobs. In fact, the opposite is true. Roles that don’t embrace AI are actually under significant threat. According to Australian Institute Centre for Future Work Director, Jim Stanford: 

  <<< Start >>>

  “Contrary to fears that ‘robots are coming’ for Australian jobs, innovation has been declining for years… At least 200,000 Australian jobs are under threat because of the absence of business innovation.” He adds, “If robots and automation were really triggering a loss of jobs for Aussies, productivity would have accelerated – but this isn’t the case; Australia is in fact facing the lowest levels of productivity we’ve ever seen.”

  <<< End >>>

•  Skills shortage – Proven AI skills are in short supply. Without an experienced team that has delivered scaled AI initiatives before, executives are shying away from making big bets, no matter how cool the POC.

• Trapped Data – With most organisations sitting on highly complex underlying tech. landscape, sometimes including pre-historic apps, getting access to the right data at the right time is challenging – which is a key requirement for any successful AI project.

And herein lies the real problem.

Executives are used to IT projects taking several months/years and then being ‘done’. Remember those early ERP implementations, which cost hundreds of millions of dollars and took 2-3 years?

In contrast, AI projects are fast and never done. You can scale AI in 3-6 months. It takes 8 weeks for a POC and 12 weeks for a minimum viable product. After 20 weeks, you’ll have machine learning code running in production.

‘Never done’ sounds bad, but actually it’s where the value comes in. For example, you’ve created an NLP solution for your call centre which now understands human language and helps automate incoming calls. Where else can you deploy it? Can it be tweaked as an HR self-serve plugin for your internal customers? What about your internal IT help desk or procurement? Every time you build an AI solution you add it to your digital asset library that can be used to solve for other use-cases – so you don’t reinvent the wheel each time.

Sometimes, digital assets lead to new revenue streams.

In 2014, Google bought a little-known AI start-up called DeepMind to help cut its energy use. The technology was predicted to cut data consumption by 15%. Given Google’s data centres used 0.01% of world’s electricity in 2011, this was a smart move. But the multi-million-dollar saving was just the start of DeepMind’s value. Now the company is selling its products and services to healthcare providers, creating a new revenue stream for Google.

How to build an intelligent enterprise

Getting to AI at scale is about cultural change, not technology transformation. It requires a new, enterprise-wide approach:

• Start with the right problem. AI is amazing, but it may not be the right solution for your problem. Legend has it that, during the space race, the Americans spent millions developing a ballpoint pen that work without gravity. The Russians simply used pencils. As they say, “If you go in with a hammer, you’ll see a lot of nails.”
• Nail it, then scale it – IF you can get in-year ROI. Think big, but start small. Once your prototype works, figure out the in-year ROI by looking for as many use cases as possible.
• Invest in adoption. Go big on change management. You won’t get your ROI if people aren’t behind the solution and adopting it.
• Reskill users. You have an AI skills shortage. You now have released capacity in your organisation with rules-based work automated and AI helping users as a coach. Encourage those with the right aptitudes to retrain as data scientists or ML enthusiasts. Those with rich business knowledge who also understand AI will find themselves on a soaring career trajectory.
• Keep improving – This is not a ‘one and done’ project. Monitor how machine learning and AI are performing and keep training the algorithms for better intelligence each time.
• Create a digital asset – Start squeezing value out of your investment. Where else can it be used? Can it be monetised? Where is your organisation’s digital asset library?

For Australian organisations deploying AI at scale is an amazing opportunity to take market leadership and accelerate ahead of the pack. With only a handful of organisations doing it well, everyone has a shot of finding the Holy Grail.

<<< Start >>>

<<< End >>>

from AI investments than companies pursuing siloed POCs

[1] Scaling enterprise AI for business value, Accenture. December 2019

[2] Ibid.

[3] Ibid.

This is for information and illustrative purposes only and is not intended to serve as advice of any nature whatsoever. The information contained herein and the references made in this document is in good faith, neither Accenture nor any of its directors, agents or employees give any warranty of accuracy (whether express or implied) nor accepts any liability as a result of reliance upon the information including (but not limited) content advice, statement or opinion contained in this paper. This document also contains certain information available in the public domain, created and maintained by private and public organizations. Accenture does not control or guarantee the accuracy, relevance, timelines or completeness of such information. This document constitutes a view as on the date of publication and is subject to change. Accenture does not warrant or solicit any kind of act or omission based on this document.

Copyright © 2020 Accenture. All rights reserved. Accenture, its logo, and New Applied Now are trademarks of Accenture. This document is produced by consultants at Accenture as general guidance. It is not intended to provide specific advice on your circumstances. If you require advice or further details on any matters referred to, please contact your Accenture representative.