AI is talked about all the time and there are many different ways to incorporate it into your organisation. But does it always provide what it promises? Paul Ducie, Partner at business transformation specialists, Oliver Wight EAME, argues that it has been oversold, underdelivered and in many businesses, it is simply leading to employee burnout and worse customer service.
So far, in many businesses, AI has been oversold, underdelivered and often is adding to employee burnout and worsening customer service. Many early adopters of AI need to simply turn it off, while any management team currently considering business cases for investment need larger buckets of scepticism and significantly better plans for its effective implementation to deliver the benefits they hope for.
Major decisions on implementing AI are made at the top by the senior team and are too often based on over-optimistic, unsubstantiated business cases that wishfully promise greater productivity at significantly lower cost.
AI has the power to transform businesses and provide market-leading capabilities, but it also has the power to really mess things up, and that is the path too many businesses are blundering down today.
For those who were around in the noughties, AI is history repeating itself as there are strong parallels to implementations of ERP systems circa 20 years ago (oversold benefits, lack of relevant education and problems from automating poor processes). But this time the pressure is even greater due to the higher costs of AI solutions which promise to deliver even higher productivity gains, but within unrealistic timeframes.
AI embedded in HR systems are a particular example – many places need to simply turn them off! Inherent biases in existing working and hiring practices can be magnified using AI-driven systems, already causing numerous problems that have been successfully challenged in court for discriminatory hiring and staff promotion practices. Also, are we in danger of bringing back the 10% promoted and 10% cut approach, all based on AI-driven quantitative not qualitative data infused with bias?
Other examples include AI demand-planning software. Putting aside that these have the inherent shortcomings of simply being a more sophisticated version of driving the business using a rearview mirror, they spit out a number without any insight on the algorithms and embedded logic that led to the predictions (critical if you are to understand why something is being recommended).
As for AI’s dehumanising impact on customer service, hands-up anyone who looks forward to first navigating the AI chatBot before they can deal with a human. Enough said! A number of household names which invested in deploying AI in real-world customer service roles, such as taking orders at drive-thrus, are already scaling back based on unsuccessful trials revealing the inevitable problems that come when an emerging technology has to deal with the foibles of humans. More businesses should follow suit!
AI is proving bad for customer relationships as businesses, once you remove the flowery justifications and hype, are generally using it to reduce cost and replace experienced people, rather than deliver a palpable improvement for the customer.
Not only is AI too often dehumanising customer service, it is also bringing tension and burnout to employees. Particular problems include:
- Middle management burnout from devising and deploying AI. With AI implementation programmes, we have teams being given little or no training and expected to deliver a major change programme, underpinned by potentially unrealistic project and operational expectations from senior management. Given the high cost and complexity that comes with implementing a new technology, it is a recipe for failure.
- Employee burnout from dealing with the problems when the productivity gains fail to appear. As with previous technology implementations (e.g. ERP) people are not being given the skills and training to properly implement the changes and then have to deal with the consequences of the change programme’s poor implementation and performance. We are seeing a backlash from employees against the drive for productivity: not only are affected employees feeling less-and-less valued, but they also recognise that often they are now competing against the AI engine, giving them unachievable targets to hit.
- Customer service deterioration. We are seeing companies race to implement AI without sufficient consideration for how it will help differentiate them in the marketplace, and then fail to provide the necessary training and change management support to their staff, so customer service levels and ultimately profitability drop while your best staff leave. The perfect doom-loop?
What should businesses do to make their AI work: humans first
Whether you have already introduced AI or just investigating, you need a ‘humans first’ approach. It is the quality of your team and customer relationships that matter, and AI has all the potential to help enhance these, and to also destroy them!
Ultimately, your profits will be delivered by your customers so take the time to deeply consider how your AI will impact how your customers think about your brand. After all, we know from bitter personal chatbot experience that it is really yelling out to us, “This business cares so little about me that it wants to waste a lot of my time so only the determined and desperate eventually get through to a human.”
At the investigation phase, make sure any proposed implementation is treated with a healthy dose of scepticism about the ability of the technology to deliver what is being promised, the unbudgeted on-costs, proposed ROI and, most importantly, what are you risking in terms of human capital and customer service if it is poorly designed and implemented.
If AI is already in place, to get its benefits you may have to re-engineer, but make sure to do this with the involvement of those who are expected to deliver the productivity gains. To successfully implement an AI capability that will drive true competitive advantage, the investment in change management must be your priority, supporting your people so that they understand the reasoning for the change and will ultimately be prepared to own and deliver the productivity improvement targets sought by the business.
Your people need to see how the integration of AI into their working life will make them more effective and successful, not subservient to the machine, with them being able to employ it as a trusted co-pilot to enhance business performance while making the working day better for every employee: the opposite of the experience of too many workers and customers so far,