Is AI Outsourcing on Steroids?

Many of the firms with expanding AI offerings were embracing Outsourcing during 2000-2010. The fit seems natural: AI improves the pitch for outsourcing by improving the offering; many of the organisations have mature capture-transition-operate capabilities in place to move business operations; and finally as most Business Process Outsourcing offerings already leverage straightforward automation to derive value, the step to using AI to further “automate” appears a straightforward gain in maturity. So is AI just Outsourcing on steroids?

Outsourcing: Where did it all go wrong?

Outsourcing had a grand aim, and a compelling case for investment, covering three main propositions. Companies could focus on their USPs and core business, rather than wasting management time and effort on running a set of supporting functions that provided no market differentiation. Service providers are arguably more experienced and more able; they have a depth of experience and the scale to make investing in skills worthwhile. Finally, of course, economies of scale and remote location in low cost economies allowed them to offer rock bottom prices. It was the perfect elevator pitch: “we remove a problem, deliver it better, and it’ll cost you less.

In reality, many business cases were not realised. No doubt there are numerous reasons, different for each case, but briefly there were 3 major issues related to the 3 sell-points.

1. It turned out that even the non-core business was flavoured with the customer’s culture. Process maps might have painted a picture of standards, but the reality relied to a certain extent on familiarity as much as skills. More than this, through extra-mile effort, value was being added in "support" activities without the organisation's appreciation or even awareness. This discretionary effort, based on pride in the company, didn't transfer to offshore 3rd parties.
2. The uplift in skills maturity available was often perceived as heavily outweighed by the barriers to leveraging these skills: different culture, remote working, dislocation of team. To the average employee it felt like there were less skilled people to hand, rather than more.
3. The deals made often focused more heavily on the cost reduction opportunity than any other line in the business case. Other upsides were icing on the cake, but the real goal was to remove cost. You get what you measure. More to the point, you only get what you measure. Service providers pared-down the value-adds to meet aggressive cost targets.

The Case for AI

So what happens when we view Artificial Intelligence as an extension of outsourcing and automation? We see many of the same value propositions. Indeed, early adoption AI will probably be supported by the same business cases that strengthened the Outsourcing proposition: back-office functions, done better and cheaper.

And just as the pitch is similar, so risk to organisations that take this approach are the same as those with outsourcing, but magnified. Re-examining the cases for Outsourcing, and how they struggled to deliver:

1. There is a strong perception that AI, as a section of IT, is unable to deliver in with the nuance required to bring value in a complex organisation - less able even than workers at an outsourcing provider. There is likely to be far less generosity when judging performance as the social inhibitors that moderate the impulse to blame and judge will not be present when assessing AI's performance.
2. The barriers to interaction are even greater than between two people. This comes not only from the natural barriers posed by human-computer interaction which are troublesome even for standard interfaces, but by the organisational barriers often present between IT services and the rest of the business. The business struggles to feel ownership for enterprise systems, and "automation AI" will be no exception. AI represents IT's offering of capability – i.e. “not invented here” for the business.
3. The temptation to drive for cost reduction has far greater potential to trigger unintended negative perceptions. Having your job transferred to someone who is cheaper is painful; having it forever destroyed as a human activity and delivered by AI removes ones feeling of self worth. There is risk of brand damage, industrial action, boycott by customers, etc.

There is also a tipping point where AI risks breaking the interworking and interaction between people. The structure and culture through which a company differentiates from rivals is swamped and replaced by a structure based on interaction between systems.

At some point the lakes and rivers of AI will join to become seas, and the dry land of the people will become islands. Where there was once a challenge to transport water between the bodies, it will become a challenge to link the bodies over the water. Although possible when heavily utilising outsourcing, the complexity of joining 3rd party providers into a single delivery framework independent of the customer made this outcome less viable. With AI, where we can control and define interfaces with certainty, it becomes inevitable once automation reaches critical mass.

This will present an organisation with radically new challenges in areas of innovation, market differentiation, identity and culture. Though it is tempting to think this will be addressed with positive thinking on leadership, EQ, social media and other current management fashions, the reality is that it will require a coordinated strategy to address and prevent the workplace becoming an isolating and sterile experience.

The Case for a Holistic Strategy

The solution lies in the development of a holistic AI strategy, and in particular the development of parallel paths which mature the workforce capability whilst developing the AI delivery capability. I argue that each and every strategic goal to embrace AI needs an active, supported and intentional partnering goal that reforms the organisation to work with the new capability and reinvent the role of the people in the organisation. These goals are symbiotic: one cannot be successful in realising its goals without the other.

This people transformation needs to establish a community to prevent the workforce becoming fragmented. It needs to equip the organisation with AI partnering skills, help them unlearn much of the scientific-method to work and reinforce innovation skills. It needs to embrace EQ skills, without allowing the workforce to become a social media talking shop that adds no value. To tie together the process, systems, data and people transformation requires senior engagement. So great is this transformation, and so diverse the areas impacted, that a CAO (Chief Automation Officer) is likely a smart move for an organisation contemplating the use of AI at anything approaching significant levels.

With the right steer, expertise and commitment to a strategic outcome, we can avoid the pain of outsourcing. The time to build this holistic strategy is now.

If you are interested in hearing how I can deliver for your organisation, please get in touch. I have solutions that can work for your company, today. Visit my company page or drop me a link request here.