Six tips for professional services leaders looking to embrace Artificial Intelligence
10 October 2018
We recently held a round table event for professional services leaders where we discussed Artificial Intelligence (AI) and the benefits available to businesses who embrace it.
As a group we discussed a broad range of productive use cases and many examples of positive deployment. But we also acknowledged that for some people in the profession it remains a step into the unknown, especially in those disciplines where traditional ways of working have remained both effective and profitable over the years.
We discussed some of the many ways professional services businesses can benefit from AI, including Natural Language Processing (NLP) for document review and narrative generation, optimisation modelling for more effective staffing and cognitive reasoning for the automation of certain areas of technical professional advice.
By the end of the event, those who were cautious about the technology recognised the significant disruptive opportunities and threats in play, and therefore the need to reimagine the way core parts of their businesses and industries might function in future. In traditionally people-centric businesses it is clear that those who will thrive in the digital revolution are the ones who most successfully complement technological innovation with deep business understanding and human insight.
Here is a roundup of the practical steps, questions and thought processes professional services leaders should go through when considering when and where it might be useful to consider deploying AI in their businesses.
1. Avoid shiny toys
Make sure the economic rationale for your investment is sound. Are you implementing it because you want it or because you need it? Do you have sufficient data volumes and data quality to drive the system? Can you scale the solution sufficiently to recoup the fixed costs of development?
2. Understand and control the scarce assets
AI can create a more complex supply chain than professional services firms are used to – hardware, software, data and people all become important factors of production. For professional services, the scarce assets will most often be people and data. Organisations should ask themselves questions such as: ‘What are the unique data assets we hold and how can they be used to create value?’; and ‘Do we have the brand permission / employee offering / work environment to attract the right talent on a permanent basis or do we need to tap the contractor market?’
3. Be clear - product or capability?
Whilst it is possible to generate new product offerings using AI, much of the upside in professional services is likely to be found in creating ‘bionic’ versions of existing product offerings. Be clear about what you are trying to do ‘more efficiently’ vs. what you are actually trying to do ‘better’. Simple tasks are likely to be automated, but for more complex cognitive processes we are likely to see the human decision making processes being augmented by technology (i.e. human remains ‘in-the-loop’ with final decision making control).
4. Know when to ‘make’ vs ‘buy’
Poorly thought through decisions in this area can waste valuable resources. The right answer will depend on the organisation, the use case and the technology in question. Relevant considerations include brand permission, access to talent, subsequent control of IP, speed to market, and fixed costs of development. Make sure you take the time to properly assess costs and benefits of each option.
5. Avoid the cost-plus trap
As technology solutions begin to replace some human tasks in professional services firms, fixed costs associated with system development may rise as marginal costs decline. The traditional professional services equation of ‘people x hours x charge-out rate’ will therefore be an inadequate mechanism to ensure profitability. New pricing models may be required – including sophisticated fixed cost allocation mechanisms or ongoing license fees for software based solutions.
6. New risks require new governance
As well as big upsides, AI presents a number of new risks. They are manageable but require an updated approach to technology governance. Issues to keep front of mind include: ‘Do I have the right to use this data?’; ‘Can I adequately explain the decisions that were made by my system?’; ‘Does it exhibit inadvertent bias?’; ‘Do I have a monitoring mechanism in place to ensure that it continues to work effectively over time?’; and, ‘Am I aware of the cyber risks (e.g. attack on my AI system through deliberate ‘mis-training’ by an adversary).
For more information please see: