Revolutionize Your Workflow with GenAI: Discover the Power of Reduce, Redesign, Reinvent!
A practical roadmap for 2024
The 3R’s of GenAI: Reduce, Redesign, Reinvent.
The biggest room in the world is the room for improvement. -- Helmut Schmidt, Chancellor of West Germany 1974-1982.
The torrent of innovations, investments, and implementations of AI and GenAI during this past year may have been the greatest flurry of deployed new function since World War II when the great powers exploded the frontiers of science and engineering desperately fighting for their existence. ChatGPT’s global adoption was faster and broader than any other technology in the history of humanity and billions of venture capital as well as in-kind services were rapidly doled out to newly born startups as technology giants and venture investors alike paddled eagerly to surf perhaps the biggest technology wave ever.
It would be hard to criticize the skeptical executive who might say, everything goes through a hype cycle. And that would feel like a reasonable stance if it were not for the fact that every study of the productivity of AI at the task level seems to show at least 14%[1] improvement of task throughput, for customer support agents using a GenAI assistant or the work of BCG consultants as studies by Karim Lakhani et al saw a 25.1%[2] decrease in time needed to complete tasks. (It is important to note that there was a productivity decrease when the tool was used on the wrong task – hence the term “jagged edge”.) In addition, both studies find that these tools help take the least productive and/or newest members of the workforce doing the task and make them more capable, faster. That is, organizational capability is enhanced.
In addition, at the corporation, level in just over a year’s time since ChatGPT hit upon the world’s day to day consciousness, major firms have deployed new models in critical business functions: McDonald’s with ask pickle for ordering and franchise operations, Coca Cola[3] in its making of ads and images, Allen & Overy a multi-billion dollar law firm using a generative AI tool for contract negotiation – just to name a few. Each of these is not some marginal task, off in the Siberia of the business, this is McDonalds on ordering and operations, Coke on brand and a Allen & Overy on negotiation!
The Three R’s
In this whirlwind of activity one should start with those tasks that are made up primarily of words, images, numbers and sounds – what we have called WINS[4] work – that are also highly digitized. Once you have chosen a function or task to begin, we suggest guiding yourself with the “three R’s of GenAI”: reduce, redesign, reinvent. As cheesy as a mnemonic can be we believe this one is useful because we have clear evidence from the academic studies that many tasks can get enhance by 10% or more. This is the domain of the first R, Reduce. Simply put GenAI can help you automate customer contact, sales preparation, document review, policy administration, etc., etc. Think of them the way a contractor thinks of power tools. No contractor in the world would deploy their talent without power tools in hand – so too your knowledge workers need better tools to be competitive.
The next level is task Redesign. With GenAI tools every individual can do some of the work a team used to do, and teams can be like teams of teams. For example, any copywriter can not only make illustrations but also animate, generate music and more for any idea. The task itself can be monitored and redesigned. For example, at GetJerry.com, a comprehensive insurance and financing site, they went from over 90% of customer interactions being elevated to a human to under 10%. And, the silicon intelligence monitors the interactions with generative AI, which means that within seconds any difficult or customer frustrating situation can be elevated immediately. This is a completely new process. We believe that with task level automation you may be able to get 10-50% productivity improvement on a one time basis, however, ifi you can redesign the task, it’s also usually much more scalable For example, GetJerry has about 15 people in customer service (over 5,000,000 customer) and they think they can grow much bigger without growing the organization much.
We have yet to see a firm that has reinvented a task with GenAI, but we predict in 2024 we will see such a revolution. Let’s take an example of a completely category reinvention with the creation of Craig’s list. In almost all cases, automation creates an economic surplus. This surplus can be shared with one of three constituencies: the capital in the form of higher returns, the labor in the form of more compensation, or the customer in terms of lower price. Craig reinvented ads by automating and aggregating classified ads and gave most of the economic bonus to the customer by only charging for job listings by employers. Even with the vast majority of the surplus, and a rudimentary user interface, the revenue per employee at Craigslist is over $14,000,000 – compared to just over a million per employee at Goldman Sachs. In the process the obliterated the local, regional and national classified ad business which as many may know was one of the most profitable parts of the newspaper business.
My prediction is that many industries will have a Craigslist enter and create a new model that has vast scalability for labor (e.g. $14,000,000/employee) and may share a lot of the economic return back to the customer. So, every executive should start to pilot where to reduce, consider where to redesign for ongoing value. And keep an eye out for reinventions in your space because once the market believes that a new model – like Craigslist – will win, your firm and its value will immediately be worth a lot less and have a hard time catching up.
[1] Brynjolfsson, Li & Raymond, National Bureau of Economic Research Working paper 31161, April 2023.
[2] Dell’Acqua, McFowland, et al., Harvard Business School Working Paper 24-013, “Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality.
[3] https://www.searchenginejournal.com/6-ways-coca-cola-uses-generative-ai-for-advertising-and-marketing/504696/
[4] https://hbr.org/2023/09/where-should-your-company-start-with-genai