AI Will Kill Offshoring
All trends begin and end. And nowadays, the cycle can be very rapid. Remember Lean Six Sigma? Remember crypto? Cycles seem to be getting shorter.
Offshoring has been a trend for quite awhile. It was driven by economics and the growing pool of tech talent in countries with low currency exchange rates. Companies offshore for one of these reasons:
Product development: Save money - offshore resources are significantly cheaper, with engineer time in the $30-60/hr range or better.
Support: Save money + cover global time zones, languages, and cultures.
Access to a broader pool of skills.
Offshoring has some high direct costs. Travel is a big one. For example, travel from the US to India can be expensive, because if one goes all that way, one will stay for several weeks. And it usually makes sense to bring a large group each way to meet the rest of the teams. That adds up. But the travel costs are overwhelmed by the low labor costs.
The indirect costs are far more significant. They include:
Collaboration challenges – most communication must be asynchronous. Collaboration needs to be more writing-oriented rather than speaking oriented, and that shifts who will be effective in a group. (Note: the shift might be for the better.) People who build trust by “looking in the eye” will have trouble working with remote people. (Note: not everyone builds trust that way.)
Leader effectiveness challenges – Leaders who are accustomed to “walking around the room” or “pulling everyone into their office” need to learn new approaches. (But arguably, remote work has already forced that shift.)
Commodity level capabilities – Resources obtained from large volume offshore resource providers tend to be “commodity” – not best in class.
Coordination challenges – Having large numbers of medium-skilled people is a more difficult situation to coordinate than having fewer highly skilled people, and can reduce agility and innovation.
And Now Comes Generative AI
If you are a business or tech leader in a large organization, you have probably looked into what the latest AI can do. You might have found that it depends. For example, for routine work, generative AI can reduce the learning time for new hires.
By “routine work” we mean things that are craft or task oriented, and that don’t require sophisticated judgment. These include things like legal research, level 1 customer support, and logistics.
But what about creative work – work that requires advanced knowledge and experience, such as engineering? The latest AI has been very effective in suggesting novel ideas and helping people to think through hard problems. For example,
But we still need the person – someone who will know what questions to ask, can make sense of the deep discussion, and can be trusted to make the final decision or complete the design.
In other words, AI can make people more productive, for both routine and creative work. For creative work, there is not much data yet on just how much more productive, but the signs are that it will be a lot. For example, some companies have decimated their marketing departments, retaining only a few of the most capable people, who now use AI to help brainstorm and do initial copy generation. Further, as people gain experience in how to use generative AI, they will begin to use it more effectively – we are in early days, and the tech itself continues to improve.
Today’s Leaders – What Is Top of Mind
There has been a shift since the pandemic. Today’s business leaders are very cost conscious. They need to support growth, but they are leery of spending or committing to future costs. That means that efficiency is paramount.
To be efficient, the leaders throughout the operational levels of the organization need to be effective. They are the ones who are spending the money. Thus, efficiency is a leadership issue – not a budgeting issue. As Byard Bogsnes of the Beyond Budgeting Roundtable has written,
“It is changes in leadership behavior that yield the largest benefits, not the process changes”.
[ref “Implementing Beyond Budgeting”, by Bjarte Bogsnes, Wiley, 2016, p211]
Another thing that is top of mind is flexibility. This is because change is the new normal. “Out” is the idea of return on investment. “In” is the view that return is a stream, and that a point at which you can measure your return never arrives like a ship coming into port – instead, you need to always be measuring, and always evaluating options against both short-term and long-term goals that are also always subject to revision.
Leaders are also concerned about their supply chains. They want to make sure that they will not be caught by surprise if there are geopolitical changes that alter access to what they need.
Finally, expertise: companies are finding that the rapid changes in technology are catching them by surprise. One day they need people of one skill, and the next day they need another skill. And these skills are hard to come by: they are sophisticated things like DevOps and AI. And the commodity level won’t do for these – not if you want results. You need people who are expert level. That’s a huge challenge, and offshoring cannot meet that challenge, although partnering with overseas firms can. Those kinds of technology partners are not low-cost suppliers: they are companies that employ experts, and they could be anywhere. Most are not within commuting distance of your offices.
AI to Amplify Productivity Instead of Offshoring
These various circumstances will converge to create a new trend: instead of offshoring to obtain large numbers of low cost commodity tech resources, it will make more economic and strategic sense to bring the work back in, to smaller teams supported by AI, and in partnership with others who can fill in pieces of the tech puzzle that you need to assemble.
By doing that, all of the direct and indirect costs of offshoring are eliminated, and people are more productive. For example, a smaller number of creative people is easier to manage, and tends to be more effective and more nimble. And fewer people translates into lower cost. If the people are highly effective using AI tools, it is not unreasonable that AI tools can eliminate the cost advantage of offshoring.
Again, there are not numbers on this yet, to our knowledge, but the signs are that it will make more sense, strategically and economically, to have fewer and smaller highly capable AI-leveraged teams for creative work that large numbers of low cost commodity offshore teams. And for routine work, it is very possible – I personally think likely – that we have not learned how to use AI effectively yet, and that AI will not only augment, but replace a lot of routine work.
Also, the increased nimbleness (agility) of having much smaller, highly effective AI-enabled people for creative work is likely so advantageous that it eclipses any cost advantages that offshoring might provide for creative work.
Global collaboration in the form of technology partnerships will continue to increase. It is only offshoring that will wane, because it will no longer make sense to seek commodity intellectual resources to lower costs. But global collaboration enables one to find the best resources, no matter where they are – cost is not the driver: expertise is.
The offshoring market will fizzle.
This is a dire prediction for countries that have large offshoring economies. Those countries will have to find ways to be the sources of innovation, instead of merely providing expertise for someone else’s products. And that – if it is done – will be a good thing.
