Software Development - Nail Guns versus Hammers... Then the Next Wave and the Giant Impact

CursorAI has a really well-put-together, team/real-world-oriented set of tools that improves the development of code. That's writing, editing, and reviewing. It has clever predictive capabilities to think ahead for you and deliver what comes next, sometimes in large chunks. It is broadly context-aware, meaning it is not just the nail gun to the hammer for the coder but an accelerator across massive code bases that live inside companies. It uses at least three of the GenAI platforms to do this.

It has grown to $100M in annual recurring revenue faster than any tech firm in history, which means that the people who actually use it find it very beneficial. Again, it is faster than any firm in tech in history.

A rough estimate is that it makes the coders 30% more productive, the code better and more secure. That means your code base becomes functionally less costly. Think of that as a factory continually becoming more efficient. I think that the estimate might be low.

The Bigger Picture

Most Product Management to Product Development/Engineering chains inside large firms are a mess. Very few people upstairs spend time on those chains' architecture or are deep enough to ask the right questions. A quick check of board meeting agendas would indicate how rare that is at that level. Generally, it is never. It may be an odd topic for the board unless, of course, there is a tech component in the product/ecosystem you sell into, and how ideas get prioritized and then turned into product matters.

There was a product development operation with 1000 engineers, with a record of zero percent customer satisfaction and 10% on time/on budget at a reasonably well-respected firm. They switched leaders, eliminated 45% of the headcount and 66% of the top management, and doubled output while hitting 100% Customer Sat and 95% on time/budget. They did that in a year. It is an example of the upside. Most of that was because of just how horrible it was, a total lack of program management, and, in all fairness, the super hardworking, intelligent, engaging, pragmatic leader they put in charge of it. I have seen worse inside a couple of giant FS firms, a global-scale retailer, and an airline.

The Next Wave

The smart CursorAI-type platform will move beyond code efficiency and effectiveness into the actual Prod Mgmt/Prod Dev world. It will incorporate innovative portfolio management and program management. It will include a war on technical debt and a continuous improvement TCO focus for the IT/Tech Factory. Most IT is managed as one thing. It is seven things inside the enterprise that need different and systematic management. It will incorporate process discovery/intelligence. It will move a bunch of coding into externalized worlds for the business where low code/no code and parametric systems will again lower the throughput cost of the IT Factory.

That will have another 30% productivity/effectiveness impact.

The Giant Impact

In the distant future, maybe two years from now, this will only be inside about 10% of the companies that thought through the company-wide view of their 'IT Factory'. They will, however, be 60% more efficient/cost-effective and twice as agile/fast as their competitors. They will be the ones in the knife fight with a gun.

That, of course, supposes they have a good strategy. No amount of great tech saves you from a bad strategy. You can Thelma and Louise yourself in a Ford F150 or a Bugatti Chiron, the end is till the same. (GenZ pls use GenAI to understand that reference...it's from a movie that predates the Web by two years and had an unknown actor in it called Brad Pitt).

This will create a migration towards this model. That will create two socioeconomic impacts. It will reduce the employment base in that area over the next decade in a way that is registered at a national level. There are about 2.0M people employed in a broad definition of software developer including the support elements around that in the US. It could be triple that in India. The second socioeconomic impact will be seen in teh gap between nations that make this transition well and those that don't because of how central this will be to revenue success across 50 verticals.

p.s. The market for outsourced software development work will not support all the current incumbents, and those that haven't modernized will fail. There will be a few emergent new players in non-traditional locations who build a next-gen model from the ground up that is a public-private partnership and architected to feel like insourced resources with a tightly bound onsite/offsite component that has deep domain expertise and cultural synergy (yes that's a real thing).