Artificial intelligence (AI) has garnered significant attention and funding ever since OpenAI’s ChatGPT became a big success. Big tech companies such as Microsoft, Google and Meta have upped their investments in AI, and companies across sectors are preparing for an AI disruption. For India’s information technology (IT) services companies, this has two implications. First, new business opportunities are opening up as their customers seek help rolling out AI projects. Second, AI is changing how software applications are created and maintained. LinkedIn, in its latest Future of Work report, said that software engineers had the highest share of skills that could be augmented by generative AI.
India’s top five IT services companies have all said they are training their employees in AI. Tata Consultancy Services claims to have trained 25,000 of its employees in Microsoft’s Azure Open AI, Infosys is training 40,000 employees on generative AI skills, and Wipro said it was spending $1 billion to train its entire staff on AI, and recently announced a new generative AI centre of excellence at Indian Institute of Technology, Delhi.
All this is expected to boost the AI skill pool, which has been picking up over the last few years. According to the LinkedIn report, the share of members in India who added AI skills to their profiles increased 14-fold over January 2016, the fourth highest rise in the world. The big question for IT companies is whether the AI skills diffusion, as LinkedIn calls it, will boost revenues and productivity.
Productivity boost
For Indian IT services companies, growth in revenues has always been accompanied by a proportional growth in employees. It has been a natural outcome of the predominant billing model, where IT companies charge their clients based on the number of people allotted to a project. Large companies have been trying to break out of this ‘linear model’, to aim for a higher revenue per employee ratio—but with limited success.
Even when the ratio went up, as it did in the most recent quarter for all the top five companies, it was a result of hiring decisions, which are mostly driven by deals in the pipeline. The total employee count for the top five companies fell by about 22,000 in the June quarter compared to the March quarter, reflecting economic uncertainties.
Now, IT companies are excited about AI, as it could potentially drive both revenues and productivity, boosting the revenue per employee ratio.
Revenue share
Using AI tools can indeed boost software productivity. According to a Microsoft Research study earlier this year, developers who used an AI tool (GitHub Copilot) finished a task 50% faster compared to those who didn’t. While it’s not clear if IT services companies can replicate or scale such productivity gains, anecdotal evidence suggests gains can be significant. However, the catch is in the billing models.
Indian IT companies mostly bill their customers based on time and effort. When that comes down, revenues drop as well, keeping the ratio the same. For this to change, the industry’s predominant billing model should move to an outcome-based model. That would give IT companies more room to save costs and improve margins. Without such changes, a higher revenue per-employee ratio is unsustainable.
In the past two quarters, while the employee additions have dropped for the top five players, the drop in operating profits has been even higher.
Value addition
A bigger attraction for IT companies is AI’s revenue potential. According to a June 2023 McKinsey report that analysed 63 use cases, generative AI had the potential to generate $2.6-4.4 trillion in value across industries. The modes could vary. For example, in retail, the bulk of value addition would happen by boosting performance in functions such as marketing and customer interactions. In high tech, the boost would come from its ability to increase speed and efficiency of software development. In banking, AI could help improve the resolution rate of customer emergencies.
IT companies are seeing early signs of this. TCS said it’s working on over 50 proofs of concept and pilots, and has over 100 projects in the pipeline. Infosys said it was working on 80 AI projects. How AI changes revenue and profit-per-customer ratios will be keenly watched.
www.howindialives.com is a database and search engine for public data
www.livemint.com
Source link