Vibecoding isn’t going to kill business software services
SaaS companies have an opportunity to use AI to make a more intelligent software
Vibecoding — using verbal prompts and AI to write code — is gunning for software giants. Shares in SAP, briefly crowned Europe’s most valuable company last March, are down by a fifth since then; ServiceNow in the US has lost two-fifths of its value in the past year.
On the face of it, there’s good reason for investors to run scared. AI tools such as Anthropic’s Claude Code theoretically enable anyone to do what these office-software stalwarts have done for the past half century or so. If the intern can create an invoicing app in their lunch hour, there’s less reason to pay monthly fees for a smorgasbord of services only some of which are frequently used, as often happens with Software-as-a-Service (SaaS).
Circumstantial evidence abounds. Searches for vibecoding, as measured by Google Trends, have spiked worldwide and especially in the US this month. Social media platforms yield plenty of “how to” videos — as well as woeful tales of lost contracts at the big SaaS players. For bearish investors, a short stroll to the nether reaches of their IT and HR departments ought to provide evidence of these new ways of working.
While barriers to entry are falling, and the so-called hyperscalers — cloud providers such as Amazon that offer their own AI tools — create a new threat, it will nonetheless take more to dislodge the SaaS incumbents.
This is an industry used to evolving, having moved from clunky discs to the cloud. Earnings, so far, are holding up. Sage Group on Tuesday said it was expecting organic revenue growth of more than 9 per cent this fiscal year. Analysts expect SAP, which reports its 2025 numbers on Thursday, to keep increasing adjusted operating profit by 15-18 per cent a year through 2030.
As such, retreating share prices are mainly the result of the lower multiples investors are willing to pay for their profit, which reflects a more pessimistic view of the future. Over the past year, multiples have come down by more than a third, to an average of 25 times forward earnings.
Yet SaaS growth is typically of pretty high quality: subscriptions mean recurring revenue, which is usually stickier, while licensing software doesn’t require chunky capital spending. Besides, these companies are also deploying AI, either organically or via partnerships and acquisitions. Israel’s Nice last year paid $1bn for Germany’s Cognigy, which uses agentic AI for customer services, while SAP has partnered with Perplexity.
Where SaaS companies have an untapped opportunity is in using AI to make the software that runs a customer’s business systems more intelligent and reactive to real-world events. Clients might be able to adjust their supply chains to cope with mooted tariffs, for instance. The odds are that companies, for now, will prefer to buy such offerings from the pros, rather than spend weeks eliminating bugs from home-cooked alternatives.