MW The old software investing playbook is dead. Here's where to put your money now.

By Christine Ji

Software companies are facing a reckoning as AI lowers barriers to entry. Experts share how to find winners in an evolving competitive landscape.

Salesforce's stock has fallen this year, and investors are questioning the impact of AI on the software sector.

U.S. and European software companies have seen their shares sell off on the back of artificial-intelligence jitters in recent days - and big questions about the future of the software sector are likely to linger.

The pandemic demand boom and near-zero interest rates once sent software companies' valuations soaring. Now, the landscape is shaped by a higher cost of capital and rapidly evolving AI technology.

That's made prominent software names big losers this year. Salesforce Inc.'s stock (CRM) has dropped 30% this year, while Adobe Inc. shares $(ADBE)$ have declined 20% year to date. At the height of the postpandemic software boom, Salesforce and Adobe shares traded at price-to-earnings ratios of over 100x and 70x, respectively - but those valuations have since come down to 37x and 22x presently, according to FactSet data.

The iShares Expanded Tech-Software Sector ETF IGV has dropped 3% since Aug. 6, the last close before OpenAI launched its ChatGPT-5 product. Some of the biggest laggards include the enterprise AI software provider C3.ai Inc. (AI), down 23%, and the data-onboarding company LiveRamp Holdings Inc. $(RAMP)$, down 21%. Both companies recently reported earnings.

Investors may have thrown the baby out with the bathwater, according to Mizuho analyst Jordan Klein. AI fears could be "short-sighted," as software companies managing valuable enterprise data will likely still maintain relevance in the years to come, he wrote in a recent note. Klein still sees compelling opportunities in Take-Two Interactive Software Inc. (TTWO) and Atlassian Corp. $(TEAM)$. However, he anticipated sentiment would only get worse as new AI tools continue to emerge and investors steer clear of what they see as a "falling knife."

As Big Tech companies and new AI-native apps launch their own AI tools, the software space is getting crowded and traditional software-as-a-service companies are facing a reckoning. The fear is that people won't need to pay for software that provides services like developer assistance and content generation, now that products from OpenAI and rivals are getting better at these tasks themselves.

That means James Dowey and Matt Ward, co-managers of the AXA Framlington Global Technology Fund, are making some changes to their investing strategy: They're paring back their software allocation and carefully curating the remaining names.

Ward told MarketWatch that software leaders like Salesforce and Adobe face a "competitive, disruptive risk" that large language models could commoditize certain functions of these companies' expensive software with much cheaper and more flexible AI-powered tools. If an AI agent can generate images or create email campaigns, then companies might opt for those options instead of paying for traditional software.

Consumers, not enterprises, have been driving most of the early AI monetization, according to Bryan Wong, portfolio manager for Osterweis Capital Management's small-cap growth strategy. That means companies are finding more success selling AI products directly to individuals rather than large businesses. Take ChatGPT, for example: Last year, OpenAI shared that 75% of its revenue came from consumers. And just a few weeks ago, the company disclosed that it had recently hit $12 billion in annualized revenue and surpassed 700 million weekly users.

The direct-to-consumer model leads to faster adoption than the enterprise sales process, which involves lengthy review periods and contract negotiations. An individual can choose to use or cancel their ChatGPT subscription at any time; the same can't be said for companies with multiyear enterprise software contracts.

Many traditional software companies have followed a seat-based pricing model, meaning that they increase their sales by growing individual users. But as AI-native programs help companies become more efficient, software clients are downsizing headcount, leading to lower demand for seats.

Don't miss: AI is eating software, and Adobe is on the menu. Why the stock could be in trouble.

How to find winners

As AI disrupts the software industry, there will be winners and losers.

Investors shouldn't write off all software names, but being a selective stock picker is key in today's environment.

Look out for a few key characteristics when screening for stocks. Companies that offer systems of record, charging through non-seat-based pricing models, a vertical business model and a regulated industry have barriers that provide insulation from "AI potentially taking over your business," Wong told MarketWatch. "The companies that serve those markets have the distribution and data."

Unlike horizontal software companies like Salesforce - which sell a generalized product that can be used across many different industries - vertical software companies provide highly tailored solutions for a specific market. Meanwhile, operating in a highly regulated end market, such as pharmaceuticals or financial services, gives software companies a competitive advantage, as there are high barriers to entry and a sticky customer base. It's also harder for newer AI technologies to disrupt these services due to the confidential nature of the data.

Dowey echoed a similar sentiment, saying that these types of companies can more easily absorb the costs of AI technology and find more ways to use the technology to solidify their dominance. While Dowey doesn't own the life-sciences software company Veeva Systems Inc. (VEEV), he likes its business model. Veeva provides regulatory compliance and data-security services for complex processes like clinical trials, meeting highly specific needs for clients.

System-of-record companies manage critical and authoritative data, and Wong is a fan of the investment-accounting company Clearwater Analytics Holdings Inc. (CWAN), which helps institutional investors streamline their operations by putting fragmented data into a singular platform.

Investors can also find safety in Big Tech companies like Microsoft Corp. $(MSFT)$, which has an advantage due to its hybrid business model that utilizes a non-seat-based model. Its Azure offering operates on a pay-as-you-go model, billing for total consumption instead of number of users.

See also: The entire stock market is being carried by these four AI stocks

Wong flagged Duolingo Inc. (DUOL) as an example of a software company successfully adopting AI. The language-learning company has adopted an "AI-first" strategy and aims to limit headcount growth. Duolingo has used AI to launch 148 new courses, according to its most recent earnings report. Its consumer-facing product also helps the company realize more "immediate" return on investment, according to Wong.

Dowey and Ward favor ServiceNow Inc. (NOW), which they believe has a stronger competitive advantage than Salesforce because its software and data stacks are more unified.

"It's no coincidence that we're starting to see ServiceNow move pretty aggressively into CRM, parking the tanks on Salesforce's lawn," Ward said, referring to customer-relationship-management software.

If you're a long-term investor, having software exposure might not be a bad thing, and the recent selloff could offer opportunities to get quality names at cheaper prices.

But at the end of the day, investors wary of the software industry's AI payout are putting their money into the AI picks-and-shovels trade, with hardware companies like Nvidia Corp. (NVDA) showing tangible financial benefits from AI. As Bloomberg has noted, comparing the aforementioned iShares Expanded Tech-Software Sector ETF to the PHLX Semiconductor Index SOX shows that software stocks are currently trading near their lowest relative valuation compared to semiconductor stocks in the past year. The ratio between them is the lowest since Oct. 29, according to Dow Jones Market Data.

Until software companies can show that their products can't easily be replicated by AI, investors will be hesitant to bet on them.

"The money is in the infrastructure; it's not really in the software as of yet," Dowey said. "That's the complete opposite of a ZIRP world," he added, referring to a zero-interest-rate policy.

Also read: It's not an AI-bubble; it's a longer-term bull market, says this Wall St firm

-Christine Ji

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

August 16, 2025 07:30 ET (11:30 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.