By James Rodda, Lead PM Antipodes Global SMID FUND (ASX: MIDS)
The market has reached a verdict on software: artificial intelligence (AI) will hollow it out. Since January that fear has driven a heavy sell-off across the sector. The market is right about the disruption and wrong about who it hurts. The next few years will separate the software AI makes obsolete from the software AI finally makes valuable – and the gap between those two outcomes is where the interesting questions sit.
The instinct in a sell-off like this is to treat a sector as a single trade, marking down the winners and the losers together. But AI is not a cyclical wobble that reverses; it is a structural shift that resets the ground rules of an industry. Structural change of that kind rewards the businesses built for the new regime and punishes those anchored to the old one – and it rarely does both to the same degree. Selling the sector as one block ignores that distinction, which is exactly why indiscriminate selling tends to leave resilient businesses priced as though they share the fate of the vulnerable ones around them.
The clearest place to see the split is in traditional software-as-a-service (SaaS) versus software built for the agentic world. A great deal of functional SaaS is genuinely exposed: if a task is simple enough, a customer can now have AI build the tool in-house. The businesses worth a second look are the ones agents make more valuable, not less.
One company that captures this distinction is Serko (SKO-NZ), an online booking tool (OBT) used by some of the world’s biggest corporate travel agents, including Flight Centre and Booking.com for Business, is down ~60% from its peak.
The bear case has three planks, and each misses something.
- The first is that customers will simply build their own software. That’s true for a lot of functional SaaS. But Serko is wired into one of the world’s most complex hotel and flight supply chains, with exclusive access to Booking.com’s hotel inventory. You can copy a front end; you can’t copy that.
- The second is that agents will replace SaaS. Done well, this is a story about operating leverage, not obsolescence. Corporate travel agents run on thin margins, and around 40% of their revenue is consumed by customer service: bookings, amendments and the like. Serko’s agentic product, Serko.ai, turns a 40-minute booking into about five minutes and removes customer service labor for the travel agent. Take that cost out and the economics change fast. That is not a threat to a business like Serko; it is the opportunity as travel agencies will want the product.
- The third is that fewer white-collar jobs means fewer seats to monetise. This is where the pricing model does the heavy lifting. Software billed per seat is directly geared to headcount: if AI thins out the ranks of the people logging in, the revenue base shrinks with them. Transaction-based pricing breaks that link. What matters is not how many staff sit in front of the tool but how much activity flows through it – and business travel does not disappear because a team is leaner. If anything, the trips that remain still need to be booked, amended and serviced, so the volume that drives revenue proves far stickier than the seat count. Serko earns on a per-transaction basis rather than per seat, and travel remains a relationship business. The impact here looks limited.
None of this is certain. Serko.ai is still in beta, the moat is real but not impregnable, and larger, well-funded players are chasing the same shift. Any of these judgements can prove early, late or wrong. But they are judgements about the business itself – the kind worth making carefully, precisely when the market is making the opposite one on emotion.
That is the broader point. The sell-off has priced software for a single outcome, as though AI is uniformly corrosive to everything with a subscription attached. The more useful question is which businesses emerge on the other side worth more than they were before – the software AI renders redundant, and the software AI finally makes indispensable. Telling the two apart is where the next few years will be decided.
Serko is held in the Antipodes Global SMID Active ETF (ASX:MIDS).
Disclaimer: This communication is prepared by Antipodes Partners Limited (“Antipodes”) (ABN 29 602 042 035, AFSL 481 580) as the Investment Manager of the Antipodes Global SMID Active ETF (ASX:MIDS) (ARSN 654 913 309) (‘the Fund’). Pinnacle Fund Services Limited (‘PFSL’) (ABN 29 082 494 362, AFSL 238371) is the product issuer and responsible entity of the Fund. PFSL is not licensed to provide financial product advice. PFSL is a wholly-owned subsidiary of the Pinnacle Investment Management Group Limited (‘Pinnacle’) (ABN 22 100 325 184). This communication is for general information only. Prospective investors should read and consider the Product Disclosure Statement and Target Market Determination before deciding whether to acquire, or continue to hold units in the Fund at www.antipodes.com. Past performance is for illustrative purposes only and is not indicative of future performance.
