This is a Forbes contributor's analysis piece, not breaking news: fintech investor Alex Lazarow uses two already-public case studies (Medallia's April 2026 debt restructuring and Chegg's multi-year decline) to argue that generative AI is bifurcating software valuations based on whether a company's moat is a repeatable output or something harder to replicate, like proprietary data, integrations, or human expertise. For practitioners, the useful takeaway is less "AI crashed two companies" and more a checklist for what still holds pricing power.

What happened

Lazarow's June 29, 2026 Forbes piece revisits two cases. Thoma Bravo, which took Medallia private for $6.4 billion in 2021, agreed in April 2026 to hand the customer-experience software company to a Blackstone-led lender group in a debt-for-equity swap, wiping out roughly $5.1 billion of Thoma Bravo's equity - reported as the second-largest private-equity loss on record, per Verdad Advisers. Thoma Bravo co-founder Orlando Bravo has since said "It was a big mistake," attributing it to underwriting fast growth that did not materialize. Separately, Chegg's market capitalization fell from roughly $14 billion at its 2021 peak to about $100 million, a decline that accelerated after ChatGPT's late-2022 launch and Google's AI Overviews began answering homework-style queries directly in search results, cutting into the click-through traffic Chegg's tutoring-answer business depended on.

Technical context

Both companies sold what Lazarow frames as commoditizable output: Medallia's customer-feedback analytics and Chegg's textbook-style answers. As generative models made producing similar outputs cheap, buyers had less reason to pay for the packaged version. That is distinct from companies whose value sits in proprietary data pipelines, workflow integrations, or regulatory/compliance lock-in, which are harder for a model to substitute.

For practitioners

The generalizable signal for product and data teams: audit which parts of your product are "outputs" a capable model could reproduce cheaply versus which parts depend on data, integrations, or trust that a model cannot easily replicate. The former is where pricing power is most exposed; the latter is where investment in defensibility pays off.

What to watch

Whether the "Saaspocalypse" framing Lazarow cites proves durable or overstated - his own piece notes causes are multifactorial, including higher rates and platform disruption (Google AI Overviews), not AI alone - and how the Blackstone-led group's planned $150 million injection into Medallia performs as a test of whether legacy SaaS businesses can reposition around AI rather than be replaced by it.

Key Points

1Forbes contributor Alex Lazarow used Medallia and Chegg as case studies for how generative AI is splitting software valuations.
2Thoma Bravo lost about $5.1 billion when Medallia was handed to a Blackstone-led lender group in April 2026, its own co-founder called it a mistake.
3Practitioners should distinguish product value tied to replicable AI-generable outputs from value tied to proprietary data or integrations.

Scoring Rationale

This is a Forbes contributor analysis piece synthesizing two already-reported case studies (Medallia's April 2026 recap, Chegg's multi-year decline) into a software-moats thesis, not fresh news - the underlying facts are 2-14 months old. The thesis is well-corroborated and practically useful for product/valuation strategy, but as derivative commentary rather than a new development it is scored in the solid tier rather than notable/major.

MoreGenerative AI news