top of page

The Three Biggest Software Investors in the World Can’t Agree on What Happens Next. That Should Concern You.

  • Unemployment Society
  • Apr 23
  • 3 min read

Workday’s CTO lasted eleven months. He left in March to join Anthropic as an individual contributor, trading a C-suite title at an $8 billion company for a technical staff role at the AI lab that is now actively hiring engineers to build HR software (Business Insider, The Information). He went to go build the thing that might replace his last employer.

If you advise executives on workforce strategy, if you serve as a fractional CHRO or COO, if your clients rely on you to tell them what tools to buy, this is the part of the story you should be paying attention to. The entire category of software your clients depend on is being repriced, rebuilt and in some cases abandoned by the people who built it.

The Software Market Is Splitting Apart

By mid-February 2026 an estimated $1 trillion in enterprise software value had been destroyed (CNBC). ServiceNow, Salesforce, HubSpot, Workday and Adobe all cratered in unison. Atlassian fell roughly 57%year to date (TIKR). Surveys indicated that 40% of IT budgets were being reallocated from traditional SaaS subscriptions to agentic platforms and LLM token usage (FinancialContent).

The three most sophisticated software investors in the world are reading this differently and none of them agree.

Thoma Bravo, $183 billion in AUM, held its annual LP meeting in Miami last month. Managing Partner Holden Spaht pushed back on the blanket AI-disruption thesis, citing 17% top-line growth in public SaaS, 74% gross margins and 80 to 95% of next year’s revenue already under contract (Augment). Their position is that mission-critical compliance-heavy software is insulated and the selloff is a buying opportunity.

Vista Equity Partners, $100 billion in AUM and 90 portfolio companies, takes a different read. Robert Smith argues that companies must maintain sovereignty and dominion over their proprietary workflows and data sets to survive, and that those who chased ARR without locking down their data contracts gave up their moat (Fortune, CNBC).

Coatue argues the model itself is being rewritten. Their thesis is that the unit of economic value is shifting from selling software per seat to selling work per output, expanding the addressable market from $200 billion to $5.5 trillion (Coatue C:\Takes).

Three firms. Zero consensus. And the enterprise software is already losing its architects.

The HR Tech Stack and the Org Chart Are Breaking at the Same Time

Anthropic has posted job openings for engineers to build “people products” covering hiring, training and promotion management, which are core functions handled by Workday’s applications (PYMNTS, The Information). ERP Today noted that AI vendors are moving closer to system-of-record territory and are no longer staying at the tooling layer. The AI lab that triggered a $285 billion single-day wipeout in software stocks when it launched Claude Cowork is now building the HR software category from the inside.

If you are recommending Workday implementations, building talent acquisition processes around Greenhouse and Lever, or integrating BambooHR into a client’s workflow, the ground underneath your recommendations is moving.

And the organizations those tools were designed for are changing shape at the same time. Amazon CEO Andy Jassy is increasing the ratio of individual contributors to managers to remove layers and streamline decision-making (Fortune). Gartner predicts 20% of organizations will use AI to flatten their structure by the end of 2026, eliminating more than half of current middle management positions (People Managing People). LinkedIn data shows that job postings with “manager” in the title declined 12% year-over-year in early 2026 while postings for “lead” and “principal” roles grew by 18% (Metaintro). Between May 2022 and May 2025 the number of managers at public companies declined by 6.1% (Lepaya).

The pyramid is compressing. Your clients are cutting middle management roles, expanding spans of control and expecting individual contributors to operate with more autonomy than traditional org charts ever required. A hiring platform built for a five-layer approval chain stops working when half those layers are gone.

What This Means If You Advise on Workforce Strategy

The tools your clients use to hire and develop people are built on a per-seat pricing model that is under existential pressure. The organizational structures those tools were designed for are flattening in real time. And the AI labs that are accelerating both of those changes are now building their own HR products to fill the gap.

If your practice depends on recommending workforce technology to your clients, the question you should be asking is whether the tools you recommend today will still make sense in 18 months, and whether your business model gives you any stake in the answer.

pēpēlwerk was built as connected workforce infrastructure for exactly this environment. If you want to understand how it works and what it means for your practice, start here: https://www.pepelwerk.com.

 
 
bottom of page