2023's pivot to AI rollups: the seeds of today's market

ChatGPT launched in November 2022. By the first quarter of 2023, buyers were already behaving as if software, data and automation assets deserved a fresh look. The archive does not show a neat set of targets labelled "AI-native". That is exactly why it is useful. Shell USA acquires Volta, NRG Energy acquires Vivint Smart Home, SoftBank acquires Berkshire Grey and Capstone Borrower acquires Cvent all point to the same shift. Buyers were not waiting for a perfect new category. They were buying platforms they believed could carry the next automation and intelligence layer.

Exit Mode's view is that 2023 planted the operating logic of the current market before the language had fully settled. By then, "AI" had become less a standalone category than a thesis layer placed on top of existing M&A habits. Strategic buyers still wanted distribution, installed customer bases and sector capability. Private capital still wanted control and optionality. What changed was the narrative around which assets would compound fastest once a new software layer could be added. That is why 2023 matters so much in hindsight. It was the year buyers started to pay up, take private or structure creatively around data-rich and workflow-heavy businesses that could become more valuable in an AI-shaped market, even if the target was not yet marketed that way.

Buyers started with software and operating assets, not slogans

The first 2023 lesson is that serious buyers did not wait for companies to rename themselves. NRG Energy acquires Vivint Smart Home took a technology asset into a broader energy platform through an all-cash deal. Shell USA acquires Volta showed the same instinct in energy infrastructure. Metropolis Technologies acquires SP Plus later in the year reinforced the appetite for assets that combine real-world operations with a technology layer.

This is an important correction to the way founders often talk about AI cycles. Buyers rarely begin by shopping for a buzzword. They begin by shopping for assets that can absorb a new capability and produce better economics once it is in place. In 2023, that meant platforms with installed customers, software control points or operating systems that could become smarter under private ownership. The premium sat on potential operating leverage, not on branding.

Take-private activity showed where buyers wanted room to rebuild

A second pattern is how often buyers chose private ownership as the vehicle. SoftBank acquires Berkshire Grey, Capstone Borrower acquires Cvent, Global Payments acquires EVO Payments and Next Holdco acquires NextGen Healthcare all moved companies away from public-market pressure or into cleaner control structures.

That matters because narrative shifts are easiest to execute away from quarterly scrutiny. When buyers think a business needs a new operating model, more automation, deeper data use or tighter product integration, public markets can be an awkward place to do the work. Going private gives the buyer time. In 2023, time itself became an acquired asset. The buyer was not just buying revenue or customers. It was buying the right to rebuild the company without explaining every step to public investors.

The market also repriced optionality through structure

Not every 2023 deal was clean cash. Medicine Man Technologies acquires Sucellus used cash, loan notes, shares and an earn-out. Gui Zhou Grand Smooth Technology acquires Zhe Jiang Xin Shui Hu Digital Information used shares only. Those are useful examples because they show what buyers do when they want exposure to a capability but are not prepared to fund the entire bet up front.

In a market being reshaped by a new technology narrative, optionality gets expensive. Buyers want room to be right without paying as if the whole thesis has already been proven. Sellers want credit for future upside. The result is more contingent value, more paper and more post-close alignment risk. Founders should read those structures carefully. They are often the first sign that the buyer sees strategic promise but not full valuation certainty yet.

Sector buyers kept looking for control points

A fourth pattern is that the 2023 pivot was not limited to software in the narrow sense. Cytek Biosciences acquires Luminex was an all-cash healthcare deal. Next Holdco acquires NextGen Healthcare took a healthcare technology provider private. Global Payments acquires EVO Payments did the same in financial services. Different sectors, same instinct: own the workflow, own the data path, own the point where future software improvement compounds.

That is why 2023 should be read as the seed year of the 2024-2026 market. Buyers were already making decisions based on which assets would matter more once smarter tooling, better automation or more integrated data became central to value creation. The explicit AI label became more common later. The underlying acquisition logic was already visible.

The pattern

What ties these deals together is not that every target was already shipping an AI story. It is that buyers had started to rank businesses by how well they could support one. Platforms with software control, operational data or customer workflows became more attractive. Public ownership became less attractive for assets that needed rebuilding. Structure became a way to price uncertainty around the new upside. That is how a narrative shift enters M&A before it fully enters the press release.

Founder's Lens

Adam J. Graham's read on this year: 2023 is when I started telling founders that the buyer narrative had moved faster than most management teams realised. You did not need to call your company AI-native to benefit, but you did need a credible answer for how your product, data or workflow got stronger in that world. Looking back from 2026, the founders who won were the ones who understood that buyers were no longer just valuing what the business was. They were valuing what it could become once a new layer of software sat on top.

What it means today

The lesson now is not to force an AI label onto a business that does not deserve one. It is to understand which part of your company a buyer will treat as the control point for future automation and value creation. If you want to watch that thesis play out across more live deals, Insider is the clearest place to do it.