For more than a decade, customers spent their software budget procuring vertical SaaS products. ACVs, or annual contract values, were modest, customer acquisition cost had to stay below a ceiling, and the resulting go-to-market playbook was product-led growth, SDR-led and content-driven.

With AI, many products are no longer SaaS but usage and outcomes based. They are replacing labor, not software. At my investment firm, , we call this new category of companies vertical AI. Vertical AI spend doesn’t just come from a customer’s software budget. It often comes out of headcount as well, a much larger line item. As a result, ACVs have jumped meaningfully to 6- and 7-figure deals.

I’ve written before about how AI for vertical SaaS, and how the value framing shifted from subscription pricing to. As ACVs have grown in vertical AI, the go-to-market motion is changing too. We’ve explored tactics to drive a more efficient sales process.

Here, I’ll explore how the channels are changing as well.

Why direct sales is back

Direct sales has historically only worked at true enterprise scale. The cost of an AE’s time wasn’t warranted for smaller ACVs. Below a certain deal size, the math didn’t work for high-touch sales. That’s why SaaS GTM became PLG and SDR-led.

With vertical AI ACVs frequently landing in the 6- or 7-figure range, founders now have room to invest meaningfully in winning each logo. We’re also seeing these smaller businesses spending relatively more with quicker sales cycles which is enabling higher volume.

AEs, in-person sales motion, and other tactics that didn’t pencil at scale under old SaaS economics now do. Direct sales now works further down market where prior SaaS economics didn’t allow it.

Two channels in particular have driven a lot of distribution and success for vertical AI companies recently. They are distinct from each other but we’ve seen companies have success with both.

No. 1: Private equity and heads of AI

Many PE firms are actively pushing their portfolio companies to drive efficiency with AI. Some have even created a new role internally to spearhead these initiatives. These AI partners are often tasked with collecting and disseminating learnings, finding good AI tools, and connecting them into the portfolio if there’s a fit.

The motivation is sometimes EBITDA driven, but can also be softer than that. Many of these execs are focused on adding value across the portfolio, helping companies build AI competency, and coming up with an execution plan.

The decision making structure also varies. Sometimes the and push adoption down to the portfolio. More often, the firm will forward information to relevant company executives and leave the decision making to them. If executed well, this can be a very efficient channel for vertical AI companies. One introduction to the PE firm surfaces many qualified leads across their portfolio companies.

Usually, companies will land one customer initially. Positive feedback then travels in two directions. Laterally to peer companies within the portfolio, and back up to the PE investor, who introduces the vendor to others in the portfolio. We’ve seen this be particularly successful in industries where rollup strategies are popular like healthcare services, dental, MSP, accounting, legal, financial advisory, insurance brokerage, home services and industrial.

No. 2: Conferences

We’ve also seen sector and function specific conferences be incredibly valuable in driving distribution for vertical AI companies. The advantage is concentrated attention and self selection by the right buyer. Buyers are captive and open to learning.

They come to these events curious to hear what’s new in their sector. Attendance allows companies to meet the right buyer, showcase the product live, and collect leads at scale. Sponsoring and attending dinners is another opportunity to meet prospects.

I’d argue that scalability of lead generation and brand awareness matters more now than ever. That requires getting the word out about your own company but also cutting through the noise of others in the market. Buyers are actively building out their AI strategies so vertical AI companies should be sprinting on GTM. Companies need to be top of mind when potential buyers are open to evaluating new tools.

Whether that becomes a sole source decision or an RFP, the prerequisite is being part of the consideration set. In order to do that, your buyer needs to know you exist, and this is a great way to spread the word efficiently.

What this means

The GTM playbook for vertical AI now looks meaningfully different from the SaaS playbook it grew out of. Distribution, pricing and sales motion have all shifted in tandem, with each piece reinforcing the others. Buyer pull justified larger ACVs, larger ACVs justified deeper investment in the sales motion, and the new economics opened up channels that didn’t work under the old model.

The companies pulling away are the ones pairing a great product with the right GTM motion. They have recognized that bigger ACVs demand a different playbook, and they have adapted before their peers.

When the gates of distribution opened, everyone walked through. The companies winning now have figured out what to do once they were inside.

If you’re a founder building vertical AI and rethinking GTM, I’d love to hear from you.

Ìę is a general partner at , where she invests in and partners with early-stage founders from inception through Series A across sectors including AI, fintech, healthcare and enterprise software. Prior to joining Defy, Agarwal spent seven years at and began her investing career at . A former founder and operator, she previously co-founded two startups and started her career at

Related reading:

• New Value Models, New Playbooks: Tactics AI-First Companies Use To Scale

Illustration:

The market believes software businesses can’t charge premiums anymore and it predicts slowing growth indefinitely.

There are two reasons. First, AI powers a 10x decrease in software production costs. Second, these AI capabilities enable a huge wave of new competitors, both VC-funded startups and in-house solutions.

Reduced software production costs and rising competition, they say, will eliminate software’s pricing power. Public software stocks traded down 20% this year through mid-May, and for the first time in history, software trades at a discount to the average S&P500 multiple on earnings. SaaS is dead.

It is true that AI has brought falling costs and rising competition. But it does not follow that SaaS is dead. Lowering the cost to produce software does not mean that software revenue will shrink. In fact, history suggests the opposite.

In certain cases, efficiency begets consumption. This is the lesson of the . It worked for coal engines, it worked for data centers, and it will also work for AI-powered software.

The Jevons Paradox

Let’s start with coal. In 1860s Britain, many worried about burning through the country’s coal resources too quickly. Conventional wisdom said that developing more-efficient coal-burning engines would make the coal last longer.

But economist William Stanley Jevons recognized that more coal-efficient engines would cause an increase in demand for coal energy with the result that Britons would burn through their coal more quickly, not less.

Jevons was right. When greater efficiency produced lower costs, it also unlocked enormous new demand. This consumed coal reserves faster, not slower.

Twenty-five years ago, I joined a buyout firm during the 2001 dot-com crash. My first investment was in a troubled datacenter company. Exodus Communication, which reached a peak market cap of $32 billion, then went through bankruptcy twice as datacenter demand continued to fall.

In 2004, I recommended that our firm acquire that datacenter business out of the second bankruptcy for $200 million and merge it into a competitor named Savvis.

At the time, the market considered datacenters a shrinking industry. Dot-com companies were pulling racks of servers out of the sites, and datacenter floors were emptying out. Industry analysts forecast that given Moore’s law about the exponential growth of chip capacity and increasing server power density, a single rack in 10 years would deliver what it took 100 racks to deliver in 2005, and that in 20 years, a rack would deliver what 10,000 racks delivered in 2005. Conventional wisdom said that more-efficient chips would require less datacenter floorspace over time.

Our thesis that demand for datacenter floorspace would grow was not a popular opinion at the time. If 10,000 racks in 2005 would be replaced by just one rack in 2025, didn’t the U.S. have plenty of datacenter floorspace already?

was running advertisements showing a room full of servers replaced by one mainframe. One skeptical investment committee member told me that this business had been through bankruptcy twice in two years, and that if it went through a third time, I would go with it.

Today, one rack can indeed deliver 20,000x the compute power of racks from 2005, and as everyone knows, far from having too much floorspace, we can’t build new datacenter capacity fast enough. Truly enormous latent demand for computing power was unlocked as rack efficiency increased. The Savvis story ended well too, sold six years later for $3.2 billion.

The Jevons Paradox was true for coal, and it was true for data centers. It will also be true for AI-supported knowledge work.

Knowledge work and market expansion

Twenty-five years ago, only the wealthy had access to personalized investment advice. In 1996, Nobel Laureate Bill Sharpe co-founded to bring personalized investment advice to anyone with a 401(k).

My firm was an investor, and I had the privilege of working closely with the company. At first it tried to sell advice about how to invest 401(k) funds, but only about 20% of employees were interested in taking advice and then managing their 401(k) positions themselves.

Financial Engines’ breakthrough innovation was to manage the 401(k) positions directly, not just advise. Employees could check a box: “do it for me”. The demand from people who previously had no access to this advice was beyond all expectation and did enormous good. I recall that an early customer was , whose tens of thousands of employees with an average age of 27 years had approximately 40% of their 401(k) monies in cash, 40% in stock of JCPenney (which would eventually file for bankruptcy in 2020), and 20% in everything else. Just moving them into sensible low-cost mutual funds appropriate to their age and other financial goals generated huge benefits.

Financial Engines went from zero to $169 billion in assets under management when it was acquired in 2018 for $3 billion.

The company delivered a service that is very similar to what we today would call agentic AI. The customer (an employee with retirement savings) was delegating a decision (invest my money) to a computer system, and the employee paid in a way tied to the outcome (~50 basis points on AUM).

Of course, the technology to deliver this was quite different, and this was a very narrow application. The lesson remains: Software enabled a massive efficiency in delivering knowledge work (in this case individual investment advice) and a huge latent market appeared to buy the service.

The end of the rationing of knowledge work

The increased cost efficiency of AI, like the increased cost efficiency of Financial Engines’ algorithms, allows demand to increase because it relaxes a supply constraint on knowledge work.

Across human history, even to today, knowledge work has always been rationed because it is supply constrained.

Knowledge workers take years of education and training, tend to want to live in high-cost places, over time want to work on only certain kinds of problems they find interesting, and require a lot of management to get along. That is why we pay them such high wages and do everything we can to make them more productive.

Business software is a tool to make knowledge workers more productive. The total business software market in the U.S. is on the order of $0.5 trillion per year, according to Gartner. The U.S. market for knowledge work, that is the amount paid to the 100 million knowledge workers in this country, is roughly $10 trillion, per numbers. Currently, we spend about 5% of the cost of knowledge workers on software tools to help them.

AI enables software companies to not just sell tools to knowledge workers, but to begin to sell the knowledge work outcomes themselves, as I have written about in prior Crunchbase articles.

Put those together: 90% cost compression in software development plus the ability to sell knowledge work. We know there is a huge latent demand for knowledge work, if only it were not so expensive and hard to access.

For the first time, millions of people and businesses who have never had access to a strategist, an analyst, a lawyer or a financial adviser are about to get one.

Software is far from dead. The increase in efficiency offered by AI will allow it to do much more for less, and just like a more efficient coal engine or data center, this will unlock huge latent demand for knowledge work.

Ultimately, this will increase revenue and the strength of software businesses that use AI to further improve knowledge workers’ productivity or deliver knowledge work outcomes directly. Software’s job today is to solve the problem of delivering this safely and reliably.

It was no small task for industry to learn how to manage knowledge workers who are human, and it will be just as big a task to learn how to manage those who are machine knowledge workers. That is the challenge.

But remember that today the market for knowledge work is 20x the size of the market for software. The scale of the prize for software companies is unlocking the latent demand for knowledge work that, if history is any guide, will dwarf today’s software market.

The market today fears that the efficiency delivered by AI will shrink the software industry. Exactly the opposite is true. AI will unlock massive latent demand for knowledge work, and the software market will explode. Long live software.

co-founded in 2014 and is managing partner. He has served on numerous private and public technology company boards, and currently is a director of , , , , and . Previously, he was a partner and member of the investment committee at . He also worked at and . Morse serves on the board of directors of and as member of the advisory board for the HMTF Center for Private Equity Finance at . He attended , graduating summa cum laude with a BSE, and , where he earned his MBA and was an Arjay Miller Scholar. Morse lives in Austin.

Related reading:

• Don’t Just Talk About AI. Measure Business Outputs. Here’s How.
• The Rise Of The AI Executive
• Invention To Innovation: Making Sense Of AI’s Moment

Illustration:

This is a weekly feature that runs down the week’s top 10 announced funding rounds in the U.S. Check out last week’s biggest funding deal roundup here.

Venture funding has always been a world of haves and have nots. And these days, the haves are having more than ever. Case in point this week was . The 5-year-old generative AI giant secured $65 billion in Series H funding this week, pushing its post-money valuation to a mind-blowing $965 billion.

After that, the next-biggest financing was a $1 billion round for AI software development tool maker , lifting its valuation to $26 billion. Companies in a range of other sectors also managed to secure sizable though smaller rounds, in areas including commerce logistics, developer AI, insurtech, fusion and more.

1. , $65B, foundational AI: Generative AI company Anthropic raised $65 billion in a Series H funding round, more than doubling its post-money valuation to a staggering $965 billion. San Francisco-based Anthropic said , , and led the financing, and that , , , , and co-led the investment.

2. , $1B, AI software development: Cognition, developer of AI software engineer Devin, has closed on over $1 billion at a $26 billion valuation. , , and 1Ìęled the financing for the San Francisco-based company.

3. , $250M, logistics: Atlanta-based Stord, developer of a fulfillment network, software and AI tools for independent brands, secured $250 million in Series F funding. The round set a $3 billion valuation for the 11-year-old company.

4. , $113M, AI for developers: OpenRouter, a marketplace for AI models, secured $113 million in Series B funding. led the financing for the New York-based startup.

5. , $106M, insurtech: San Francisco-based Corgi Insurance, developer of an AI-native insurance platform for startups, picked up $106 million in Series B1 funding led by . The financing, which set a $2.6 billion valuation, comes just three weeks after Corgi $160 million in Series B funding at a $1.3 billion valuation.

6. (tied) , $100M, fusion energy: Kearny, New Jersey-based Thea Energy, a developer of technology for fusion energy systems, raised $100 million in Series B funding led by . Thea says the funding will go toward manufacturing infrastructure.

6. (tied) , $100M, healthcare data: Garner Health, a platform for finding healthcare providers, closed on $100 million in Series E funding led by . The financing set a $2.74 billion for the New York-based company.

8. , $90M, space tech: Observable Space, a space tech startup that develops and builds advanced optical systems, says it raised $90 million in Series A funding led by to scale manufacturing and develop its technology. The Santa Monica, California-based company also announced that it secured a $94 million contract with the.

9. , $59M, AI video: Reactor, a San Francisco-based developer platform for real-time generative video, emerged from stealth with $59 million in funding led by .

10. , $52M, cancer detection: San Diego-based ClearNote Health, a developer of early detection and monitoring tests for multiple forms of cancer, picked up $52 million in Series D financing. Founding investor led the round.

Methodology

We tracked the largest announced rounds in the Crunchbase database that were raised by U.S.-based companies for the period of May 23-29. Although most announced rounds are represented in the database, there could be a small time lag as some rounds are reported late in the week.

Illustration:

The global innovation economy remains largely defined by agglomeration dynamics. Worldwide, 19 ecosystems dominate the innovation landscape, increasingly concentrating innovation demand (corporates) and supply (scaleups) — attracting further growth capital (investors).

Meanwhile, other ecosystems struggle to achieve a meaningful presence on the global innovation map and are at serious risk of technological disruption and economic downfall.

Yet something is happening below the surface. Over the past decade, the composition of the Global Innovation Ecosystems Life Cycle Curve changed dramatically, as the number of scaleup ecosystems worldwide has more than doubled.

The trend is not stopping just here: we expect these figures to even triple in the coming years.

In this new scenario, emerging innovation economies hold the potential for disrupting the agglomeration paradigm, toward a new scheme of interconnected networks of specialized local innovation hot spots.

Among them, there is also Africa. While the continent still lacks ecosystems at the most advanced stages of maturity, it now counts four ecosystems at the startup stage and 40 at the standup stage, compared with respectively 25 of those 10 years ago, according to by my organization, , in collaboration with and .

Africa: the awakening giant of the coming decade?

As of today, Africa’s innovation economy includes 883 tech scaleups that have raised a combined $24.7 billion. Despite this progress, the continent still represents only about 1% of global figures.

The African innovation landscape remains highly concentrated around four main hubs: South Africa, Egypt (North-East), Nigeria (West Africa) and Kenya (East Africa). The North-Western corner of the continent still lacks a dominant hub, although Tunisia, Morocco and Algeria remain the leading candidates.

A testbed for clean technologies?

Emerging innovation economies that thrive on the global innovation map typically build on top of highly specialized, unique local strengths.

Our recent analysis has identified clear evidence that Africa holds significant potential over the development of clean energy systems and technologies.

The relative prominence of the cleantech sector in Africa is evident from the data:

• Africa is home to 95 cleantech scaleups, representing roughly 11% of the total scaleup base.
• Collectively, they have attracted approximately one-fifth of all capital deployed to African ventures.
• Cleantech has also generated a disproportionate share of high-growth leaders, accounting for around 20% of both scalers (scaleups that raised more than $100 million) and super scalers ($1 billion-plus).

Within cleantech, a highly specialized vertical is also emerging, what we might call “gridtech”:

• It comprises 16 scaleups (17% of the cleantech total) and two scalers (25% of total).
• It has attracted around 30% of total cleantech funding.
• Africa’s sole cleantech tech giant, Kenya-based , operates within this gridtech vertical.

That said, the numbers still point to a gap.

The elephant in the room

The main challenge is the grid infrastructure deficit, which remains the primary bottleneck to scaling energy system technologies. As shown in the map below, Africa’s grid infrastructure is highly fragmented: High-voltage networks are concentrated in a few densely populated areas, while large parts of the continent remain largely disconnected.

As a result, grid infrastructure development and electrification are key to unlocking Africa’s growth — consider that Africa still accounts for only about 5% of global energy supply — and its innovation potential.

At the same time, the continent holds world-class renewable resources, including approximately 13% of global technical hydropower potential and around 60% of the world’s best solar resources.

Africa’s energy system is expanding, but fully unlocking its economic and innovation potential will depend on accelerating electrification and strengthening grid infrastructure.

Blended finance will be critical to enable this growth. Both private and public capital are required: private capital drives innovation, while public finance enables foundational infrastructure such as grid expansion.

In particular, private capital needs to be complemented by structured public finance initiatives to address the inherent limitations of a relatively small domestic VC market, which remains heavily focused on early-stage investments.

Public capital will be essential for infrastructure development. In gridtech especially, public investors are expected to account for up to about 80% of total investments by 2030, reflecting the capital intensity and risk profile of grid infrastructure.

International capital still dominates the market, with approximately 69% of active investors originating outside Africa, underscoring continued reliance on foreign capital despite growing local participation.

Get the full story in our report:

is chairman of and a professor at . He is a serial entrepreneur who has started three startups in his career, the last of which is , among the five Italian scaleups that have raised the largest amount of capital. He is recognized among the leading international experts in open innovation and has wide experience in setting up and managing open innovation projects — venture clients, venture builders, intrapreneurship, CVCs — with large multinational companies, as well as advising and training on this subject. Onetti has a column on () and several other tech blogs.

Photo by on .

At least not these days. The space and AI pioneer posted a net loss of $4.28 billion in the first quarter of 2026, up more than 700% from a year ago. Revenue, meanwhile, totaled $4.69 billion in Q1, up 15% from a year ago.

As a public company, SpaceX is reportedly seeking a valuation of around $1.5 trillion or more, . It’s aiming to raise up to $80 billion or more in the offering, which would make it the largest IPO in history.

At its target valuation, SpaceX would join a rarified club of just seven U.S. public technology companies with market caps of $1.5 trillion or more. Of those, just five have crossed the $2 trillion mark.

Of course, those companies took time to grow into their 13-digit valuations. But at some point, they too made their first public IPO filings. And they too had revenue.

The similarities end there. For a sense of how SpaceX compares at IPO time to other members of the trillion-plus-club, we took a look at their original S-1s from the 1980s and onward. Here’s what their numbers looked like just before their public market debuts:

: Today, the Silicon Valley chip designer is a $5.3 trillion market cap company. Anyone who invested in its 1999 IPO, needless to say, has done extraordinarily well.

At the time of its market debut, of course, such a trajectory was not obvious. Still, it looked like a solid bet. The company, which then focused on designing 3D graphics processors for the PC market, had $93 million in revenue for the three reported quarters prior to its IPO, growing severalfold year over year. Over the same period, it posted a modest $3.5 million loss.

: Google was already the dominant player in online search when it went public in 2004, with impressive financials to boot. Revenue for the first half of that year totaled $1.35 billion, more than doubling in a year, paired with a $326 million profit.

While that was impressive, so is Google’s ongoing growth. Currently, its market cap is $4.7 trillion and it posts more than $400 billion in annual revenue, with massive profits as well.

: The iconic smartphone and computing giant knows a thing or two about longevity. Apple turned 50 last month, and it went public over 45 years ago, in 1980.

It was an impressive and attention-getting offering for the time, with $118 million in sales and nearly $12 million in profit. It helped that Apple was already a prominent consumer brand at the time due to its popular home computers. These days, its market cap hovers around $4.5 trillion.

: Microsoft went public in 1986, so it’s had some 40 years to grow into its current $3.1 trillion valuation. But even back in the era of big hair and floppy disks, the software giant’s IPO prospectus showed clear signs this would be no ordinary market entrant.

In the year before its IPO, Microsoft had revenue of $140 million and net income of $24 million. That income figure, however, includes stepped-up spending on marketing and R&D. Without those expenses, profit margins looked astoundingly high for a time before software business models were status quo.

: At the time of its public offering in 1997, Amazon was known as an online bookseller, branding itself as “Earth’s Biggest Bookstore.” All the other stuff came later.

Still, it was a compelling offering at the time, with Amazon growing annual sales from zilch to around $16 million in just two-and-half years after its inception. It pitched losses as part of its growth strategy, which called for investing heavily in marketing and promotion, site development and operating infrastructure.

Needless to say, things worked out well, with Amazon currently valued at more than $2.8 trillion.

SpaceX is not like the others

If we look at the most valuable public tech companies, a few commonalities about their earlier days stand out. All went public relatively early in their operating histories and debuted with sharply growing revenue and either profits or losses in the single-digit millions.

SpaceX, founded in 2002, looks by comparison like an oldster for a company on the cusp of a public market debut. It’s also worth pointing out that Google, founded in 1998, is only four years older than SpaceX. That means, it’s had 28 years to grow into becoming a company with over $400 billion in revenue over the past 12 months and $138 billion in operating income.

SpaceX, by contrast, has had 24 years to grow into becoming a company that loses $4.3 billion in a single quarter.

Related Crunchbase query:

Related reading:

• SpaceX IPO At $1.5T Valuation Would Be 10x Larger Than Biggest VC-Backed Listing Of All Time
• The IPO Pipeline Finally Gets Interesting
• PwC’s US IPO Lead On The 2026 Outlook, IPO Timing And The Secondary Boom
• IPOs Are Holding Up In 2026, But SaaS Debuts Aren’t Happening

Illustration:

We’re not talking about confidential filings either. Sure, they signal serious intent and contain valuable information for regulators. But for the rest of us, it’s the public S-1 filing that signifies an IPO is actually imminent.

By this latter measure, the past few weeks have been pretty busy for venture-backed startups. , the designer of speedy AI inference chips, filed publicly last week for an offering expected to raise around $2 billion. The Silicon Valley company, which withdrew plans for an IPO last fall, is reportedly seeking a valuation upwards of $35 billion this time around.

That alone would be enough to set IPO market watchers abuzz. Per Crunchbase data, it stands to be the largest initial share offering of a U.S. semiconductor company to date.

However, Cerebras wasn’t the only venture-backed company seeking a multibillion-dollar IPO valuation.

Power players

Another, albeit smaller, contender is nuclear power startup , which is making its debut today. The Rockville, Maryland-based company priced shares at $23 each late Thursday, above the projected range, raising around $1 billion. Shares closed up 27% in first-day trading Friday.

Meanwhile, on the geothermal power front, is also looking to take its clean energy ambitions to the public market. The Houston-based company filed last week for a offering that could bring in around $250 million.

Biotech IPOs heating up

Biotech is also heating up. Last week delivered a big debut from , a Waltham, Massachusetts-based developer of oral and injectable treatments for obesity and metabolic disease that $718 million in its Nasdaq offering. , a Fremont, California-based startup applying proteomics to early disease detection, made its market entry as well, securing a current market cap around $1.6 billion.

More biotech debuts are on deck too. Austin-based , a venture-backed developer of a nerve stimulation device for stroke survivors, filed last week for an offering. The prior week brought S-1 filings from Boston’s , a developer of medicines for depression, anxiety and other neuropsychiatric disorders, and , a Denmark-based biotech which focuses on treatment of blood coagulation disorders.

Space and defense on the rise

Of course, everyone knows the Texas-based company on deck to publicly file for a space tech offering of unprecedented magnitude. for an IPO a few weeks ago, with media reports pegging its target valuation around $1.75 trillion. If the company forges ahead with reported plans for a June market debut, a public filing should follow in the next few weeks.

In the interim, another, much, much smaller offering in the defense tech space is on track to hit the market much sooner. , a Herndon, Virginia-based developer of radio frequency intelligence for military customers, filed earlier this month for a offering. It comes amid a period of heightened investor appetite for defense tech, with an expectation of more debuts in the space likely in coming months.

Now we just need some software

Of course, it’s not an IPO market that is welcoming to all venture-backed startup sectors. One area noticeably absent from the impending offering list is enterprise software. While SaaS has long been a mainstay of the IPO pipeline, the sector has taken a hit of late amid investors’ concerns of AI disruption.

That said, it’s still encouraging to see a swathe of other sectors dipping a toe in IPO waters.

Related Crunchbase queries:

Related reading:

• After Swarmer’s Soaring Debut, Here Are 12 Other Potential Defense Tech IPOs
• IPOs Are Holding Up In 2026, But SaaS Debuts Aren’t Happening

Illustration:

Look closer, and the story changes completely. Early-stage funding was up 41% year over year. AI/ML deal count , up from roughly 5,600 the year before. More companies are getting funded at the early stage, not fewer.Ìę The concentration at the top? That’s an infrastructure play. The application layer looks entirely different.

Build vertical, not horizontal

The real signal is in the shift from horizontal to vertical. shows horizontal SaaS down 35% over the past 12 months while vertical SaaS is essentially flat (up 3%). That divergence matters for founders deciding what to build.

Horizontal software (project management, general productivity, collaboration) is commoditizing fast as AI agents handle coordination natively. But vertical software? That’s where proprietary data shines and industry-specific compliance workflows matter. AI makes the first category less valuable and the second category more valuable.

If you’re starting a company right now, the data says: Pick an industry, not a feature. Claims processing in insurance, scheduling in healthcare, compliance in financial services, job costing in construction. These are workflows where software penetration has been shallow for decades because the problems were too specific for horizontal tools. AI changes that math.

Build for the $6T, not the $500B

The addressable market for software is expanding, not contracting. In Redpoint’s CIO survey, 58% cite AI as the top driver of increased software spend. As agents move from copilot features into autonomous workflow execution, the addressable market grows from roughly $0.5 trillion in current U.S. enterprise software spend toward $6 trillion or more, because AI starts capturing portions of the knowledge-worker payroll that software never could.

This is classic : When a resource gets dramatically cheaper to produce, consumption goes up. AI makes software dramatically cheaper to build, deploy and maintain. Suddenly, job costing for midsize contractors pencils out. Inventory optimization for independent pharmacies becomes economically viable. The cottage industries that enterprise software ignored for decades? They’re all in play now.

Build for acquirability, not just IPO optionality

But let’s talk about exits, because that’s where the rubber meets the road. The IPO market remains largely closed. In 2025, roughly 2,300 VC-backed startups were acquired compared to just 65 IPOs, per Crunchbase data.

LPs have seen nearly $200 billion in cumulative negative net cash flows since 2022. The pressure to return capital through M&A is real and growing.

Smart founders are building for this reality from day one. They’re building products that integrate into existing enterprise stacks, accumulating proprietary data that makes them expensive to replicate and cheap to integrate. Strategic acquirers in insurance, healthcare, logistics and financial services are actively buying vertical software companies. Why? Because these buyers can’t build this stuff internally — they lack the talent, the focus, and frankly, the DNA.

Start with the workflow, not the technology

So while everyone’s mesmerized by the infrastructure megarounds, the real opportunity is staring us in the face. Pick a specific industry workflow that’s still manual or stitched together with Excel. Build the AI-native solution that actually works for that vertical. Get to revenue before the market catches up.

The record quarter and the shrinking fund base are telling the same story from different angles. Infrastructure capital is concentrating at the top while the application layer is wide open for those willing to roll up their sleeves and solve real problems for real industries. That’s where I’m putting capital, and that’s where smart founders should focus their energy.

As the co-founder and managing partner of , is committed to establishing MGV as the premier venture firm for world-class tech entrepreneurs to accelerate their visions. Under Schröder’s stewardship, MGV has swiftly ascended to a top-quartile firm, surpassing the performance of 95% of venture funds. The performance of MGV is driven by Schröder’s unique approach to venture investing — that providing intensive sales training, devising robust fundraising strategies and securing follow-on investments is the best way to support founders and drive the deepest return for investors. has recognized him as one of the Top 100 global seed investors, and his perspectives are published regularly in Crunchbase News and other leading publications.

Related reading:

• These 3 Charts Show How Venture Capital Has Concentrated At The Top In 2026
• Q1 2026 Shatters Venture Funding Records As AI Boom Pushes Startup Investment To $300B

Illustration:

led the financing, which included participation from , , and . It brings Boulder, Colorado-based Schematic’s total funding since its 2023 inception to $12 million.

Schematic builds entitlements and enforcement infrastructure for SaaS and AI companies. Put more simply, it serves as a digital gatekeeper for software and AI companies. For example, if a company’s sales team wants to give a major client a special discount or extra storage, they have to ask an engineer to go in and “move the walls.” The process can be slow, expensive and tedious.

That’s where Schematic comes in. It essentially acts like a universal remote control for a company’s features.

Instead of burying those rules in the code, a company can plug Schematic into its product. Then, if it, for example, wants to launch a new “AI Tier” or change how many users a client can have, a person in marketing or sales can flip a switch in a simple dashboard.

“When a software company sells you a plan, something inside their product has to enforce what you can do and access based on what you paid for,” said CEO and co-founder . “Most companies build that enforcement infra themselves, often badly, and it becomes the thing that slows down every future monetization change. Schematic is the infrastructure that handles it, so engineering doesn’t have to.”

In addition to the fundraise, Schematic is also announcing that payment giant has tapped it “to solve entitlements as a first-class primitive: decoupled from code, enforced at runtime, on top of Stripe Billing.”

Schematic will be launching its new Stripe app publicly on stage next week at Stripe Sessions.

Systems like Stripe currently handle the money, sending invoices and charging credit cards. But Stripe doesn’t actually sit inside the app to block or allow a user from clicking a button. Schematic claims it will now serve as the “muscle” that actually enforces the rules that a platform like Stripe sets.

‘An emergent crisis’

By using Schematic, Glover said that companies like went from taking weeks to change their pricing to just 10 minutes. The startup’s other customers include , and .

AI has made entitlements an emergent crisis, in Glover’s view.

“Neither underlying costs nor customer value are predictable, and both accrue at runtime,” he told Crunchbase News. “This is why we describe what we’re building as runtime monetization infrastructure: Value is now accruing nondeterministically at runtime, and as a result, pricing and packaging have to be enforced at runtime. A shadow enforcement system catching webhooks from a billing platform cannot support this inflection.”

, general partner at S3 Ventures, said his firm was drawn to invest in Schematic for a few reasons.

“As operators and through our portfolio companies, we’ve seen firsthand how often pricing changes get delayed or deprioritized because entitlement logic is buried in application code. On top of that, AI is accelerating a structural shift away from seat-based pricing; hybrid and consumption-based models now represent 38% of SaaS companies and that number is rising as companies hone their AI pricing strategies, putting real pressure on legacy monetization architectures,” he wrote via email. “Finally, Fynn, Ben, and Gio have worked together for nearly a decade, and each of them encountered this specific problem while running pricing and packaging at growth-stage SaaS companies.”

Fintech startups have benefited from increased investment in recent quarters. Total global funding to VC-backed financial technology startups totaled $53.8 billion in 2025, per Crunchbase . That’s a more than 29% increase from 2024’s total of $41.6 billion raised.

Related Crunchbase query:

Related reading:

• Fintech Startups Globally Raise More Money In Far Fewer Deals In Q1 2026

Illustration:

His company raised a few million in seed, and lifted the curtain, and 
 crickets. Nothing happened. Businesses didn’t sign up. My friend was confused while prospective clients hesitated. This simply didn’t sit well with them.

Then the founder decided to do something odd. He charged money on top of the original monetization plan. Same product, same value proposition, but now there is a monthly subscription. Almost overnight, new businesses began signing up.

Today, that startup is worth billions.

This highlights a counterintuitive truth in strategy: In real-world markets, free or lower prices don’t always drive demand. Frequently, they achieve the opposite.

Higher prices amplify perceived value

The price-quality heuristic is a cornerstone of behavioral economics. When buyers lack complete transparency, they use price as a shortcut for quality. This is why identical items, from fine wines to electronics, are rated higher when they cost more. In B2B, this effect is amplified: A cybersecurity solution priced far below the market doesn’t look like a bargain; it looks like a risk.

Pricing dictates customer behavior and expectations

Low entry points tend to attract price-sensitive users who optimize for cost over outcomes. These cohorts are often more prone to churn and demand excessive support. Conversely, premium pricing attracts partners who value reliability and performance. Opting for higher pricing means going after clients with a different mindset. Even in strategic advisory, I see premium pricing as a filter for commitment.

Your price defines your competitive landscape

Pricing at the bottom floor frames the company within a commoditized segment where differentiation is minimal. Pricing at a premium forces a higher standard of depth, service and trust. Price defines who you are competing against and how you will be compared to them.

is a strategic adviser to tech companies and investors, specializing in strategy, growth and M&A, a guest contributor to Crunchbase News, and a seasoned lecturer. Learn more about his advisory services, lectures and courses at . for further insights and discussions.

Illustration:

“I recognized both the massive potential AI brought to the tax world, as well as the risks to firms and clients by making mistakes and hallucinations,” he told Crunchbase News.

The accounting industry has historically been slow to adopt new technologies. As of today, the majority of small to mid-sized accounting firms — which make up 90% of the market — remain stuck in a cycle of manual data entry.

Addressing both the opportunities — and risks — that came with advances in AI, Haase started building , a tax prep automation startup, on the side in 2023. Rather than targeting the self-prep market, like does, or the mega-enterprise firms that can afford $15,000-per-return software, Juno was built for the underserved SMB accounting firm.

“We continuously ‘dog fed’ the early Juno prototypes into the firm to see what worked best, what slowed things down, and to make it the most efficient tax preparation platform as possible,” Haase said.

It took about a year and a half just to build integrations. “We had to do a bunch of hacky things to be able to work with the existing tax software,” he explained, “because your typical tax software is actually around 15 to 20 years old and they don’t have public APIs.”

By 2024, Juno had launched a co-pilot. Then, in July 2025, it had a tax product. The startup began onboarding other tax firms, growing to nearly 500 customers over the past year. Last year, Haase sold his accounting firm to focus on growing Juno full-time.

Today, he’s announcing that San Diego-based Juno has raised $12 million in a seed funding round led by , including participation from and .

AI to help humans ‘be the advisers they were trained to be’

What makes Juno different from others in the market, Haase believes, is that it operates on the premise that, at least for the foreseeable future, human tax preparers should be the ones driving the tax-return preparation process.

“A business or high-net-worth tax return requires hundreds of calculations, edge cases, deductions and more,” said Haase, who holds an MBA from . “AI simply can’t do that with the 100% accuracy required not to get audited or charged with tax fraud.”

Describing much of the manual work that most accountants must perform to complete returns as extremely tedious, Haase acknowledges that it’s also very easy for accountants to make mistakes that could prove very costly.

“In school, if you get a 93, an A, you get all the credits,” he said. “But on a tax return, if you have a 99%, you fail, and your client could pay the price in penalties.”

In a nutshell, Juno acts as the bridge between a client’s raw documents and the accountant’s filing software. It performs tasks like pulling data from IRS forms and even unstructured documents, such as business financial statements. Overall, it automates 90% of data entry across more than 90 document types while also flagging prior-year changes and inconsistencies for human validation.

The result is that a process that typically takes a human two to three hours is shrunk down to seven to 10 minutes, Haase estimates.

“We do 95% of a tax return in minutes, leaving the accountant to handle the strategic human decisions — the parts that actually save the client money,” he said.

While he declined to reveal hard revenue figures, Haase said that in just eight months, Juno grew to mid-seven-figure annual recurring revenue.

The startup sells on a per-return basis, starting around $45, dropping to the low $30s for high-volume firms.

‘s recent move into consumer taxes and OpenAI’s hiring of a tax director show that the bigger players are eyeing the tax market. But Haase doesn’t feel threatened.

“High-wealth individuals want assurance. If you’re paying $40,000 in taxes, you don’t want to ‘cross your fingers with a chatbot,” he said. “You want a human to talk to, someone who understands the context of your life.”

Juno isn’t trying to replace accountants, he added.

“It’s trying to rescue them from the data-entry basement so they can actually be the advisers they were trained to be,” Haase said.

The startup plans to roll out business returns soon, a move that Haase expects will significantly scale its customer base.

‘A huge, obvious pain point’

, co-founder and managing director of Bonfire Ventures, said he was drawn to invest in Juno because he believes the company is going after “a huge, obvious pain point in a category that hasn’t been meaningfully modernized in a long time.”

“The workflow pain is real, the labor dynamics make the timing right, and Dave brought exactly the kind of founder-market fit you hope to see,” Andelman told Crunchbase News via email. “He lived this problem before he built the company. That always matters.”

The investor believes that tax prep is a category where trust is crucial to product success.

“If you’re going to bring AI into that workflow, it has to be transparent, auditable, and built with a human in the loop,” Andelman added. “That’s what Juno understood early, and I think that’s a big part of why the product is resonating.”

Fintech startups, particularly those that apply AI to traditionally manual or burdensome processes, have benefited from increased investment in recent quarters. Total global funding to VC-backed financial technology startups totaled $53.8 billion in 2025, per Crunchbase . That’s a more than 29% increase from 2024’s total of $41.6 billion raised.

Related Crunchbase query:

Illustration: