AI is making software easy to copy and physical consumer businesses more valuable. Here's why we think the next decade belongs to brands that code can't replicate.

Download the full report PDF · 13 pages · 204 KB

For twenty years the smartest money in the world chased software. It made sense. Software was hard to build, and once you built it, customers stayed and the margins were huge.

AI just changed that math. Code is getting cheap to write and easy to copy. A frozen pasta, a vet visit, a dog chew and a brand people trust are not.

That is the bet we are making at Cutting Horse. Six things are happening at once, and every one of them pushes value toward physical consumer businesses.

1. Twenty years of capital chased software

In 2025 software took 52% of US venture dollars. Consumer goods and services took 2.8%. For every dollar that went into a consumer company, almost $19 went into software. In 2018 it was $1.50.1

Software earned that premium. It was predictable, customers stuck around, margins were high, and it cost a lot to build. That last part is the one AI is breaking.

Share of US venture dollars, %
SoftwareConsumer goods and services
2017: Software 33%; consumer goods and services 13%
2018: Software 30%; consumer goods and services 21%
2019: Software 34%; consumer goods and services 12%
2020: Software 33%; consumer goods and services 9%
2021: Software 40%; consumer goods and services 8%
2022: Software 39%; consumer goods and services 8%
2023: Software 40%; consumer goods and services 6%
2024: Software 46%; consumer goods and services 5%
2025: Software 52%; consumer goods and services 3%
201720182019202020212022202320242025

As of December 31, 2025. Source: NVCA 2025 and 2026 Yearbooks (PitchBook data). Shares derived from sector totals; software includes AI model companies.

2. AI is collapsing the cost of building software

About 75% of Google's new code is now written by AI.2 The price of AI output has dropped more than 280x in two years.11 When anyone can build it, what you built is worth less.

The market already sees it. The median public SaaS company traded at 19.8x revenue in 2021. Today it trades at 4.7x.4

Not all software goes to zero. Proprietary data and regulated systems will hold up. But the easy moat is gone.

3. The money has to go somewhere

Nearly $4 trillion of private capital is sitting undeployed. Funds have about five years to invest it or lose it. It cannot sit still.7

Venture is pouring into AI. Buyout money cannot follow, because buyout borrows against cash flow and most AI companies do not have any. So it is moving. Global buyout value in retail rose 197% in 2025 and consumer products rose 96%, against 30% for tech.8 Tech's share of US platform buyout value fell from 30% to 13%.9 Bain says firms are rotating toward businesses with "physical or labor-intensive components."23

Those buyers are who our companies sell to.

Change in global buyout deal value, 2025 vs 2024
Retail
+197%
Healthcare
+115%
Consumer products
+96%
Services
+91%
All buyouts
+44%
Technology
+30%

Source: Bain Global PE Report 2026 (Dealogic data). Tech buyout value then fell 70% from Q4 2025 to Q1 2026 (Bain).

4. You can't prompt a frozen pasta into existence

Consumer spending is 67.6% of the US economy.10 In 2008 and 2009 it fell 2.3% while business investment fell 16.6%.10 People keep eating, keep taking care of their dogs, and keep buying what they trust.

Real growth in recession years, %
Consumer spendingBusiness fixed investment
Consumer spending and business investment in recession yearsAnnual real growth. Positive values rise above zero and negative values fall below it. 1975: consumer spending +2.3%, business fixed investment −9.0%. 1982: consumer spending +1.5%, business fixed investment −2.9%. 1991: consumer spending +0.2%, business fixed investment −3.9%. 2002: consumer spending +2.5%, business fixed investment −6.9%. 2009: consumer spending −1.3%, business fixed investment −14.5%. 2020: consumer spending −2.5%, business fixed investment −4.6%.+2.3−9.01975+1.5−2.91982+0.2−3.91991+2.5−6.92002−1.3−14.52009−2.5−4.62020

Scroll to see all six years →

Source: BEA via FRED, annual real growth; Cutting Horse calculations. Data through 2025. These are calendar-year changes; the paragraph above uses cumulative changes from late 2007 to mid-2009.

Physical things also do not get cheaper the way code does. In the year to September 2025, prices for medical care services rose 3.9% and food away from home rose 3.7%, while core goods rose 1.5%.24

Brands hold up too. In blind taste tests people show no preference between Coke and Pepsi. Put the label on the cup and about 75% pick Coke.13 AI does not change that.

5. AI is the small brand's new back office

This is the part most people miss. AI does not just spare consumer companies. It makes them better.

McKinsey estimates digital and AI transformation can add 7 to 16 points of EBITDA margin for consumer companies.15 By our math, a $10 million brand used to need about $2 million a year of specialists in finance, analytics, content and customer service to compete. With AI, that looks more like $280,000.15

That is why insurgent brands are taking share from the giants. They get big-company capability without big-company headcount.

6. The entry point is private, and the buyers are waiting at the exit

The median private consumer deal priced at 9.2x EBITDA in 2025, a ten-year low.18 Public consumer staples trade at 23.7x forward earnings, above tech at 17.3x.25 The discount is in private, smaller companies.

At the exit, strategic buyers have paid 3 to 5x revenue for insurgent brands.20 They need growth, and they would rather buy it than build it.

Enterprise value / revenue at announcement, select strategic deals
Alani Nu (Celsius) ~
3.0x
Simple Mills (Flowers)
3.3x
Rhode (e.l.f.)
3.8x
Poppi (PepsiCo) ~
4.0x
LesserEvil (Hershey) ~
4.6x
Bachan's (Marzetti)
4.6x
Touchland (Church & Dwight)
5.4x

Announced 2025–2026. Upfront consideration; most deals carry earnouts. ~ marks reported prices or estimated revenue. Sources: company releases and SEC filings; trade press; Cutting Horse arithmetic.

Consumer did not fail in 2021. The venture model did.

In 2021 a lot of consumer companies were priced like software and bought their growth on paid social. US venture into DTC consumer products went from more than $5 billion in 2021 to about $130 million in 2023.26 When the money stopped, so did the growth.

The brands that came through it did the unglamorous things. They paid rational prices, grew profitably, and built something customers came back for. Same category, different playbook.

Two approaches to investing in consumer businesses
Consideration2015–21 venture playbookRational consumer playbook
Entry pricePriced off software compsPriced backward from where the business will trade
GrowthTriple-digit growth bought with paid socialSteady growth across retail, wholesale and online, funded by gross margin
Profit"Later"A path to positive EBITDA; cash flow is the scorecard
ExitIPO at peak multiplesA mid-market PE buyer as the base case, a strategic as the upside

Software's golden age was built on code being scarce. Code is not scarce anymore.

Consumer's golden age will be built on products, places and brands people trust. That is what we invest in at Cutting Horse: growing, profitable consumer brands, early. If you are building one, we want to talk.

Sources and notes
  1. NVCA 2025 and 2026 Yearbooks (PitchBook data), as of December 31, 2025. Shares derived from sector totals.
  2. Alphabet executive statements (October 2024 to April 2026); Anthropic (June 2026). Figures as stated by each company.
  3. a16z, "LLMflation" (2024); Epoch AI.
  4. Clouded Judgement weekly medians, EV / next-twelve-months revenue, as of November 13, 2026.
  5. Retool 2026 Build vs Buy Report (February 17, 2026; vendor survey, n=817).
  6. Gartner via CIO Dive (July 1, 2026).
  7. PitchBook (December 31, 2025); S&P Global / Preqin (June 30, 2025); Bain Global PE Report 2026. Definitions overlap.
  8. Bain Global PE Report 2026 (Dealogic data).
  9. PitchBook 2026 US PE Outlook Midyear Update.
  10. BEA via FRED, data through 2025; Cutting Horse calculations. Consumer spending growth · Business investment growth
  11. Stanford HAI AI Index 2025: $20.00 per million tokens (November 2022) to $0.07 (October 2024).
  12. BLS Monthly Labor Review (2014), CPI-U December to December.
  13. McClure et al., Neuron (2004).
  14. Wood, Quinn and Kashy, Journal of Personality and Social Psychology (2002).
  15. McKinsey, "Fortune or fiction?" (2024). Back-office figures: BLS median wages (May 2025) with benefits, Clio 2025 legal rates and AI vendor list prices (November 2026); illustrative Cutting Horse arithmetic that excludes a CX manager, controller and agencies.
  16. Bain, 2026 US Insurgent Brands; company reports via trade press.
  17. BCG, "Processed and pressured: CPG's lost decade of R&D" (June 11, 2026).
  18. Capstone Partners Annual Consumer M&A Report (April 27, 2026).
  19. stockanalysis.com, trailing twelve months as of November 19, 2026. Private figures are control-deal multiples, so the public premium is understated.
  20. Company press releases and SEC filings; Axios, FT and WSJ via trade press; Capstone Partners. Cutting Horse arithmetic.
  21. CNBC; company filings; stockanalysis.com (market values August to November 2026). Peak is the highest public or private value.
  22. Company filings and releases; Kellogg. Public company examples only; these are not Cutting Horse investments or returns.
  23. Bain, Private Equity Midyear Report (June 2026), on businesses with "physical or labor-intensive components."
  24. BLS CPI release, October 24, 2025. Price changes over the 12 months to September 2025.
  25. FactSet Earnings Insight, November 6, 2026. Forward earnings multiples for consumer staples and information technology.
  26. Crunchbase News, "VCs No Longer Do DTC" (November 2, 2023). The 2023 figure is year to date at publication.

Figures are as of November 19, 2026 unless noted. Derived shares and multiples are Cutting Horse arithmetic on source figures. This article is for informational purposes only. It is not investment advice and not an offer to sell or a solicitation of an offer to buy any security, including any interest in a Cutting Horse fund. Forward-looking statements are opinions and may not be realized. Third-party data has not been independently verified.