A venture firm’s interview with HubSpot’s co-founder argues that the software opportunity is bigger than ever, and for builders it probably is. For the incumbent telling the story, the numbers read differently: slower growth, weaker retention, fewer new customers and a staff cut. The video supports the opportunity, not HubSpot’s place in it.
Table of Contents
SaaS companies are growing more slowly and cutting staff while the case for building software gets louder. On 4 November HubSpot, which expects about $3.7 billion in revenue this year, told investors it sees fourth-quarter revenue growth of 12%, down from 17% in the third quarter, and that net new annual recurring revenue will fall in 2026. Its board had authorized job cuts affecting about 7% of the workforce on 1 October, and after the results Cantor Fitzgerald downgraded the stock, which was then about 46% lower for the year.
That is the backdrop for an a16z interview with HubSpot co-founder and CTO Dharmesh Shah, posted on 10 November under the title “Why HubSpot’s Co-Founder Is Betting Against the SaaSpocalypse.” Shah argues that cheaper software creation widens the market and that startups should build away from the road the AI labs are racing down. He also concedes that some software companies will die. The argument is sound for builders. HubSpot’s own filings, published six days before the video, show what the squeeze looks like inside one incumbent.
A friendly interview on a venture firm’s own channel
The video is a conversation between Shah and a16z partner Joe Schmidt, built around Schmidt’s 27 May essay “Avoiding Death on the Yellow Brick Road.” The essay says the AI labs will win problems whose quality rises with raw model capability, such as code, writing and images, while multi-step, vertical work leaves room for startups. Shah adds two tests of his own. If you can append the word yet to the sentence ChatGPT can’t do that and it still holds, the idea is dangerous. And judge the lab’s whole company, not just its current product.
The relationships matter. a16z’s description says the firm and its affiliates may hold investments in companies discussed, Shah says in the interview that he is an investor in OpenAI, and he is HubSpot’s CTO and the builder of the product he promotes. This is an interested party talking to an interested host, not independent reporting.
The “yet” test meets OpenAI’s developer week
Shah dates the recording to 30 September and says it was OpenAI’s developer day. OpenAI held DevDay on 29 September, so the conversation was taped on one of those two days. The event is a useful test of his rule. OpenAI announced Dots, always-on agents with their own cloud computer and browser that work across ChatGPT, text messages, email and Slack, and said it is testing specialist Dots for specific business roles, according to Business Standard.
Agents built for a business role are the kind of product a founder should check twice against the yet test. That reading is my inference and OpenAI has not made the claim. Schmidt’s essay points the other way and notes that OpenAI and Anthropic have announced joint ventures to configure their models for enterprises, which it reads as proof that a generic AI coworker cannot solve every problem. Both readings can be true at once: labs reaching into vertical work and labs admitting how hard it is.
The one-person market is real but counted loosely
Shah says the United States set a record for new companies in 2025, at close to six million, and that more than 70% of new ones are not expected to hire anyone within two years. Census data support the direction. A Founder Reports summary of Census Bureau figures puts 2025 business applications at 5,671,836, an annual record, of which 1,708,842, about 30%, were flagged high-propensity, meaning likely to become employers.
Three footnotes apply. Applications are filings for tax identification numbers, not companies, and many are side projects or tax vehicles. High-propensity is a statistical flag and carries no forecast that nobody will hire. And the Census Bureau announced that from its January 2026 release, applications tied to internet sales are excluded from that series. The industry split also matters: retail, mostly online sellers, led with 1,166,015 applications at roughly 19% high-propensity, while professional services, closer to the consultants and creators Shah describes, had 764,007 at about a quarter. His figure of more than four million applications by August 2026 is not confirmed by any Census release cited here.
YouSpot proves interest at a one-dollar price
Shah soft-launched YouSpot, a CRM for one-person companies, on 27 August and published a longer note the next day. Go Big News reported that the first 1,000 paid seats cost $1 a month and that 547 were taken by the time of the 28 August post. Early coverage called it a separate product; Martech Notes reported that Shah clarified on 30 August that it is a HubSpot Next project aimed at the low end of the small-business market.
In the interview he calls willingness to pay with a credit card on a consumption model the most encouraging result of any HubSpot Next project he has run, and he says the product is 30 days old. No revenue, conversion or retention figures have been published. At an introductory price of one dollar, early seats measure curiosity at least as much as budget, and Go Big News notes that the price is a cap for a limited cohort and that the product needs inbox access to work well. Go Big News describes itself as software coverage reviewed by AI, which is one reason to rely on Martech Notes for the same price and seat terms.
HubSpot’s own quarter shows the squeeze at the low end
Third-quarter revenue reached $947.6 million, up 17%, and non-GAAP operating margin rose to 22.6%. But fourth-quarter guidance of $945 million to $947 million implies growth of 12%, net revenue retention stood at 101% and is expected to finish 2026 about two points below 2025, and net new annual recurring revenue is now expected to decline. HubSpot added 5,300 net customers in the quarter, down from 7,000 in the second, and expects fewer than 5,000 in the fourth. Chief executive Yamini Rangan said the headwinds seen in the second quarter “continued and became more pronounced in Q3,” and chief financial officer Kathryn Bueker said they would persist into 2027.
The weakness sits where Shah says the new market is. Bueker said net additions are driven disproportionately by Starter and Pro customers, who arrive through an inbound, product-led motion and have become more price sensitive. Management also said inbound demand is under pressure because AI changes how buyers find software. Shah describes YouSpot as built for the low end of small business. This is an inference, but it follows from the filings: the segment he calls an opening is the one where his company reports its softest demand.
The interview was taped a day or two before the board authorized the cuts. HubSpot’s 8-K says the plan eliminates roles for about 7% of staff at a cost of $65 million to $75 million, mostly in the fourth quarter, to build a leaner organization of agile teams. The coverage reviewed does not say what Shah knew when he spoke, and nothing in the video’s description marks that the conversation is older than the news.
The chat window cuts both ways
Shah’s central analogy is that ChatGPT and Claude are the new browsers: the interface is a means to an end, and the value sits in the work and data behind it. HubSpot’s results give that view some support. Management says HubSpot is the most popular CRM connector on both ChatGPT and Claude, and that customers who use connectors show gross revenue retention 13 points higher and upgrade paid seats at three times the rate of those who do not. Those are company figures, and customers who connect their tools are probably the more engaged ones to begin with, so the data show a correlation.
The same call shows the other edge. Inbound leads built HubSpot, and management said they remain under pressure as buyers use AI to find software, while new sources such as connectors and answer-engine optimization are still nascent. The browser era also reshaped who found customers and where software ran. A chat window that hosts HubSpot’s product can also stand between HubSpot and its next buyer.
The evidence backs the opportunity, not the incumbent
The SaaSpocalypse label fails as a verdict on demand for software. The Census series, the new solo-founder products and a venture firm’s thesis all point to more builders and more problems for software to solve. It describes this year’s numbers at an incumbent more accurately: slower growth, softer retention, a staff cut and a stock down by nearly half. HubSpot is still profitable. It reports a 22.6% non-GAAP margin, about $720 million of expected free cash flow and 19% growth in deals above $120,000. Its problem is the price of moving to agents and credits while its entry-level market hesitates.
For a founder, Shah’s advice to start narrow and solve the whole problem fits the a16z essay’s tests, but nothing published shows that one-person businesses pay full price. For a HubSpot buyer, the signal to watch is the EMEA pricing pilot, which lowers the jump from Starter to Pro and which Cantor Fitzgerald, as reported by Investing.com, reads as an effective price cut. For everyone else, the next evidence arrives in February, when HubSpot has said it will report fourth-quarter results and 2027 guidance: net customer additions and net revenue retention will say more than any interview.
Author:
Jan Bielik
CEO & Founder of Webiano Digital & Marketing Agency
This article is an original analysis supported by the sources cited below
Why HubSpot’s Co-Founder Is Betting Against the SaaSpocalypse
The a16z video page, published on 10 November 2026, giving the title, the description with its investment disclaimer and the chapter list the article tests.
Avoiding Death on the Yellow Brick Road | Andreessen Horowitz
Joe Schmidt’s 27 May 2026 essay, which defines the Yellow Brick Road argument, the three tests and the point that labs cannot solve every problem with a generic coworker.
HubSpot Reports Q3 2026 Results
HubSpot’s filed earnings release of 4 November 2026 with 17% revenue growth, 311,704 customers, 22.6% non-GAAP margin and fourth-quarter guidance implying 12% growth.
HubSpot, Inc. Form 8-K (report date October 1, 2026)
The filing showing the board authorized role eliminations affecting about 7% of the workforce on 1 October, with 5 million to 5 million in charges.
HubSpot (HUBS) Q3 2026 Earnings Call Transcript | The Motley Fool
Management’s account of 101% net revenue retention, falling net additions, pressure on inbound demand, connector data and the EMEA pricing pilot.
Business Formation Statistics
The Census Bureau’s program page, which describes the series and announces the January 2026 change excluding internet-sales applications from high-propensity counts.
How Many Businesses Are Started Each Year? – Founder Reports
A secondary summary of Census series giving 5,671,836 applications in 2025, about 30% high-propensity, with industry splits and the caveat that applications are not businesses.
HubSpot co-founder Dharmesh Shah launches YouSpot, a CRM for one-person companies — Go Big News
Launch coverage with the 27 and 28 August dates, the monthly price for the first 1,000 paid seats and 547 seats taken, from a site that says AI reviews its coverage.
HubSpot Next clarifies YouSpot as a company-backed AI CRM experiment – Martech Notes
Independent corroboration of the offer and the 30 August clarification that YouSpot is a HubSpot Next project aimed at the low end of small business.
Cantor Fitzgerald cuts HubSpot stock rating on weak demand trends By Investing.com
Report of the 5 November downgrade, the stock down 46% for the year at 16.64 and the pricing pilot read as an effective price cut.
OpenAI DevDay 2026: Dots agent, GPT-6.1 Sol, new plans and more announced | Tech News – Business Standard
Independent account dating OpenAI’s DevDay to 29 September 2026 and describing Dots agents and specialist Dots for business roles.
| Citing this article? Brief excerpts are welcome. Please credit Webiano.digital, name the author where stated, and include a link to https://webiano.digital and to this original article. Full or substantial republication requires prior written permission. Read our Copyright and Content Use Policy. |
This article was prepared with the assistance of artificial intelligence tools. The content underwent expert human review, and Webiano Digital & Marketing Agency assumes editorial responsibility for its final version and publication.
