Mallard Bay founder says AI shifts niche SaaS toward payroll budgets

1. Adoption without vibe-coding — ▶ 3:54

   Outfitters are adopting technology quickly, though Morrow says none are close to building their own workflows, a risk he associates with desk-heavy markets such as medical. In rough polling of about 500 active customers, roughly 30% or fewer use ChatGPT or Claude beyond graphics and search. Mallard Bay's in-app agent, Suzy, is often a first chatbot encounter. Morrow does not currently treat customers building their own tools as a risk, but outfitters have begun requesting Claude MCP integration.

2. Consumer AI primes SMBs — ▶ 7:42

   Nick argues LLMs are consumer-first, unlike enterprise-first SaaS. Field-service owners who disliked sitting at a software interface have had a lightbulb moment through consumer AI, creating latent demand among SMBs that were previously cheap about software. Many then hit the limit of their own time and want help putting AI to work. Morrow adds that model labs' advertising, such as Claude ads during NFL games, warms outfitters up for an in-app agent.

3. Moat beyond software — ▶ 11:32

   Morrow concedes lower technical barriers invite point-solution rivals, such as upstarts serving only Montana fly-fishing operators. He sees three defenses: an all-in-one suite (CRM, payments, bespoke reporting) that is hard to match; 5 years of relationships with guide associations, brands and trade shows; and network effects, including referrals between outfitters and Stripe Capital. Nick adds that investors avoid businesses where the database or UI is the core defense, and that SMBs need a familiar working system while they learn to trust AI.

4. Workflows impossible before LLMs — ▶ 15:53

   Morrow names three. Onboarding, once a 2-3 week process that stalled many outfitters, now runs through Suzy, which builds listings from a website, interview or uploaded documents, and some operators are live in 15 minutes. Reporting no longer requires endless standard templates for state and Bureau of Land Management filings, because outfitters describe what they need in chat. Sales follow-up, once a leased human team, is now handled by agents; the alternative would have been hiring more people.

5. Payroll pricing, still unproven — ▶ 18:40

   Outfitters have judged Mallard Bay against the 7 or 8 tools it consolidates. The Alabama marina asked whether it could cut five full-time phone staff at $60,000 a year each to two. Mallard Bay now tracks token usage and is weighing credits on top of subscriptions, but has not rolled out new pricing. Nick floats a 5-10x willingness-to-pay gap in some cases. Luke calls the shift from software to labor budgets plausible, worth trillions rather than billions, but says founders must prove it, and he has seen few who have.

6. Investors split on niches — ▶ 22:11

   Luke finds investors more accepting of historically small TAMs. Nick disagrees: the Series A bar has risen, and he sees vibes-driven fear of missing outsized outcomes rather than math. Morrow's experience supports Luke. Founders of vertical SaaS for batting cages and funeral homes, based away from the coasts, are getting growth equity conversations. He cites 57 million people who hunted or fished last year, more than golfers and pickleball players combined. Nick calls the market bifurcated, with some funds chasing $100 billion-plus outcomes that exclude most vertical businesses.

7. The 3-10M ARR gap — ▶ 27:00

   Luke recalls that in 2019 a Series A needed $1 million of ARR growing above 100%, while Series B required close to $10 million, leaving $3-10 million as no man's land that growth equity has filled. His example fund model puts in about $5 million for roughly 20%, then moves to majority deals at $50-60 million enterprise value. Nick worries these investors may behave like PE, seeking 30% cheaply and deterring VC follow-on. Luke notes a bear case: PE is reportedly down 30-something percent year over year. Morrow finds $100 million exit underwriting appealing.

8. Trust as distribution — ▶ 37:07

   In offline markets, trust is the variable to solve. After cold calls and trade shows, Mallard Bay hired a go-to-market leader in 2024, formerly first sales hire at TechMetrics, and now spends about 90% of performance marketing on Facebook. The Montana Outfitters and Guides Association supplies warm introductions. Morrow's advice is to find whom the customer already trusts and give them incentives to introduce you. He also recommends connected TV, targeted from uploaded lead lists for about $50 a day, to create demand, though it needs TV-quality creative and attribution is harder.

9. Endemic founders win — ▶ 43:17

   Nick's example is FareHarbor: a founder who lived in tourism hotspots for 3 months signing guides one by one, then sent young salespeople to do the same for 90-120 days. Participants cited $110,000 raised and a $250 million figure, apparently the exit to Booking.com, though the transcript is ambiguous. Morrow says Mallard Bay's team, with ranch, duck-lodge and fishing-charter backgrounds, follows a similar playbook, including free-crawfish events. FareHarbor's later layer, listing supply across about 200 OTAs, is the direction Mallard Bay hopes to take next.

10. Search changes, taste persists — ▶ 47:48

    Morrow cites a report that the average search query has grown from 5 or 6 words to 26, as travelers describe group size, dates and lodging in full sentences. Websites must be optimized for answer engines, and social captions must match site copy. Yet taste remains scarce. His marketing bots do about 80% of the work in a 10-80-10 split, with the final 10% needing human judgment. Even outfitters new to AI call out AI-generated images. Nick adds that AI reuses what others have already said, making curation more valuable.

11. Agent booking timelines diverge — ▶ 55:58

    A host cites about 10% of US consumers (range 8-14%) as comfortable letting AI book autonomously. Morrow expects roughly 2 years, once people trust personal agents. Luke thinks 50% of 30-55-year-olds could get there in 3 years. Nick argues for 2030 or later, calling behavior change and trust the bigger problems than product failures, and noting a report that 4% of Americans pay for AI. Morrow later retreats to about 10 years, pointing to an education gap and a younger generation that may pull older users along.

The size of a niche market depends on which budget a product addresses. Joel Morrow's example is a large Alabama marina that told him it pays five full-time people $60,000 a year each to answer phones, and asked whether Mallard Bay could cut that to two. Set against software plans of roughly $400 a month, that is a different calculation. Morrow's company serves about 600 outfitters, a market many investors would call too small. His argument, which organizes the rest of the conversation, is that AI agents change the denominator. When software does the work rather than organizing it, the relevant comparison is payroll, not the tools it replaces.

The product evidence is concrete, if self-reported. Onboarding used to take 2-3 weeks and often an account manager. Mallard Bay's agent, Suzy, now builds listings from a customer's website, an interview or uploaded documents, and some operators are running in 15 minutes. Reporting once demanded endless standard templates for state and Bureau of Land Management filings. Now an outfitter describes what is needed in chat and can save the report. Sales follow-up, previously sold as a leased human team, is handled by agents. In that case the alternative was hiring more headcount.

On the demand side, consumer AI appears to be doing the market education. Nick argued that LLMs are consumer-first, unlike enterprise-first SaaS, so owners who were indifferent to software have already played with ChatGPT and then hit the limit of their own time. Morrow adds that outfitters watching NFL games see Claude ads, which warms them for an in-app agent. His rough polling of about 500 active customers suggests the real base is narrower, with roughly 30% or fewer using ChatGPT or Claude productively. That gap is both the opportunity and the caveat. Latent demand is not realized demand.

Pricing is where the thesis remains unproven. Mallard Bay still sells subscriptions. It now tracks token usage and is weighing a credit system on top, and Morrow admits he has not worked out how to explain usage-based pricing to outfitters, who know it mainly as a cut of payments. Nick floated a gap in willingness to pay of 5x or 10x in some cases. Luke called the move from software to labor budgets the obvious answer, with trillions rather than billions at stake, but stressed that founders must prove it. He has seen only a few who did.

If customers will pay for outcomes, the next question is who keeps them. Morrow concedes that lower technical barriers invite point-solution rivals, such as upstarts focused only on Montana fly-fishing operators, and he expects some to reach feature parity on calendaring or resource management. His defense has three layers: an all-in-one suite combining CRM, payments and bespoke reporting; 5 years of relationships with guide associations, brands and trade shows; and network effects such as referrals between outfitters. Nick agreed that the database or UI is no longer a durable moat, and added that SMBs need a familiar working system while they learn to trust AI with more tasks.

In offline markets, distribution follows trust. Morrow and his co-founder were 20-year-old LSU students when they began with cold calls and trade shows. In 2024 they hired a go-to-market leader who had been the first sales hire at TechMetrics. Performance marketing, about 90% of it on Facebook, now sits alongside association partnerships, with the Montana Outfitters and Guides Association supplying warm introductions. His advice is to find whom the end customer already trusts and give them a reason to introduce you. He also recommends connected TV, targeted from uploaded lead lists for about $50 a day, noting it needs TV-quality creative and makes attribution harder.

The precedent the panel cited is FareHarbor, which Nick described as built by a founder who lived in tourism hotspots for 3 months signing guides one by one, then sent young salespeople to repeat it for 90-120 days. Participants cited $110,000 raised and a $250 million figure, apparently the exit to Booking.com, though the transcript is ambiguous. Morrow sees the same logic in his own team's ranch, duck-lodge and fishing-charter backgrounds. In a small market where everyone knows everyone, being endemic is an asset a Silicon Valley sales rep cannot copy. FareHarbor's next layer, distribution across about 200 OTAs, is where Mallard Bay hopes to go.

AI is also changing how end buyers find outfitters, with a twist. Morrow cited a report that the average search query has grown from 5 or 6 words to 26, as travelers describe group size, dates and lodging needs in full sentences. Websites must be tuned for answer engines, and social captions should match site copy. Yet the same tools make taste scarcer. Morrow's marketing bots do about 80% of the work, with the first 10% setup and the last 10% human judgment. Even outfitters new to AI spot AI-generated images, a real problem in a photography-driven industry.

Capital markets pull in two directions. Luke reports that investors have grown more accepting of historically small TAMs. Nick sees vibes-driven thinking, a higher Series A bar and fear of missing the next outsized outcome. Morrow's experience resembles Luke's. Founders building for batting cages and funeral homes are getting growth equity meetings, and he points to 57 million people who hunted or fished last year. The apparent contradiction may reflect different investor types. Nick's term is bifurcation: some funds chase $100 billion to $1 trillion outcomes, which excludes most vertical businesses, while others fund the middle.

The middle has a structural explanation. Luke recalls that in 2019 a Series A required $1 million of ARR growing above 100%, while a Series B demanded close to $10 million, leaving $3-10 million as no man's land. Growth funds filled it, writing roughly $5 million for about 20% and targeting majority deals at $50-60 million enterprise value. Nick's worry is that these funds act like PE, wanting 30% cheaply and discouraging VC follow-on. Luke adds that PE is reportedly down 30-something percent year over year. Morrow finds a $100 million exit target appealing, since it makes $3-10 million of ARR feel reachable.

Whether the larger market gets captured may depend on trust more than capability. Morrow expects consumers to hand agents purchases of $5,000 or $6,000 within about 2 years, then moved toward 10 years after hearing Nick, who argued for 2030 or later because behavior change and trust, not product failures, are the binding constraints. One host cited about 10% of US consumers as comfortable with autonomous booking, and a report that 4% of Americans pay for AI. By editorial inference, if operators are equally slow to trust agents with their phones, the marina's 5-to-2 headcount math will arrive later than the pricing thesis assumes.