Want to start the next billion-dollar AI company? Seven frameworks for AI-enabled vertical SaaS

Forget general-purpose copilots and chatbots. The next wave of AI will be built by those who can reimagine a specific industry vertical from the ground up. The potential of vertical AI represents a staggering $6 trillion opportunity in annual services spend in the U.S. alone—that's ten times greater than the $600 billion market cap of all cloud software combined.

Vertical AI can tackle previously intractable problems by addressing unstructured data, which makes up about 80% of all data and has been challenging for traditional SaaS to handle effectively.

But success in this space requires a fundamental shift in product development, monetization, and growth. Consider the “Fosbury flop”: just as Dick Fosbury revolutionized high jumping by perfecting the backward technique, vertical AI startups should find unconventional ways to overcome industry challenges rather than just try to “jump higher.” By taking an unconventional approach to creating and delivering value, startups can disrupt established players and capture market share more effectively.

In this guide, we’ll outline seven frameworks we believe founders can use to turn these ideas into reality.

Seven powerful frameworks for vertical AI

1. Networked SaaS

Establish multisided platforms that connect various stakeholders in an industry ecosystem to address a constellation of needs. We're seeing the emergence of “networked SaaS”—combining elements of the vertical SaaS and marketplace models to produce network effects that entrench the platform. While this is the most complex framework we’ll discuss in this post, we think it’s a model with a lot of potential.

Where a typical vertical B2B marketplace connects supply and demand, a networked SaaS marketplace goes beyond, providing different SaaS functionality for different stakeholders across the stack. Imagine a startup that builds back-office automation and billing software for home services professionals, funnels demand to them with a website for homeowners suffering weather damage that matches them with these professionals, and provides analytics dashboards to insurers who pay for the home repair. This may sound like trying to build too many products at once, but it’s possible within niche verticals that constrain the feature set required. This is especially true within fragmented markets that include individuals or very small businesses on the supply side that want a consolidated software stack but do not require the most sophisticated enterprise capabilities. 

To understand networked SaaS, it's crucial to differentiate it from other common models:

• Traditional vertical SaaS: Focuses on building a software suite for a single, specific niche buyer within an industry
• Horizontal SaaS: Offers a single product that can be sold to a wide range of buyers across different industries
• Two-sided marketplace: Simply connects supply and demand, allowing stakeholders to find each other

In contrast, networked SaaS:

• Provides specific point solutions for different stakeholders across the industry
• Unifies this into a common set of workflows to eliminate systemic coordination problems
• Doesn't necessarily monetize each stakeholder but can generate value for or drive the acquisition of all stakeholders

This approach creates a more efficient value chain and deeper lock-in for the company than traditional SaaS models, positioning a startup at the center of their industry.  

Example: Verse Medical starts with an AI-enabled ordering platform for hospital supplies as a wedge to become an in-home care provider.

Verse offers a free digital ordering workflow for clinicians who support patients transitioning from hospitals back to their homes. It replaces the inefficient process of calling and faxing orders to medical suppliers. For patients, Verse tracks those medical supply orders, reviews how patients are adhering to their medical protocols, reorders supplies, and communicates the patient’s status to clinicians. For medical suppliers, Verse provides streamlined orders and analytics dashboards to help them understand their market share. For insurers, Verse demonstrates improved clinical outcomes from patient interventions, lowers re-admission rates, and now it’s building deeper integrations and partnerships with these insurers. 

Verse can become a central coordination layer on behalf of hospitals for in-home patient care in the future, adding another stakeholder of in-home health agencies and skilled nursing facilities to the Verse platform.

Verse's approach demonstrates key aspects of the networked SaaS model:

• Multistakeholder solution: Verse offers tailored tools for nurses, patients, medical suppliers, and insurers—next, home health companies.
• AI-driven efficiency: The platform uses AI for efficient billing/RCM, claims denials, patient follow-up, and more.
• Bottom-up adoption: By allowing direct clinician adoption, Verse sidesteps long, drawn-out hospital sales cycles.
• Network effects: Aggregating clinicians' purchasing power enables valuable negotiations with suppliers and insurers.

With a networked SaaS model, Verse is building a highly defensible market position that poises it for expansion.

2. Focus on data defensibility

Build proprietary, high-value datasets by focusing on unique, industry-specific information that’s difficult for competitors to replicate. This could involve aggregating data from multiple sources or generating novel, proprietary data through a workflow that the company has built, aggregating data through integrations or generating novel data through AI-powered processes.

Just expecting customers to blindly pass you their data may not be enough; according to McKinsey, IP infringement is one of the top concerns for organizations adopting generative AI, cited by 52% of respondents. Startups that help their customers safely leverage internal data stores with AI have massive potential. 

Example: EvenUp's proprietary settlement data from personal injury cases creates a strong competitive advantage and enables more accurate predictions for clients. By securely analyzing settlement data from a huge volume of cases, EvenUp provides unparalleled insights that help lawyers and clients make more informed decisions than any individual law firm could with its own data.

3. AI-business-in-a-box

Empower individuals to start AI-driven businesses by providing comprehensive, turnkey solutions that handle complex backend operations. This approach democratizes entrepreneurship and taps into the growing side hustle / gig economy while addressing critical labor shortages in various industries.

The business-in-a-box model is experiencing a moment of growth driven by several key factors:

• Labor shortages: Many industries, particularly healthcare, face severe worker shortages and burnout.
• Technological efficiency: These platforms use AI and automation to help practitioners do more with less. 
• Flexibility and remote work: Unlike traditional models such as private equity roll-ups, business-in-a-box platforms cater to younger workers’ desire for flexible hours and remote work. 
• Comprehensive solutions: By bundling all necessary tools and services, these platforms remove the overwhelming friction of starting a business, allowing practitioners to focus on their core competencies.
• Deep monetization: By providing a full suite of services, these companies can take a significant share of revenue (as much as 40%–50% of payment volume in some cases), making individual practitioners economically viable customers.

Example: Grow Therapy improves client access to care, but it also helps therapists simplify or automate tasks such as scheduling, billing, compliance, and customer relationship management. This allows mental health professionals to focus on their core expertise—providing therapy—while the AI-powered platform handles the administrative burden. The result is increased access to care and more efficient practice management for therapists.

The business-in-a-box model: Grow Therapy’s playbook

A look at how Grow Therapy used the business-in-a-box model to build trust, expand, and scale quickly

Read More

4. Automate the work, not the worker

Focus on automating specific, repetitive outputs rather than entire roles, which may comprise a wide range of tasks that are hard to automate. This targeted approach enhances productivity without threatening job security, making it more approachable for workers and easier to implement.

This strategy offers several key advantages:

• Improved job satisfaction by eliminating tedious tasks
• Faster adoption and integration into existing workflows
• Building in the potential to scale as tasks are automated
• Improved output quality and consistency

Example: EvenUp's legal document automation and Peer AI's clinical trial patient narrative summaries showcase how AI can tackle complex, time-intensive tasks. By automating the creation of legal demand packets or lengthy clinical study summary reports, these tools free up professionals to focus on strategic work and client interactions. In many of these high-stakes industries where work can be a matter of life and death, the key is to augment human capabilities rather than replace them entirely.

5. Build a full-stack AI company

If you’re facing complex, high-value problems that require deep expertise, become an end-to-end service provider to get and maintain a hold in the market. Lean on AI to deepen your competitive edge.

By offering a comprehensive solution, we believe vertical AI companies can:

• Capture more value: Control over the entire process allows for higher margins and multiple revenue streams.
• Ensure quality: End-to-end oversight enables consistent quality control and faster iterations.
• Build deeper relationships: Direct interactions with end-users foster stronger customer relationships and valuable feedback loops.
• Create higher barriers to entry: The complexity of an end-to-end solution makes it harder for competitors to replicate.
• Accelerate innovation: Full control allows for rapid testing and deployment of new AI-driven features across the entire service.

This approach has its challenges: increased operational complexity, higher capital requirements, and potential conflicts with existing market players. Carefully assess your company's capabilities, market dynamics, and potential ROI before pursuing this strategy. A hybrid model combining platform provision with select end-to-end services can offer balance between control and operational manageability.

Example: Justpoint's approach to mass tort litigation demonstrates how a full-stack AI company can revolutionize an entire legal process. By handling everything from case identification to settlement negotiations, Justpoint combines AI with legal expertise to create solutions that go beyond traditional SaaS. This model allows them to take on operational responsibilities and share in both the risks and rewards of outcomes.

6. Own the origin dataset in the first of a series of actions

Companies that aggregate and structure the initial dataset (the first domino in a series of actions) can build a natural advantage extending into subsequent actions. This creates opportunities to serve customers upstream in the process and potentially displace companies further down the action chain.

Example: One of our portfolio companies, Stampli, exemplifies this approach in the accounts payable space. Their AI-powered platform, Billy the Bot, captures and processes invoice data (input A), which triggers an approval process (action A), and subsequently a payment (action B).

By owning the origin dataset (invoice information), Stampli can upsell subsequent actions:

1. Advanced analytics and reporting features
2. Vendor management tools
3. Integrated payment solutions (Stampli Direct Pay)

By digitizing and optimizing these traditionally disconnected workflows, Stampli generates significant customer value through reduced errors, faster approvals, real-time visibility, and improved vendor relationships.

7. Move from the outer edges of a customer’s needs to the core

It's often difficult for new companies to start in the core processes of a customer's business. These areas are typically the most scrutinized, optimized, and guarded. Processes that are far from the core of a business can be seen as a more accessible entry point.

Many of these workflows are too small and idiosyncratic to automate easily, and front-line employees are given localized autonomy for these smaller decisions. However, in aggregate, these small workflows and decisions can be substantial and underoptimized relative to core processes.

Example: Longtail.ai starts by optimizing pricing for the “long tail” of flight connections that are too low-volume for airlines to justify optimizing with human pricing analysts. Over time, customers have pulled Longtail into the “midtail” of flight routes and other high-value services for the airlines.

Implementing the frameworks: Strategies for success

Choosing the right framework(s) for your vertical requires a deep understanding of industry dynamics, pain points, and potential for AI-driven disruption. For maximum impact, you can combine multiple frameworks by identifying alignment between different strategies. For example, a data defensibility approach could be complemented by a full-stack AI model to fully capitalize on proprietary insights. By applying one or more of the above frameworks, we believe founders can deliver huge value in traditionally underserved industries.

Are you building an AI-enabled vertical SaaS solution? Reach out to Wayne and sign up for our Vertical SaaS newsletter for periodic updates from SignalFire.