gp0577

Segment and ICP first

Vertical SaaS fails most often at the definition stage, not the execution stage. "Healthcare" is not a segment. "Independent dental practices with two to six operatories, one location, using Dentrix or Eaglesoft, billing PPO plans" is a segment. The difference matters because everything downstream — the demo script, the integration list, the compliance posture, the price point, the channel — is derived from that definition. Get it loose and every function builds slightly different assumptions into their work.

Build the ICP on four axes, in this order. First, the workflow you own: the specific recurring task the operator performs daily or weekly that your platform absorbs. Scheduling, claim submission, route planning, permit tracking, inspection logging. If you cannot name the task in five words and say how often it runs, you do not have a wedge. Second, the system of record they already run: the vertical ERP or practice-management system that holds their data. You are almost never replacing it in year one; you are attaching to it. Third, the buying unit: how many people must say yes, and whether procurement, legal, or a compliance officer touches the contract. Fourth, the count of firms that match — a real number, sourced from association membership rolls, state licensing databases, NAICS-coded business registries, or trade-show exhibitor lists.

That fourth number is the one most teams skip and it is the one that determines whether the business works. If the segment contains 3,000 firms and realistic penetration over five years is 8-12%, you land 240-360 customers. At $9,000 average annual contract value that is a $2-3M revenue ceiling for the wedge — fine as a beachhead, fatal as the whole plan. Run this arithmetic before you hire the first rep, because it tells you whether you need a $30K ACV motion into larger operators or a $200/month self-serve motion into thousands of them. Those are entirely different companies.

Segment by operational shape rather than by size band alone. Two 40-employee firms in the same trade can have completely different buying behavior if one is private-equity-rolled-up and one is owner-operated. Roll-ups buy centrally, evaluate against a standardized stack, sign multi-site contracts, and take 6-9 months. Owner-operators decide in a week, pay by card, and churn when the owner retires or sells. Pick which one your product and cash position can actually serve, and write the other into a "not now" list that the whole go-to-market team can see. A published anti-ICP is one of the cheapest sales-efficiency tools available: it stops reps from burning quarters on deals that were never going to close well.

Layer in regulatory exposure as an explicit ICP dimension. If the segment is subject to HIPAA, you need a Business Associate Agreement template and audit logging before your first paid customer, not after. If it is subject to OSHA recordkeeping, PCI, or state-level licensing rules, those requirements shape the product roadmap and the security questionnaire you will answer hundreds of times. Vertical SaaS teams that treat compliance as a later-stage enterprise concern discover mid-cycle that their smallest deal has the same legal surface as their largest, because the regulation applies to the operator, not to the contract size.

Finally, pressure-test the segment against three questions before committing a year to it. Does the operator lose money or face penalty when the workflow goes wrong? Is there a budget line today, even if it is spent on labor, spreadsheets, or a legacy vendor? Can you reach 200 of these firms through fewer than five channels? A yes on all three means the market is addressable. A no on the budget question in particular usually means you are selling a vitamin into an industry with thin margins, which is the slowest possible path.

The motion that fits that segment

Once the ICP is fixed, the motion follows almost mechanically from price point and buying-unit size. There are three viable motions in vertical SaaS and mixing them incoherently is the single most common structural error.

Self-serve with assisted conversion works when ACV runs roughly $600-6,000 and one person decides. The product must deliver a visible result inside the first session — a completed schedule, a submitted claim, a generated compliance checklist — using templates pre-built for the trade. Free-tier design should solve one narrow, real problem completely rather than gate a broad product at 20% capability. Conversion happens at an operational ceiling the user hits naturally: a second location, a fifth employee, a monthly volume threshold. Human touch enters only after a product-qualified signal, and the first conversation is an onboarding call, not a discovery call.

Practitioner-led sales fits ACV between roughly $8,000 and $60,000 with two to four people in the buying unit. Here the differentiator is who is on the call. A rep who ran a practice, managed a fleet, or held the license outperforms a generalist because they diagnose in the customer's vocabulary and can challenge a bad answer. Hire the first two of these before hiring a sales manager. Sales cycles in this band typically run 30-90 days, and the reliable accelerant is a structured pilot: a fixed 30-day scope, one location, a written success metric agreed in advance, and a pre-signed conversion order so the pilot does not restart procurement.

Multi-site and enterprise applies above roughly $75,000 ACV, where procurement, security review, and often a legal redline are guaranteed. Cycles run 4-9 months. This motion needs a solutions engineer, a completed security questionnaire library, SOC 2 Type II in hand, and a reference customer of comparable size. Do not attempt it before the mid-market motion is repeatable; the resource drain kills two motions instead of building one.

mermaid flowchart LR A["Weekly: pipeline by stage, loss reasons, first-30-day onboarding"] --> B["Monthly: channel CAC, usage cohorts vs retention, partner check-ins"] B --> C["Quarterly: segment penetration, pricing review, expand or deepen"] C --> D["Pod decision: next adjacent segment or deeper in current"] D --> A A --> E["Field signal: top 3 requests with accounts and dollars"] E --> F["Product commits: yes, no, or dated"] F --> A </invoke>

Keep the reporting stack simple enough that the pod trusts it. Three dashboards suffice: acquisition by channel with cost and conversion, activation and time-to-first-value by cohort, and retention with the workflow-usage signal alongside it. Vertical teams that build twenty dashboards end up debating definitions instead of acting; teams that build three end up acting on them weekly.

The decision to open a second vertical should have written criteria set in advance rather than being made in a moment of optimism. Reasonable gates: the first segment is producing predictable monthly new revenue over at least two quarters, gross retention is holding above your threshold, onboarding is productized to a documented number of hours, and you can name the shared asset — integration, compliance posture, data model, or channel — that transfers to the next segment. If nothing transfers, the second vertical is a second company, and it should be resourced and expected to behave like one.

Related questions

How long should a vertical SaaS beachhead take before expanding?

Typically two to four quarters of predictable acquisition plus stable gross retention. Expand only when a concrete asset transfers — an integration, a compliance posture, a data model, or a channel. Without a transferable asset, the next segment is effectively a separate company.

Should vertical SaaS publish pricing?

For self-serve and lower mid-market tiers, yes. Fragmented small-business segments compare vendors before contacting anyone, and hidden pricing removes you from the shortlist. Multi-site and enterprise tiers can stay quote-based, since scope genuinely varies by site count and integration depth.

What is the right first hire for a new vertical?

A practitioner from the industry who can both sell and shape the roadmap. They validate the workflow wedge, open association relationships, and translate customer language into product requirements — work that a generalist rep and a generalist PM would take three times as long to do.

How do you compete with an entrenched system of record?

Attach rather than replace. Own an adjacent workflow the incumbent handles poorly, integrate cleanly with their data, and prove value in weeks. Replacement conversations become possible later, once you hold enough of the daily workflow that the incumbent is the secondary system.

Does usage-based pricing work in small-business verticals?

It works when the meter matches a unit the operator already counts and when the model has a floor and a cap. Uncapped consumption pricing produces bill surprises that damage renewals, and no floor makes revenue unpredictable through the customer's seasonal trough.

FAQ

How do you size a vertical market without buying an expensive research report?

Count firms directly. Association membership directories, state professional licensing databases, business registries filtered by industry code, and trade-show exhibitor and attendee lists give you a defensible population figure. Cross-check two independent sources and use the lower number. Then apply a realistic five-year penetration assumption and your target contract value to see whether the segment supports the business you intend to build.

What compliance work is required before the first paid customer in a regulated vertical?

Whatever the regulation applies to the operator, not to your contract size. In practice that usually means encryption in transit and at rest, role-based access control, audit logging, a documented data-retention policy, an incident-response process, and any industry-specific agreement your customers must have on file. Formal certification such as SOC 2 Type II can follow, but the underlying controls should exist from the start.

How should a vertical SaaS company approach trade associations?

Lead with value to the members rather than a partnership request. Offer a practical education session, contribute data to a member survey, or sponsor a member-benefit resource. Endorsed-vendor status generally follows a demonstrated relationship, not a first email. Expect a revenue share or sponsorship fee, and evaluate that channel on the same cost-per-acquisition basis as any other.

When does product-led growth make sense in a complex industry?

When a meaningful slice of the workflow can be completed by one person in a single sitting without touching the system of record. Narrow the free experience to that slice and make it genuinely complete. If the first useful outcome requires a data migration or an administrator's approval, PLG will underperform and assisted onboarding is the better path.

How do you decide which integrations to build first?

Survey your current and target customers for what system of record they run, and build to the two or three that cover the clear majority. Publish that list. Treat a prospect on an unsupported system as a fast disqualification rather than a custom build, because every one-off integration becomes permanent maintenance that grows faster than the revenue it protected.

What retention numbers indicate the product is genuinely embedded?

Gross revenue retention above roughly 90% in an SMB-heavy segment, paired with a workflow-usage metric that stays steady week over week. Logins alone are a weak signal. If retention is in the 75-85% band, the platform is likely adjacent to the operator's daily work rather than inside it, and that is a product problem that no go-to-market change will resolve.

Sources

• https://www.bvp.com/atlas — Bessemer Venture Partners State of the Cloud and vertical software research
• https://www.saastr.com/ — SaaStr, operating benchmarks and sales playbooks for SaaS companies
• https://a16z.com/enterprise/ — Andreessen Horowitz enterprise and vertical software analysis
• https://openviewpartners.com/blog/ — OpenView, product-led growth benchmarks and pricing research
• https://www.gartner.com/en/information-technology — Gartner IT and software market research
• https://hbr.org/topic/subject/business-models — Harvard Business Review on business models and network effects
• https://aws.amazon.com/marketplace/ — AWS Marketplace, cloud marketplace distribution mechanics
• https://appexchange.salesforce.com/ — Salesforce AppExchange, partner and marketplace listing model
• https://www.aicpa-cima.com/topic/audit-assurance/audit-and-assurance-greater-than-soc-2 — AICPA SOC 2 reporting framework
• https://www.hhs.gov/hipaa/for-professionals/index.html — U.S. HHS guidance on HIPAA for covered entities and business associates

Related on PULSE

• [Vertical SaaS go-to-market playbook for healthcare in 2027](/knowledge/gp0507)
• [What is the go-to-market playbook for streaming media platforms in 2027?](/knowledge/gp0552)
• [What is the go-to-market playbook for edtech platforms in 2027?](/knowledge/gp0535)
• [How do you build a vertical SaaS for auto dealers (Reynolds / CDK) go-to-market motion in 2027?](/knowledge/gp0078)
• [How do you build a vertical SaaS for electrical contractors go-to-market motion in 2027?](/knowledge/gp0087)
• [How do you build a vertical SaaS for HVAC and plumbing contractors go-to-market motion in 2027?](/knowledge/gp0085)