Preface · a hypothesis, not a diagnosis
Every observation in this document is drawn from public signals: product pages, review platforms, investor announcements, job postings, LinkedIn org patterns, contractor forums, and market reports. Public signals are directionally accurate and independently verifiable, but they do not carry internal context. They cannot see the roadmap items already in flight, the architectural decisions already made, or the operating dynamics already in motion. Consider this a starting thesis, not a final verdict. The work gets sharper once I am inside the room.
01 · Context
The founder built the company over 14 years, raised $383M across an early growth investor ($53M, 2021) and a control investor ($330M, November 2024), and made three roofing-adjacent acquisitions. More than 8,000 contractors run the platform across roofing, gutters, siding, and solar. The $330M the control investor round was attributed publicly to 'record revenue, strong growth, and investments in AI.' The board has underwritten a specific future: the company as the intelligence layer for the roofing industry, not just the system of record. The product foundation still reflects a 14-year CRM spine — rules-based automation, two AI modules (both narrow, one still in closed beta). Independent assessments call the AI story 'two narrow products, one in beta.' The gap is not effort. The gap is architecture.
02 · Market framing
ServiceTitan ships predictive job scoring and smart dispatch. AccuLynx has AI lead scoring. QuoteIQ markets directly against this category on AI. CompanyCam is advancing visual intelligence faster than any CRM in the space. One competitor built a storm-triggered lead engine that creates contractor jobs from NOAA hail data before homeowners know they need a roofer. According to the 2026 ServiceTitan Market Report, 47% of contractors now prioritize production features over sales features. Speed-to-lead automation is the single most requested capability across roofing contractor forums in 2026. The window to own predictive roofing software is measured in quarters, not years.
03 · Executive thesis
The company has earned category trust, contractor fluency, and a founder-led operating history that has survived multiple market cycles. A feature set does not do what The company has done. A platform does. The next question is whether the platform can become intelligent. The 8,000-contractor base is a data asset competitors cannot recreate from scratch. But data only becomes a moat when the product knows what questions to ask of it. The intelligence layer must turn weather events into jobs, field activity into forecasts, customer behavior into next best actions, and operational data into strategic advantage. That is the work the next VP of Product owns.
04 · Root causes
- 01
The intelligence gap is structural, not a backlog problem
The two AI modules sit beside the existing platform, not intelligence embedded in its data model. The Insights dashboard describes what happened — stage durations, bottlenecks, workflow throughput. The automation engine executes predefined rules. There is no predictive feature anywhere in the visible product: no job profitability forecasting, no win/loss prediction, no cash flow modeling, no anomaly detection, no next best action.
- 02
AI is being retrofitted onto the platform, not architected into it
The Chief AI Officer (hired October 2025) operates with a mandate over 'product, engineering, and operations' that is parallel to product leadership, not inside it. A predictive platform does not simply add AI features. It changes what the product believes a job, contact, estimate, crew, payment, and workflow are supposed to do.
- 03
Generational debt under the surface of the data model
Core CRM spine (Contacts, Jobs, Boards, Automations) still reflects its original 2011 design. V1 and V2 of the Jobs entity coexist in production today. The most recent acquired product still functions primarily as a semi-standalone app connected via API rather than fully unified.
- 04
Product authority has been in transition for 16 months
The last VP with full mandate moved to COO in late 2022 and left in January 2025 (now Executive in Residence at an earlier growth investor). The Group PM for AI and Core Platform left for another SaaS company in December 2025. The VP of Engineering departed May 2025 as a post-the control investor restructure was underway. Engineer-to-PM ratio is approximately 10:1 against an industry standard of 5:1. The next VP does not inherit a quiet roadmap — they inherit a trust-rebuilding moment.
05 · Operating problems
- 01
Speed-to-lead is reactive, not event-driven
GTM partners (Roofle, SalesRabbit, Spotio, Angi) are all reactive lead sources. The contractor waits for the homeowner to raise their hand. The competitive advantage in storm restoration belongs to the contractor who is on the doorstep before the homeowner has called their insurance company.
- 02
Pricing model designed for seats, not intelligence value
The current three-tier model caps automations and integrations at Growing and gates API access entirely behind Established. Independent analyses estimate $619/mo on Growing and $800–$950/mo on Established for a 5-person team. AI value does not map cleanly to seats — it maps to jobs created, estimates generated, claims improved, payments collected faster, churn prevented. Packaging intelligence as another add-on underprices strategic value.
- 03
Mobile foundation distrust will infect the field AI story
The most consistent complaint across G2, Capterra, and contractor forums is mobile. Verbatim Capterra review: 'TERRIBLE mobile apps. Our field guys have resorted to opening the desktop version on their phones.' The closed-beta field AI tool will be one of the most visible AI experiences. If contractors meet the closed-beta field AI tool through a foundation they already distrust, field AI inherits the mobile trust problem.
- 04
Competitive scorecard is a quarterly slide, not a product signal
The JD names competitive scorecard maintenance as a VP responsibility. A live scorecard functions as a continuous product signal — sharper battlecards for Sales, clearer roadmap tradeoffs for Product, retention risk signals for CS, executive confidence about where to lead, follow, or ignore. The goal is not competitor obsession. The goal is conviction.
06 · Organizational readiness
The VP of Product seat has been in transition for 16 months. The last full-mandate VP became COO in late 2022 and exited in January 2025. The Group PM for AI and Core Platform left for a Utah SaaS competitor (another SaaS company) in December 2025. The VP of Engineering departed in May 2025. A Glassdoor review from an Account Executive in October 2024 used this exact phrase as its title: 'Product and PX are dragging everyone else down.' The next VP rebuilds the function from a position of maximum leverage — right after $330M landed, right before the AI roadmap is set, right at the moment the board has signaled it wants a CPO-level upgrade.
07 · Product leadership mandate
- Define the intelligence architecture before the next acquisition — a written thesis covering canonical objects, events, and relationships AI can safely reason across.
- Ship the event-driven lead engine that turns storm activity, hail maps, permit filings, and claim patterns into contractor opportunities before the homeowner picks up the phone.
- Rebuild the PM bench and clarify product authority over AI — AI strategy, architecture, quality, GTM, and roadmap all integrated into product accountability.
- Redesign pricing and packaging around intelligence value — tier-embedded, usage-based, role-based, or premium intelligence suite tested deliberately.
- Fix the mobile foundation before the closed-beta field AI tool becomes the field AI story. Mobile parity is the trust layer for field intelligence.
- Turn the competitive scorecard into a live operating system that feeds Sales, Product, Marketing, CS, and the executive team continuously.
08 · The first 90 days
Days 1–30 · Diagnose
Diagnose the inheritance
Listen with structure. Interview every PM, the CAIO, CTO, CRO, CEO, and key functional leads. Talk to five to eight customers across contractor size and use case. Review product analytics, roadmap history, support themes, churn reasons, sales loss data. Audit AI initiatives and data architecture. Day 30 deliverable: an inheritance document for the founder & CEO — what is load-bearing, what is debt, what is unclear, what needs sequencing.
Days 31–60 · Operate
Establish the operating model
Turn diagnosis into an operating model. One planning rhythm across Product, Engineering, AI, Design, Sales, and CS. Documentation standards for roadmaps, PRDs, decision memos, AI evaluation, launch readiness. Bring Sales into competitive review. Begin pricing and packaging review tied to customer value. Day 60 deliverable: a proposed product operating model covering decision rights, team structure, AI governance, roadmap cadence, and competitive intelligence workflow.
Days 61–90 · Ship
Ship something visible. Set the 12-month roadmap.
Select one high-confidence, visible improvement — General availability of the closed-beta field AI tool is the clearest AI proof point tied to the board's thesis. Sequence the 12-month roadmap around three near-term milestones connected to the intelligence thesis. Define the 18-month horizon around the event-driven lead engine and predictive workflow architecture. Day 90 deliverable: a 12-month product thesis. A sequenced operating plan that replaces Gantt charts and feature wishlists with conviction.
09 · Metrics to watch
| Metric | Target |
|---|---|
NPS The JD targets this directly. NPS is the cleanest signal that product credibility is being rebuilt after a 16-month leadership transition. | Above 50 within 12 months |
90-day churn The leading indicator that field workflow trust — especially mobile — has been restored. | Below 5% |
Event-driven lead conversion Proves the intelligence thesis. Without this metric, the AI roadmap is feature-shipping rather than category-defining. | Measurable jobs created from storm/permit signals before the homeowner inbound, instrumented at the cohort level |
Intelligence ARPU lift Confirms that pricing and packaging align with AI value, not seats. | Quantifiable expansion attributable to intelligence packaging within 12 months |
10 · Risks & mitigations
AI shipped as features beside the platform, not architected into it
MitigationDay-30 inheritance document forces the architectural call before any new AI feature ships. Decision rights move AI roadmap inside product accountability.
Pricing change destabilizes installed-base economics
MitigationTest packaging deliberately by tier, usage, role, and premium suite with a small cohort before broad rollout. The control investor's economics depend on retention.
Mobile rebuild gets deprioritized in favor of visible AI
MitigationSequence mobile parity ahead of the field AI tool's GA. Field AI fails if the trust layer underneath it has not been earned.
11 · Why now
The control investor's $330M investment carries return expectations that are specific and time-bound. The press release attributed the round to investments in AI — the board has funded a platform expansion thesis, not a vague narrative. ServiceTitan, AccuLynx, QuoteIQ, and CompanyCam are all moving. The window to own predictive roofing software is measured in quarters, not years. The 8,000-contractor data asset is the moat. It only compounds if the intelligence layer is built on top of it. The VP of Product hired now is not inheriting a steady-state org — they are rebuilding the function from a position of maximum leverage. That combination of urgency and mandate does not last.