Convex
Winter 2019AcquiredSoftware for the commercial services industry.
Pressure-test this opportunity
Explore the risks and possibilities with a prompt for ChatGPT, Claude, or your agent.
Software for the commercial services industry.
Explore the risks and possibilities with a prompt for ChatGPT, Claude, or your agent.
Convex built a map for the people who sell services to commercial buildings. Founded in 2017 and accepted into YC's Winter 2019 batch, it combined property, permit, equipment, ownership, company, and contact data with sales workflows for HVAC, fire safety, security, elevators, janitorial, and other contractors. Convex raised $60 million and became a meaningful vertical sales platform.
ServiceTitan acquired Convex in April 2024 for $25.8 million, less than half the disclosed funding.[1] That makes the venture outcome weak even though the product survived. ServiceTitan kept the brand, founders, data, and workflow, then connected Convex to quoting, jobs, and realized revenue. Convex found a valuable top-of-funnel wedge; its independent limit was that the incumbent system of record could make the same wedge more complete.
Charlie Warren grew up in Detroit around his family's wiring business. He understood commercial service work before he entered software: technicians maintained physical systems, but the companies behind them often sold through relationships, territories, paper lists, and institutional memory. After college and graduate work in energy and finance, he began collaborating with Blake Meulmester, whose prior work included industrial software, data, and analytics engagements at Goldman Sachs and Deloitte.
Warren and Meulmester started Convex in 2017 in Meulmester's San Mateo garage. Warren later recalled that they “bootstrapped for six months with one whiteboard to our names.”[2] Brandon Dement and Sven Linsmaier were also named as early builders in investor accounts. The central idea was that a commercial property was a rich data object. Its location, size, use, owner, tenants, equipment, permits, and prior work could tell a contractor whether the building fit its ideal account.
The first product was Atlas, a map-based sales-intelligence application. A representative could draw or select a territory, filter buildings, identify owners and decision makers, inspect permit or equipment signals, and manage outreach. The product replaced a familiar process: driving around looking for promising buildings, searching county records, browsing LinkedIn, and assembling lists in spreadsheets.
The first onboarding exposed a positioning problem. Warren sold Atlas, flew to the customer, and demonstrated it to a 12-person team. Their reaction was to ask what else it did. He took away a sentence that became his operating rule: “always lead with outcomes over features.”[2] Convex would sell faster prospecting, shorter rep ramp time, and more service revenue, not a map.
Convex entered YC in winter 2019.[3] Emergence Capital led a $17 million Series A in 2020. In 2021, Fifth Wall, Emergence, and GGV led a $39 million Series B, bringing disclosed funding to $60 million.[4]
Atlas made commercial-property intelligence usable by sales teams. The database covered buildings, owners, tenants, contacts, permits, and physical attributes. A rep could filter by building size, owner occupancy, property type, equipment age, or recent work, then decide which accounts deserved attention. The map matched how contractors thought about territories and service density.
Convex added sales execution around that intelligence. Teams could create accounts and opportunities, manage outreach, record activity, call or email prospects, and track pipeline. Later products added buyer-intent signals and automated lead discovery. The present ServiceTitan product also offers AI-assisted searches and outreach, though those features expanded after acquisition.[7]
The product mattered because commercial services have recurring, property-specific demand. An HVAC unit ages. A permit can reveal a recent installation. A building owner changes. A warranty expires. The best prospect is not merely a company with the right industry code; it is a physical site with equipment, budget, and a decision maker inside a contractor's service territory.
A Comfort Systems USA Southwest case study shows the workflow. Before Convex, reps used LinkedIn or drove building to building and could spend two days filling the pipeline. Convex let them screen by size, occupancy, and prior work. One sales consultant said he could reduce weekly prospecting to a few hours. Management said new-rep ramp time fell from six-to-nine months to two-to-three months and that Convex became the center of prospecting. The same manager explicitly said the tool did not cause all of the division's growth.[8]
Under ServiceTitan, Convex moved from a separate prospecting layer toward an integrated commercial workflow. A meeting found in Convex can become a ServiceTitan CRM opportunity, then move into quoting, job execution, invoicing, costing, and reporting.[6]
Convex served commercial contractors with dedicated revenue teams: HVAC and mechanical firms, building-automation vendors, fire and life-safety providers, elevator companies, electrical contractors, security integrators, solar businesses, and facility-service firms. Individual users included territory sales representatives, sales managers, marketers, and executives.
The best accounts had high customer lifetime value, recurring maintenance, defined geography, and salespeople who otherwise spent hours constructing prospect lists. Convex sold both data and a change in selling behavior. The software had to be accurate enough to earn trust and simple enough for a rep who worked from a vehicle or customer site.
Emergence estimated the commercial-services industries at more than $1 trillion in annual global revenue in 2021.[9] That figure describes contractors' revenue, not software spending. Convex's available software market was the fraction of those firms with enough sales headcount, contract value, and data maturity to pay for specialized intelligence.
ServiceTitan's acquisition materials claimed that Convex helped customers generate nearly $1.5 billion of incremental growth and an estimated $3 billion of revenue in 2023, with a 9x median first-year return.[5] The footnote assumed 95% lifetime-value retention and a 45% margin. Those assumptions make the figures directional marketing evidence, not audited customer economics.
Convex competed with ZoomInfo and other horizontal data vendors, Salesforce and HubSpot for CRM, local permit or assessor databases, Google Maps, spreadsheets, and the contractor's own relationships. Its advantage was a property-first model. Horizontal tools organize around companies and people; Convex organized around the commercial building where service work occurred.
That same workflow sat adjacent to ServiceTitan. ServiceTitan already managed customers, dispatch, technicians, work orders, estimates, invoices, and job costs. Convex supplied the prospect before that prospect became a customer. A combined product could trace one record from property discovery to gross profit, improve lead scoring with job outcomes, and eliminate duplicate data entry. The buyer therefore had a natural data and distribution advantage that a standalone prospecting product could not match.
Convex sold vertical SaaS subscriptions plus data access and enterprise onboarding. Public pricing, contract length, gross margin, retention, and standalone revenue were not disclosed. Disclosed funding reached $60 million across seed, Series A, and Series B.[4]
ServiceTitan's fiscal 2025 Form 10-K supplies the exit economics. It paid $25.8 million: $23.5 million in ServiceTitan shares and $2.3 million cash. The accounting assigned $15.0 million to goodwill and $9.5 million across developed technology, customer relationships, and trade name, alongside other acquired assets and liabilities.[1]
The purchase price was 43% of disclosed funding. That comparison does not reveal the payout to common holders because the public record lacks the preferred-stock terms and capitalization table. It does establish that the company did not exit above invested capital in aggregate.
Convex published customer stories across HVAC, fire safety, building automation, electrical work, security, and facility services. Comfort Systems said Convex reduced prospecting time and new-hire ramp time.[8] Another later case study said Mechanical Services and Design added more than $42 million to its sales pipeline over 18 months, a pipeline figure rather than realized revenue.[10]
The strongest evidence of durable product value is post-acquisition investment. Convex still operates under ServiceTitan, now advertising 146 million contacts, intent signals, property search, outreach, and a direct connection into commercial CRM.[7] ServiceTitan did not buy the customer list and retire the software; it made Convex the front end of its commercial product.
Convex raised $39 million in October 2021 after an earlier $21 million. The funding announcement said it would hire, scale the platform, and develop new products.[4] This was a familiar 2021 plan: use venture capital to capture a large vertical quickly.
ServiceTitan paid $25.8 million less than three years later.[1] No public founder account explains the gap. There is no sourced evidence of a shutdown threat, missed payroll, or forced process, so the sale should not be described as a rescue. Still, a price below total funding means the venture-scale capital structure did not convert product value into a strong standalone exit.
The attempted remedy was continued product expansion. Convex moved beyond Atlas into engagement, CRM, intent, and broader revenue intelligence. That widened the contract but also pushed it toward crowded horizontal categories and ServiceTitan's product boundary. More features could raise revenue per customer; they could not give Convex the buyer's installed base or job-level outcome data.
Convex owned the first stage of commercial service revenue: identifying a building and reaching a buyer. ServiceTitan owned the stages after that: quoting, dispatch, work, invoicing, and costing. When separate, both systems had incomplete feedback. Convex could show pipeline but not always closed-job margin. ServiceTitan could show execution but not the entire addressable property market.
Acquisition closed that loop. In 2025, ServiceTitan demonstrated AI Leads, meeting-to-opportunity handoff, quoting, and embedded reporting across Convex and its CRM.[6] This is the structural mechanism behind the exit. Vertical data applications become features of the operating system when outcome data and distribution matter more than the standalone interface.
Convex tried to resist that outcome by becoming a larger revenue platform. ServiceTitan could reach the same destination faster by buying it. The independent product's strength therefore increased its strategic value while weakening the case for remaining separate.
The strongest countercase is visible today. Convex remains active, customers still use it, and ServiceTitan has added data and workflow. The acquisition announcement said Charlie Warren would remain CEO and Blake Meulmester would lead product; Meulmester later appeared as a ServiceTitan product vice president.[11] This was not a product failure.
The investor outcome is harder. A $25.8 million acquisition after $60 million raised cannot return aggregate invested capital before preferences, transaction costs, or dilution. The lesson is not that vertical SaaS lacks value. It is that a good vertical feature can have a natural owner, and raising for independent category dominance makes that otherwise useful acquisition look small.
Model the physical object. Convex organized sales around buildings rather than generic company records. That domain model made ordinary public data useful to contractors with territories, installed equipment, and recurring service work.
Lead with the customer's result. An early customer asked what else Atlas did after seeing the demo. Convex learned to sell shorter prospecting time, faster rep ramp, and won service agreements instead of map features.
Pipeline is weaker than job economics. Convex could measure opportunities; ServiceTitan could connect them to quotes, work, invoices, and margin. The system with realized outcome data had the stronger product position.
Capital changes what counts as success. Convex built a lasting product and found a strategic home, but $60 million of funding made a $25.8 million sale a weak venture return.
An acquisition can complete the product thesis. Convex by ServiceTitan now spans property discovery through commercial CRM. The original wedge became more useful after it joined the system of record.