A product change begins quietly. An engineer swaps a component. A quality manager revises a procedure. A supplier gets a new drawing. Marketing still has the old specification, sales promises the old configuration, and service learns the truth when a customer calls. The part may cost $4. The organizational chase can cost weeks.
Propel Software built its business around that chase. The Redwood City company keeps the bills of materials, engineering changes, quality records, supplier conversations, product descriptions, digital assets, sales context, and service history on one Salesforce-native platform. Its term for the bundle is Product Value Management, or PVM. The less branded explanation is better: one governed product record, used by nearly everyone who touches the product.
This is not task management for people making another app. Propel is aimed at companies that make complicated things: medical devices, safety equipment, smart-home gear, industrial machinery, electronics, consumer goods. Their products have revisions, approvals, regulations, suppliers, variants, labels, complaints, and a long afterlife in the field. Publicly named customers include VIZIO, Zoetis, Samsara, SimpliSafe, Rawlings, Guardant Health, Desktop Metal, Savant Systems, and Breg.
One record, six impatient departments
01The bet was boring. That helped.
Founder Ray Hein came from the world that Propel intended to disturb. He had worked at Agile Software, later bought by Oracle, and held leadership jobs at Apttus, Vendavo, and Centric Software. On April 22, 2015, after about a month of groundwork and notes on one sheet of paper, he incorporated Propel with fellow enterprise-software veterans Brian Sohmers and Ron Hess.
Their key decision was architectural: build PLM natively on Salesforce. Traditional product-lifecycle systems were engineering vaults, often installed on premises and difficult to extend. Salesforce already held the customer record and offered permissions, workflows, reporting, collaboration, mobile access, and an ecosystem. If the product record lived there too, a customer complaint could lead to the affected revision; a sales request could become a product requirement; a released specification could move toward a channel without manual re-entry.
Investors paid for that wager in three visible stages. Cloud Apps Capital Partners led an initial $4.2 million Series A close in 2016, with Salesforce Ventures and SignalFire participating; Propel later described the full Series A as $10 million. Norwest led an $18 million Series B in 2018 after Propel reported 500 percent annual revenue growth and a tripling of customers. Salesforce Ventures led a $20 million Series C in 2021, joined by Norwest, Cloud Apps, Green D Ventures, TIFF, and Ankona Capital. The last announcement put total capital raised at $48 million. Propel remains private, and no reliable public valuation is available.
That choice also created the company's sharpest boundary. Propel is attractive when a manufacturer wants Salesforce adjacency or a broad cloud platform. It is less obvious for a tiny team with a simple bill of materials, no regulated quality process, and no reason to connect engineering to customer data. In that case, a lighter PDM or document tool may be enough. Platform leverage becomes platform dependence if the surrounding architecture is a bad fit.
“Product success is a team sport.”Ray Hein, founder of Propel
02What failed first was the handoff
Propel's customer stories repeat one plot with different props. The software did not arrive because a bill-of-materials screen looked old. It arrived because the gaps between systems had become operational liabilities. Breg, a maker of orthopedic products, had product and supplier information scattered across 20 systems. Advanced Sterilization Products inherited six separate quality applications after a divestiture and had one year to stand up its own technology stack. Girtz Industries wanted product data to travel across engineering, finance, and sales instead of being retyped at each border.
The first failure, then, is usually continuity: somebody copies data, somebody misses a change, or somebody cannot see the context. The visible symptom might be a slow change order, a labeling error, an audit scramble, a supplier risk discovered late, or a product page published from last week's spreadsheet. Companies change their minds about replacement when those inconveniences gain an owner and a number.
Breg reported 186 percent ROI one year after a big-bang PLM and QMS replacement, alongside a 77 percent reduction in product-development time and 50 percent less document-control work. Guardant Health reported 27,000 hours saved annually across core processes. Savant Systems reported a 74 percent reduction in time to market. Imperative Care said one management-review report fell from five days to five minutes. These are vendor-published customer results, not promises that every deployment repeats them. But they identify where buyers expect the money to appear: fewer handoffs, less searching, faster approvals, cleaner audits, and earlier launches.
03The invoice is not a mystery
Propel is unusually direct about the shape of its pricing. Licenses are annual and user-based. Salesforce platform access is included. Optional products cover supplier workspaces, visual collaboration, digital FDA submissions, and premium support. Public starting prices on Salesforce AppExchange begin at $73 per user per month, but enterprise math quickly leaves that neat number behind.
Add the published bands and a large deployment can approach $900,000 before internal labor. Deployment estimates run from three weeks or less for small projects to three to six months for large ones. Cost is therefore not merely the subscription. It is data cleanup, process decisions, validation, integration, and the political work of persuading six departments that “their” record is now shared.
The counterargument is equally concrete. How many hours disappear into reconciliations? What does one late launch cost? What is the price of using the wrong revision in a regulated product? Breg's 20-system starting point explains the purchase better than any feature list. Propel works when the cost of fragmentation is visible, recurring, and bigger than the program required to remove it.
04PLM grew tentacles
The company began with cloud PLM: bills of materials, change control, documents, projects, costs, and supplier collaboration. It then joined Quality Management System functions such as training, CAPA, complaints, audits, nonconformance, equipment calibration, and supplier quality. Product Information Management added descriptions, assets, variants, categories, and channel-ready content. In 2022, Propel renamed itself from PropelPLM and introduced the broader PVM label.
The products that followed made the shared record more useful. Supplier Community gave outside partners controlled access. Component Insights brought SiliconExpert sourcing and compliance data for more than one billion standard parts into context. Product Spotlight created curated views for Salesforce users. DesignHub connected mechanical and electrical CAD data, automating part creation, file updates, and BOM structures. Rootstock's Salesforce-native ERP partnership connected product decisions to inventory, production, and cost.
The photograph also says something about culture that a values poster cannot. Propel calls its employees “Propellerheads” and publishes four current values: Growth, Customer Success, Trust, and Bias for Action. It pledges one percent of employee time to local causes that include food insecurity, disease awareness, and inclusion. In 2022, Salesforce veteran Ross Meyercord replaced Hein as CEO while Hein moved into strategy. The handoff signaled a familiar stage in enterprise SaaS: the founding thesis had survived, and the next job was scale.
05AI arrived after the data
Propel One, launched in 2025 on Salesforce Agentforce, put role-based agents into engineering, quality, product management, supply chain, marketing, sales, and service. The pitch is practical rather than cinematic: summarize a change, identify an at-risk component, extract facts from a document, generate a training quiz, create product content, or move an approved routine step forward while keeping a person in the loop.
In June 2026, Propel added Model Context Protocol connectivity. Approved tools such as Claude, ChatGPT, or Copilot can query live Propel records under existing permissions; Propel agents can also reach external MCP servers for ERP, supplier, or component information. The company described it as the first production MCP implementation in PLM. The important part is not conversational access. It is that the answer can inherit a governed record, current revision, audit trail, and user-level access instead of rummaging through a folder of stale PDFs.
The timing explains what changed manufacturers' minds. AI made old data architecture newly expensive. A legacy PLM might still store a BOM, but a sealed system keeps the relationships and context that an agent needs out of reach. Propel reported 42 percent year-over-year bookings growth for the fiscal year ended January 2026, citing DesignHub, Propel One, partner sales, and replacements of end-of-life systems such as Oracle Agile PLM. It also reached the Deloitte Technology Fast 500 for a fifth straight year in 2025.
Where Propel wins
Complex, cross-functional products; regulated quality; supplier networks; Salesforce-heavy estates; and organizations ready to retire several systems at once.
What buyers compare
Teamcenter, Windchill, ENOVIA, Oracle Agile, SAP, Arena, MasterControl, ETQ, Centric, Autodesk, and combinations of specialist tools.
What can go wrong
Weak data ownership, endless configuration, scope creep, poor CAD or ERP integration, and teams that keep their shadow spreadsheets after launch.
The decisive difference
PLM, QMS, and PIM share a Salesforce-native model, so product and customer context can meet without another synchronization layer.
That competitive table hides an important distinction. Siemens, PTC, Dassault Systèmes, Oracle, and SAP can sell deep portfolios into enormous manufacturing estates. Arena and Autodesk offer cloud PLM choices; MasterControl and ETQ are familiar quality alternatives; Centric is strong in retail and consumer products. Propel does not win simply by having every possible specialist function. It wins when a buyer values one extensible model, quicker cloud deployment, Salesforce familiarity, and participation beyond engineering. The same trade can lose when a manufacturer needs unusually deep domain functionality, must preserve a heavily customized legacy environment, or prefers several best-of-breed systems over one platform.
06Copy the rollout, not the slogan
What can another company borrow? Not the PVM acronym. Borrow the sequence. Begin with the handoff that causes measurable pain. Name the records and owners involved. Put one workflow into a shared model. Keep business users close to the configuration. Track cycle time and error rates before and after. Then extend the thread only when the first group actually uses it.
A rollout worth copying
- Choose one expensive break: change orders, document control, complaints, supplier onboarding, or product-content release.
- Clean the minimum data needed to run that workflow well, with a named owner for every critical field.
- Use standard configuration first; custom work should answer a proven business need, not preserve an old habit.
- Measure adoption and cycle time in weeks, then widen to the adjacent department that reuses the same record.
This approach will not work when leadership wants a software installation but declines a process decision. It will not work when no one owns the master data, when every exception becomes a customization, or when suppliers and commercial teams are invited only after engineering declares victory. Nor is a unified suite automatically superior. A company with excellent specialist systems and disciplined integrations may prefer depth over consolidation.
Propel's strongest idea remains its oldest: the value of product data rises as more of the company can safely use it. In 2015 that meant putting PLM next to CRM. In 2026 it means letting an AI assistant ask the product record a question without losing permissions, provenance, or context. Between those moments sits the less glamorous work - revisions, approvals, training, suppliers, labels, and a thousand small handoffs. That is exactly why the idea matters.