Last public equity mark 01
13000 USDm Diligence report
Grammarly
Grammarly refresh report
Grammarly has real scale and product pull, but the public evidence still supports a cautious track/research-more posture rather than a clear buy.
Coverage and disclosure
Public coverage is strongest on product, scale, financing, and legal-risk signals. Retention, margin, covenant, and concentration details remain under-disclosed.
Cover facts
Non-dilutive capital 02
1000 USDm Total raised 03
1400 USDm Daily users 04
40000000 users Organizations 05
50000 organizations Company profile
Grammarly is a late-stage private AI writing and productivity platform with a freemium consumer base, growing enterprise monetization, and a broader suite strategy tied to Superhuman and adjacent workflow products. The public record supports strong distribution and product utility, but the company still discloses limited financial, retention, and capital-structure detail.
- Website
- www.grammarly.com
- Founded
- 2009-01-01
- Founders
- Alex Shevchenko, Dmytro Lider, Max Lytvyn
- Founding location
- Kyiv, Ukraine
- Headquarters
- San Francisco, CA
- Product
- AI writing assistance, editing, tone, agents, and productivity workflow surfaces across browser, desktop, mobile, docs, and enterprise settings.
- Customers
- Consumers, students, teams, and large enterprises
- Business model
- Freemium consumer subscriptions plus enterprise contracts and newer suite/agent monetization
- Stage
- Late-stage private, pre-IPO
- Funding status
- Last widely cited equity valuation was $13B in 2021; in 2025 Grammarly announced $1B of non-dilutive CVF capital from General Catalyst.
Executive summary
Top strengths
- Massive consumer distribution and recognizable brand
- Clear monetization paths across freemium, enterprise, and suite expansion
- Demonstrated customer proof in enterprise and workflow use cases
Top risks
- Public financial and retention disclosure remains thin
- Legal / reputation risk from the current publicity-rights controversy
- Competitive pressure from bundled editors, LLMs, and platform substitutes
Open gaps
- Enterprise NRR, gross margin, and cohort retention are still not public.
- General Catalyst CVF covenant terms and downside structure are undisclosed.
- Post-rebrand KPI trajectory and concentration risk remain difficult to model.
- Public evidence still does not justify a stronger than track recommendation.
Contents
01Company Overview
1.1 What Grammarly is
Company — Grammarly is an AI writing and productivity platform.. This anchors the company identity and business model. Founding — The company dates to 2009 and a Kyiv origin story.. That origin frames the rest of the historical analysis. Stage — The business remains late-stage private and pre-IPO.. Stage matters for how we underwrite the risk/reward profile. Taken together, this what grammarly is section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CO001, CO002, CO003, CO004, CO005, CO011, CO012, CO013, CO014, CO015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Founded | 2009 | The company's origin story starts in Kyiv and is consistently dated to 2009. |
| Founders | Alex Shevchenko, Dmytro Lider, Max Lytvyn | Three founders are named across current and historical coverage. |
| Current product | AI writing assistant and productivity platform | The product spans writing help, agents, and workflow support. |
| Current stage | Late-stage private / pre-IPO | Public sources continue to describe the company as private and late-stage. |
| Mission | Improve lives by improving communication | The mission statement has remained stable across historical coverage. |
Direct company and press sources; factual snapshot with historical and current framing.
[CO001, CO002, CO003, CO004, CO005]| Aspect | Fact | Why it matters |
|---|---|---|
| Founders and board | Founder presence remains strategically important | Governance and key-person risk still matter. |
| General Catalyst | $1B CVF partner and financing counterparty | Its capital structure has material strategic influence. |
| Customers | Large consumer and enterprise base | Usage scale is the platform’s core economic stakeholder. |
| Employees | Distributed product and engineering teams | Retention and geo concentration affect execution. |
| Legal/public stakeholders | Courts, regulators, and the public record | They shape risk, reputation, and disclosure discipline. |
Stakeholder map connects financing, governance, users, and legal pressure points.
[CO011, CO012, CO013, CO014, CO015]A compact visual summary of the chapter evidence.
[CO001, CO002, CO003]1.2 Who runs it
CEO — Shishir Mehrotra leads the post-Coda integration.. Leadership continuity affects execution risk. Offices — The workforce is distributed across four key offices.. Geographic distribution is a resilience and coordination issue. Governance — Founders still matter to the story and the board.. Key-person and governance concentration remain relevant. Taken together, this who runs it section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CO006, CO007, CO008, CO009, CO010]
| Aspect | Fact | Why it matters |
|---|---|---|
| CEO | Shishir Mehrotra | The Coda acquisition brought a new CEO and a broader productivity narrative. |
| Board continuity | Founders remain visible on the board | Historical and finance coverage still references founder involvement. |
| Offices | San Francisco, Kyiv, New York City, Vancouver | The company remains geographically distributed. |
| Key-person dependence | High | A platform tied to a few public leaders carries execution concentration risk. |
| Material change | Coda + Grammarly integration | The leadership and product stack changed materially in 2025. |
Leadership snapshot based on company history and financing announcements.
[CO006, CO007, CO008, CO009, CO010]1.3 Why the current financing matters
Users — 40M+ daily users support the platform thesis.. Scale is what makes the CVF financing possible. Funding — $1B CVF capital and $13B historical equity frame the stack.. Capital structure and valuation are central to the refresh. Gaps — Retention, margin, and headcount remain undisclosed.. The missing metrics are the main diligence blockers. Taken together, this why the current financing matters section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
02Market Analysis
2.1 Where the market begins and ends
Boundary — The market is writing assistance plus adjacent productivity surfaces.. A tight boundary prevents an inflated TAM. Substitutes — Built-in editors, chatbots, and human editing are all real alternatives.. Substitutes keep pricing power in check. Fragmentation — The category is broad but not winner-take-all yet.. Fragmentation explains why sizing is uncertain. Taken together, this where the market begins and ends section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CM001, CM002, CM003, CM004, CM005]
| Aspect | Fact | Why it matters |
|---|---|---|
| Included | AI writing help, editing, tone, and workflow agents | This is the core addressable category. |
| Excluded | General-purpose chatbots without writing workflow | Those tools are adjacent, not identical. |
| Adjacencies | Email productivity, docs collaboration, and review automation | The category bleeds into broader productivity software. |
| Status quo | Built-in spellcheck and manual editing | These remain the main substitutes. |
| Substitute risk | Users can multi-home across several writing tools | Low switching costs make the market more competitive. |
Boundary table uses a narrow-but-practical market definition rather than a single broad TAM number.
[CM001, CM002, CM003, CM004, CM005]A compact visual summary of the chapter evidence.
[CM001, CM002, CM003]2.2 Who pays and why
Buyers — Consumers, teams, enterprises, and education each buy for different reasons.. The same product must serve multiple segments. Drivers — Speed, quality, scale, and trust are the core adoption drivers.. Those drivers make the ROI argument understandable. Channels — Freemium and lifecycle marketing still do most of the work.. That channel mix shapes the market thesis. Taken together, this who pays and why section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CM006, CM007, CM008, CM009, CM010, CM011, CM012, CM013, CM014, CM015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Consumers | Individual writers, students, and job seekers | They drive freemium acquisition and top-of-funnel scale. |
| Teams | Marketing, support, and operations teams | Teams buy for consistency and time savings. |
| Enterprise | Large organizations with governance needs | Enterprise budgets support higher ACV and admin features. |
| Education | Students and instructors | Education broadens the funnel and helps habit formation. |
| Payer complexity | User and payer are often different people | That complicates conversion and messaging. |
Segmentation combines company claims with partner/customer case-study evidence.
[CM006, CM007, CM008, CM009, CM010]| Aspect | Fact | Why it matters |
|---|---|---|
| Speed | Real-time editing reduces manual back-and-forth | Time savings are the most obvious ROI story. |
| Quality | Tone, clarity, and style improvements are visible quickly | The product shows value in the first use. |
| Scale | The product works across many apps and surfaces | Distribution breadth raises total adoption. |
| Security | Privacy promises lower procurement friction | Trust is a buying criterion in enterprise deals. |
| Education | AI literacy is becoming a job-market expectation | This widens the long-term demand pool. |
Driver table focuses on adoption catalysts, not valuation.
[CM011, CM012, CM013, CM014, CM015]2.3 Why sizing stays fuzzy
Constraints — Trust, policy, and budget friction slow adoption.. These frictions make top-down sizing too optimistic. Lenses — Usage, organization, education, and workflow all provide different size lenses.. Using several lenses is more defensible than one broad number. Gaps — Public TAM and retention data remain incomplete.. The market thesis should remain evidence-constrained. Taken together, this why sizing stays fuzzy section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
03Competitors
3.1 Who competes with Grammarly
Direct peers — Standalone writing tools and bundled editors occupy the same user problem.. This is the direct competitive set. Adjacents — Generic LLMs and workflow suites increasingly overlap.. Adjacents are the most realistic displacement path. Status quo — Native spellcheck and human editing still matter.. The status quo remains the largest substitute. Taken together, this who competes with grammarly section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CP001, CP002, CP003, CP004, CP005]
| Aspect | Fact | Why it matters |
|---|---|---|
| Bundled editors | Microsoft Office and Google Docs | These win on default distribution. |
| Standalone writing tools | ProWritingAid, LanguageTool, Getsolved | They compete on feature depth and price. |
| AI assistants | ChatGPT and general LLMs | They substitute for drafting and rewriting. |
| Workflow suites | Superhuman Go and broad productivity platforms | These can absorb Grammarly as a feature. |
| DIY/open source | Open-source alternatives and internal tooling | They pressure the low end of the market. |
Competitive landscape with direct, adjacent, and status-quo alternatives.
[CP001, CP002, CP003, CP004, CP005]A compact visual summary of the chapter evidence.
[CP001, CP002, CP003]3.2 How the product compares
Features — Inline editing, tone, and brand voice are still central.. Product breadth is one moat. Pricing — Free plus paid tiers keep the funnel broad.. Pricing power is constrained by alternatives. Distribution — Cross-app presence helps but also creates platform dependence.. Distribution is both advantage and risk. Taken together, this how the product compares section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CP006, CP007, CP008, CP009, CP010, CP011, CP012, CP013, CP014, CP015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Real-time edits | Grammarly still wins on inline, cross-app suggestions | This is the company's core product claim. |
| Brand voice | Tone and style controls remain a differentiator | Brand consistency is key for teams. |
| Agents | Competitors are adding agentic editing experiences | The feature gap is narrowing fast. |
| Plagiarism / originality | Some rivals bundle checkers; others do not | Trust and academic use cases depend on this. |
| Distribution | Browser, desktop, mobile, and docs coverage matter | Surface area is a moat and a battleground. |
Comparison table focuses on product scope and delivery surfaces.
[CP006, CP007, CP008, CP009, CP010]| Aspect | Fact | Why it matters |
|---|---|---|
| Free | $0 entry point | This keeps Grammarly at the top of the funnel. |
| Pro | $12/member/month annually or $30 monthly | This is the core self-serve monetization tier. |
| Enterprise | Custom pricing with admin and governance controls | This supports larger ACV and procurement. |
| Competitive pricing | Many substitutes are bundled or cheaper | Pricing pressure stays intense. |
| Switching | Multi-homing is common | Low switching costs weaken lock-in. |
Pricing table emphasizes how bundled and low-cost substitutes constrain the market.
[CP011, CP012, CP013, CP014, CP015]3.3 What can displace it
Lock-in — Switching costs are modest in consumer use cases.. Low lock-in keeps churn risk alive. Bundling — Bigger suites can bundle similar capability.. Bundling pressure is the main strategic threat. Future — The moat likely has to move up the workflow stack.. A feature-level moat is not enough. Taken together, this what can displace it section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
04Financials
4.1 How Grammarly makes money
Revenue — The business monetizes freemium users, teams, and enterprise buyers.. This underpins the business model. Pricing — Public pricing is simple but conversion is what matters.. The economics come from conversion and expansion. Growth — Growth uses lifecycle marketing and enterprise upsell.. Channel economics are central to the model. Taken together, this how grammarly makes money section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CI001, CI002, CI003, CI004, CI005, CI006, CI007, CI008, CI009, CI010]
| Aspect | Fact | Why it matters |
|---|---|---|
| Free | Freemium top-of-funnel | The free product drives acquisition. |
| Pro | Per-seat subscription | This is the core consumer monetization path. |
| Enterprise | Custom contracts and admin features | Enterprise expands ACV and retention. |
| Education | School and student plans | Education broadens usage and conversion. |
| Platform / AI | New agents and suite attach | The product is being monetized upward. |
Revenue mix remains inference-heavy because public segment disclosure is limited.
[CI001, CI002, CI003, CI004, CI005]| Aspect | Fact | Why it matters |
|---|---|---|
| Free | $0 | The free tier remains the primary acquisition wedge. |
| Pro annual | $12/member/month | Annual pricing anchors the self-serve model. |
| Pro monthly | $30/month | Monthly pricing provides an upcharge option. |
| Enterprise | Custom | Enterprise pricing is negotiated, not listed. |
| Upsell logic | Higher usage and trust drive upgrades | Conversion is tied to habit and ROI. |
This table is a pricing snapshot, not a revenue forecast.
[CI006, CI007, CI008, CI009, CI010]Bridge from free usage to enterprise monetization and higher-ARPU products.
[CI001, CI002, CI003]4.2 What is and is not disclosed
Proxies — The best public proxies are users, organizations, and email volumes.. The company is scale-rich but metric-light. Capital — The 2025 CVF deal adds capital without resetting valuation.. That changes the financial risk profile. Gaps — NRR, margin, and runway remain undisclosed.. These are the main diligence blockers. Taken together, this what is and is not disclosed section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CI011, CI012, CI013, CI014, CI015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Acquisition cost | Not publicly disclosed | Paid marketing spend is material but not itemized. |
| Payback | Not publicly disclosed | Public data is not enough to model payback directly. |
| Retention proxy | Weekly insights and product usage loops | Owned channels imply a retention flywheel. |
| Gross margin | Not publicly disclosed | Software economics likely help, but public proof is thin. |
| Working capital | Light relative to hardware businesses | The model should be capital-light absent acquisitions. |
Proxy table preserves the fact that direct unit-economics disclosure is not public.
[CI011, CI012, CI013, CI014, CI015]4.3 What the financing implies
Monetization — The model appears healthy enough to support growth capital.. That is a positive signal. Risk — Opaque covenants and missing margin data cap conviction.. Downside modeling stays difficult. Verdict — The financial picture is good but incomplete.. The chapter supports a cautious track posture. Taken together, this what the financing implies section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
05Product & Technology
5.1 What the product is
Surface — Grammarly now spans many writing surfaces.. It is more than a grammar checker. Agentic — Specialized agents are the core roadmap.. The platform is becoming agentic. Docs — A dedicated writing surface anchors the suite.. This adds control and retention. Taken together, this what the product is section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CE001, CE002, CE003, CE004, CE005, CE006, CE007, CE008, CE009, CE010]
| Aspect | Fact | Why it matters |
|---|---|---|
| Browser extension | Inline writing suggestions | Core distribution wedge. |
| Desktop app | Cross-app editing support | Extends the product beyond the browser. |
| Mobile | On-the-go writing support | Expands the workflow. |
| Docs | AI-native writing surface | A new product surface. |
| Agents | Specialized writing agents | The most strategic product evolution. |
Module table reflects the current product stack.
[CE001, CE002, CE003, CE004, CE005]| Aspect | Fact | Why it matters |
|---|---|---|
| Drafting | First-pass writing support | Core productivity use case. |
| Revision | Tone and clarity edits | This is where Grammarly adds value. |
| Originality | Plagiarism / AI detection | Trust and authenticity controls. |
| Research | Citation and evidence support | The new agentic layer expands scope. |
| Collaboration | Brand, team, and enterprise controls | Team workflow is part of the value proposition. |
Workflow table ties product features to customer jobs-to-be-done.
[CE006, CE007, CE008, CE009, CE010]The product spans browser, desktop, mobile, docs, and agent surfaces.
[CE001, CE002, CE003]5.2 How it works
Architecture — The product is a cross-app layer atop writing workflows.. Distribution is the operating model. Controls — Security and privacy are key to enterprise use.. Trust is a product feature. Quality — Latency and reliability remain central.. A lot rides on execution quality. Taken together, this how it works section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CE011, CE012, CE013, CE014, CE015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Cross-app layer | Works across apps and websites | Distribution is the operating model. |
| Agent layer | Task-specific agents on top of writing | Agentic expansion is the key evolution. |
| Docs layer | Dedicated writing surface | A new owned surface improves control. |
| Data layer | Writing and preference data | Data supports personalization. |
| Platform dependence | Browser/OS/platform permissions matter | Distribution is partly outside control. |
Architecture table summarizes the product stack without over-specifying internals.
[CE011, CE012, CE013, CE014, CE015]5.3 What could break
Roadmap — The moat depends on moving up the workflow stack.. Feature parity is a risk. Gaps — Model-level details are not public.. The technical moat is hard to underwrite without them. Taken together, this what could break section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
06Customers
6.1 Who buys and uses it
Segments — Consumers, teams, enterprise, and education all use the product.. Customer variety drives scale. Payers — The payer and user are often the same, but not always.. That affects conversion. Channels — Partners and embedded surfaces support reach.. Distribution matters. Taken together, this who buys and uses it section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CU001, CU002, CU003, CU004, CU005]
| Aspect | Fact | Why it matters |
|---|---|---|
| Consumers | Writers, students, and professionals | This is the widest top-of-funnel segment. |
| Teams | Marketing, support, operations | Team workflows support paid conversion. |
| Enterprise | Large organizations | Enterprise drives higher ACV. |
| Education | Students and instructors | Education broadens long-term adoption. |
| Channel / partner | Embedded workflows and platform partners | Distribution partners amplify reach. |
Segmentation remains broad because the product serves several customer groups.
[CU001, CU002, CU003, CU004, CU005]Named customers and public outcomes by segment.
[CU001, CU002, CU003]6.2 What adoption looks like
Scale — Users and organizations are both large.. Scale and enterprise proof coexist. Proof — Customer case studies show real ROI.. The product has visible business impact. Durability — Retention metrics are not public.. This is the biggest gap. Taken together, this what adoption looks like section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CU006, CU007, CU008, CU009, CU010, CU011, CU012, CU013, CU014, CU015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Daily users | 40M+ daily users | Reach is still growing at massive scale. |
| Organizations | 50,000+ organizations | Enterprise adoption is meaningful. |
| Institutions | 3,000 educational institutions | Education is a major wedge. |
| Apps | 500,000+ applications and websites | Ubiquity supports adoption. |
| Lifecycle | Billions of emails every year | Owned channels power retention. |
Adoption table uses public scale claims and customer case studies.
[CU006, CU007, CU008, CU009, CU010]| Aspect | Fact | Why it matters |
|---|---|---|
| Databricks | Saved $1.4M annually | Strong enterprise value proof. |
| OSV | 90%+ employee adoption | Broad internal adoption proof. |
| Salesforce case | 80% plan-upgrade lift | Shows conversion optimization impact. |
| Hightouch | Personalizes billions of emails | Demonstrates scale and automation. |
| UserEvidence | Storytelling engine scaled lean | Proof content can be operationalized. |
Customer proof is strongest in the enterprise and growth-marketing segments.
[CU011, CU012, CU013, CU014, CU015]6.3 Where risk concentrates
Expansion — Land-and-expand is plausible but not fully measurable.. Up-sell is a key thesis driver. Concentration — Top-customer concentration is not disclosed.. This leaves a material risk gap. Verdict — Customer adoption is strong, but the cohort math is not public.. Track the business, but keep gaps open. Taken together, this where risk concentrates section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
07Risks
7.1 Legal and regulatory risk
Consent — The lawsuit and expert-review controversy are the biggest near-term risks.. This can damage brand trust. Disclosure — AI disclosure and attribution remain sensitive.. This is both a legal and reputational issue. Privacy — Enterprise buyers care about data handling.. This can affect the sales cycle. Taken together, this legal and regulatory risk section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CR001, CR002, CR003, CR004, CR005, CR006, CR007, CR008, CR009, CR010]
| Aspect | Fact | Why it matters |
|---|---|---|
| Consent / publicity rights | High | Recent lawsuit coverage shows real exposure. |
| AI disclosure / attribution | High | The expert-review controversy is reputationally sensitive. |
| Privacy / data handling | Medium | Enterprise buyers care deeply about this. |
| IP / training data | Medium | The AI category is still legally unsettled. |
| Regulatory change | Medium | Future AI rules could require product changes. |
Legal risk table centers on the current publicity-rights dispute and AI governance concerns.
[CR001, CR002, CR003, CR004, CR005]| Aspect | Fact | Why it matters |
|---|---|---|
| Quality failures | Medium | Bugs and edge cases can damage trust. |
| Platform dependence | High | Browser / OS / app surfaces can change rules. |
| Security incidents | Medium | Security promises are part of the enterprise sale. |
| Support burden | Medium | Multi-surface products increase support complexity. |
| Brand/reputation | High | One product failure can travel fast. |
Operational risk is driven mostly by product trust and platform access.
[CR006, CR007, CR008, CR009, CR010]A compact visual summary of the chapter evidence.
[CR001, CR002, CR003]7.2 Operational and dependency risk
Platform — Browser and platform dependencies are structural.. This is a non-trivial risk. People — Leadership concentration and rebrand execution matter.. Execution can be slowed by complexity. Support — A broad product surface increases support burden.. Quality issues matter more at scale. Taken together, this operational and dependency risk section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CR011, CR012, CR013, CR014, CR015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Browser platforms | High | Core distribution depends on third-party ecosystems. |
| AI / model providers | Medium | Model access and pricing could change. |
| Acquisition capital | Medium | General Catalyst terms are opaque. |
| Channel partners | Medium | Lifecycle and CRM integrations matter. |
| Customer concentration | Medium | Large deals are helpful but not public. |
Dependency risks reflect the app/platform model and the CVF capital stack.
[CR011, CR012, CR013, CR014, CR015]7.3 What would break the thesis
Kill criteria — Prolonged legal failure or no enterprise retention evidence would be thesis-breakers.. These are the key downside triggers. Mitigations — Privacy, DLP, and suite integration are the main mitigants.. The business must keep trust high. Watch list — Legal updates, customer proofs, and product trust metrics should be watched closely.. Those are the fastest-moving signals. Taken together, this what would break the thesis section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
08Valuation
8.1 How to think about the stock
Thesis — Grammarly has real scale and product pull.. That supports a constructive baseline. Anti-thesis — Key metrics and legal risk remain opaque.. That limits conviction. Recommendation — Track rather than buy today.. The current evidence is not enough for an aggressive entry. Taken together, this how to think about the stock section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CV001, CV002, CV003, CV004, CV005, CV006, CV007, CV008, CV009, CV010]
| Aspect | Fact | Why it matters |
|---|---|---|
| Recommendation | Track | Public evidence is good but not complete. |
| Confidence | Medium | Key metrics remain undisclosed. |
| Risk rating | High | Legal and platform risks remain material. |
| Valuation stance | Expensive | The last public equity mark is still 2021-era. |
| Action | Wait for more proof | The setup is not yet compelling enough for a buy. |
This table is the final recommendation summary, not a full model.
[CV001, CV002, CV003, CV004, CV005]| Aspect | Fact | Why it matters |
|---|---|---|
| Thesis | Huge distribution and real product utility | The company has durable usage. |
| Thesis | Agentic suite expansion could widen ARPU | Product breadth is moving up the stack. |
| Anti-thesis | Legal and consent risk could impair the brand | The downside is asymmetric. |
| Anti-thesis | NRR and margin remain hidden | The unit economics remain underwritten. |
| Anti-thesis | Feature commoditization is accelerating | The moat must keep evolving. |
Thesis table keeps both sides of the debate visible.
[CV006, CV007, CV008, CV009, CV010]What supports and what blocks a buy decision.
[CV001, CV006, CV011]8.2 What the market is pricing
Capital — The 2025 CVF deal did not reset equity value.. The old valuation still anchors expectations. Comp set — Private-market and productivity comps imply premium expectations.. The market is still paying for the thesis. Risk — The current setup still looks expensive.. Entry discipline matters. Taken together, this what the market is pricing section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.[CV011, CV012, CV013, CV014, CV015]
| Aspect | Fact | Why it matters |
|---|---|---|
| Bull | Suite adoption and enterprise growth re-rate the business | Upside requires strong retention and lower legal drag. |
| Base | Good growth but still opaque metrics | This is the current evidence-weighted case. |
| Bear | Lawsuit / platform issues slow growth and compress valuation | Downside comes from trust and distribution failure. |
| Bull trigger | Public NRR / margin disclosure | Would materially improve underwriting. |
| Bear trigger | Prolonged adverse legal outcome | Would weaken the brand and business model. |
Scenarios are qualitative because key financial metrics are undisclosed.
[CV011, CV012, CV013, CV014, CV015]8.3 What could change the call
Bull case — Disclosure of retention and margin could improve the story.. That would increase confidence. Bear case — Legal or platform issues could impair value.. That would reduce the multiple. Diligence asks — NRR, margin, CVF terms, and legal status are the key asks.. These are the missing pieces. Taken together, this what could change the call section uses public evidence to triangulate the most defensible view, while keeping the remaining uncertainty explicit instead of smoothing it over. The intent is to preserve decision-useful detail, show where the public record is strong, and isolate the missing evidence that still controls the judgment. That framing should be read as a working diligence note rather than a final verdict, which helps keep the chapter usable when new evidence lands.
Appendix A: Decision snapshot
Grammarly remains a strong product and distribution asset, but the current public record leaves too many valuation and risk questions unanswered for a more aggressive rating.
Disclaimer
This report is based on public sources and current-run evidence only; it is not investment advice.
Evidence index
| ID | Statement | Confidence | Sources |
|---|---|---|---|
| CO001 | Founded: 2009 | Medium | SO001, SO002, SO003 |
| CO002 | Founders: Alex Shevchenko, Dmytro Lider, Max Lytvyn | Medium | SO001, SO002, SO003 |
| CO003 | Current product: AI writing assistant and productivity platform | Medium | SO001, SO002, SO003 |
| CO004 | Current stage: Late-stage private / pre-IPO | Medium | SO001, SO002, SO003 |
| CO005 | Mission: Improve lives by improving communication | Medium | SO001, SO002, SO003 |
| CO006 | CEO: Shishir Mehrotra | Medium | SO002, SO004, SO010 |
| CO007 | Board continuity: Founders remain visible on the board | Medium | SO002, SO004, SO010 |
| CO008 | Offices: San Francisco, Kyiv, New York City, Vancouver | Medium | SO002, SO004, SO010 |
| CO009 | Key-person dependence: High | Medium | SO002, SO004, SO010 |
| CO010 | Material change: Coda + Grammarly integration | Medium | SO002, SO004, SO010 |
| CO011 | Founders and board: Founder presence remains strategically important | Medium | SO001, SO002, SO004 |
| CO012 | General Catalyst: $1B CVF partner and financing counterparty | Medium | SO001, SO002, SO004 |
| CO013 | Customers: Large consumer and enterprise base | Medium | SO001, SO002, SO004 |
| CO014 | Employees: Distributed product and engineering teams | Medium | SO001, SO002, SO004 |
| CO015 | Legal/public stakeholders: Courts, regulators, and the public record | Medium | SO001, SO002, SO004 |
| CM001 | Included: AI writing help, editing, tone, and workflow agents | Medium | SM001, SM002, SM003 |
| CM002 | Excluded: General-purpose chatbots without writing workflow | Medium | SM001, SM002, SM003 |
| CM003 | Adjacencies: Email productivity, docs collaboration, and review automation | Medium | SM001, SM002, SM003 |
| CM004 | Status quo: Built-in spellcheck and manual editing | Medium | SM001, SM002, SM003 |
| CM005 | Substitute risk: Users can multi-home across several writing tools | Medium | SM001, SM002, SM003 |
| CM006 | Consumers: Individual writers, students, and job seekers | Medium | SM001, SM006, SM007 |
| CM007 | Teams: Marketing, support, and operations teams | Medium | SM001, SM006, SM007 |
| CM008 | Enterprise: Large organizations with governance needs | Medium | SM001, SM006, SM007 |
| CM009 | Education: Students and instructors | Medium | SM001, SM006, SM007 |
| CM010 | Payer complexity: User and payer are often different people | Medium | SM001, SM006, SM007 |
| CM011 | Speed: Real-time editing reduces manual back-and-forth | Medium | SM001, SM002, SM007, SM010 |
| CM012 | Quality: Tone, clarity, and style improvements are visible quickly | Medium | SM001, SM002, SM007, SM010 |
| CM013 | Scale: The product works across many apps and surfaces | Medium | SM001, SM002, SM007, SM010 |
| CM014 | Security: Privacy promises lower procurement friction | Medium | SM001, SM002, SM007, SM010 |
| CM015 | Education: AI literacy is becoming a job-market expectation | Medium | SM001, SM002, SM007, SM010 |
| CP001 | Bundled editors: Microsoft Office and Google Docs | Medium | SP001, SP003, SP004 |
| CP002 | Standalone writing tools: ProWritingAid, LanguageTool, Getsolved | Medium | SP001, SP003, SP004 |
| CP003 | AI assistants: ChatGPT and general LLMs | Medium | SP001, SP003, SP004 |
| CP004 | Workflow suites: Superhuman Go and broad productivity platforms | Medium | SP001, SP003, SP004 |
| CP005 | DIY/open source: Open-source alternatives and internal tooling | Medium | SP001, SP003, SP004 |
| CP006 | Real-time edits: Grammarly still wins on inline, cross-app suggestions | Medium | SP001, SP002, SP004, SP006 |
| CP007 | Brand voice: Tone and style controls remain a differentiator | Medium | SP001, SP002, SP004, SP006 |
| CP008 | Agents: Competitors are adding agentic editing experiences | Medium | SP001, SP002, SP004, SP006 |
| CP009 | Plagiarism / originality: Some rivals bundle checkers; others do not | Medium | SP001, SP002, SP004, SP006 |
| CP010 | Distribution: Browser, desktop, mobile, and docs coverage matter | Medium | SP001, SP002, SP004, SP006 |
| CP011 | Free: $0 entry point | Medium | SP001, SP004, SP005, SP006 |
| CP012 | Pro: $12/member/month annually or $30 monthly | Medium | SP001, SP004, SP005, SP006 |
| CP013 | Enterprise: Custom pricing with admin and governance controls | Medium | SP001, SP004, SP005, SP006 |
| CP014 | Competitive pricing: Many substitutes are bundled or cheaper | Medium | SP001, SP004, SP005, SP006 |
| CP015 | Switching: Multi-homing is common | Medium | SP001, SP004, SP005, SP006 |
| CI001 | Free: Freemium top-of-funnel | Medium | SI001, SI002, SI004 |
| CI002 | Pro: Per-seat subscription | Medium | SI001, SI002, SI004 |
| CI003 | Enterprise: Custom contracts and admin features | Medium | SI001, SI002, SI004 |
| CI004 | Education: School and student plans | Medium | SI001, SI002, SI004 |
| CI005 | Platform / AI: New agents and suite attach | Medium | SI001, SI002, SI004 |
| CI006 | Free: $0 | Medium | SI001, SI002, SI004 |
| CI007 | Pro annual: $12/member/month | Medium | SI001, SI002, SI004 |
| CI008 | Pro monthly: $30/month | Medium | SI001, SI002, SI004 |
| CI009 | Enterprise: Custom | Medium | SI001, SI002, SI004 |
| CI010 | Upsell logic: Higher usage and trust drive upgrades | Medium | SI001, SI002, SI004 |
| CI011 | Acquisition cost: Not publicly disclosed | Medium | SI003, SI004, SI005 |
| CI012 | Payback: Not publicly disclosed | Medium | SI003, SI004, SI005 |
| CI013 | Retention proxy: Weekly insights and product usage loops | Medium | SI003, SI004, SI005 |
| CI014 | Gross margin: Not publicly disclosed | Medium | SI003, SI004, SI005 |
| CI015 | Working capital: Light relative to hardware businesses | Medium | SI003, SI004, SI005 |
| CE001 | Browser extension: Inline writing suggestions | Medium | SE001, SE002, SE003 |
| CE002 | Desktop app: Cross-app editing support | Medium | SE001, SE002, SE003 |
| CE003 | Mobile: On-the-go writing support | Medium | SE001, SE002, SE003 |
| CE004 | Docs: AI-native writing surface | Medium | SE001, SE002, SE003 |
| CE005 | Agents: Specialized writing agents | Medium | SE001, SE002, SE003 |
| CE006 | Drafting: First-pass writing support | Medium | SE001, SE002, SE005 |
| CE007 | Revision: Tone and clarity edits | Medium | SE001, SE002, SE005 |
| CE008 | Originality: Plagiarism / AI detection | Medium | SE001, SE002, SE005 |
| CE009 | Research: Citation and evidence support | Medium | SE001, SE002, SE005 |
| CE010 | Collaboration: Brand, team, and enterprise controls | Medium | SE001, SE002, SE005 |
| CE011 | Cross-app layer: Works across apps and websites | Medium | SE001, SE002, SE003, SE004 |
| CE012 | Agent layer: Task-specific agents on top of writing | Medium | SE001, SE002, SE003, SE004 |
| CE013 | Docs layer: Dedicated writing surface | Medium | SE001, SE002, SE003, SE004 |
| CE014 | Data layer: Writing and preference data | Medium | SE001, SE002, SE003, SE004 |
| CE015 | Platform dependence: Browser/OS/platform permissions matter | Medium | SE001, SE002, SE003, SE004 |
| CU001 | Consumers: Writers, students, and professionals | Medium | SU001, SU002, SU004 |
| CU002 | Teams: Marketing, support, operations | Medium | SU001, SU002, SU004 |
| CU003 | Enterprise: Large organizations | Medium | SU001, SU002, SU004 |
| CU004 | Education: Students and instructors | Medium | SU001, SU002, SU004 |
| CU005 | Channel / partner: Embedded workflows and platform partners | Medium | SU001, SU002, SU004 |
| CU006 | Daily users: 40M+ daily users | Medium | SU001, SU003, SU004, SU009 |
| CU007 | Organizations: 50,000+ organizations | Medium | SU001, SU003, SU004, SU009 |
| CU008 | Institutions: 3,000 educational institutions | Medium | SU001, SU003, SU004, SU009 |
| CU009 | Apps: 500,000+ applications and websites | Medium | SU001, SU003, SU004, SU009 |
| CU010 | Lifecycle: Billions of emails every year | Medium | SU001, SU003, SU004, SU009 |
| CU011 | Databricks: Saved $1.4M annually | Medium | SU002, SU003, SU004, SU005 |
| CU012 | OSV: 90%+ employee adoption | Medium | SU002, SU003, SU004, SU005 |
| CU013 | Salesforce case: 80% plan-upgrade lift | Medium | SU002, SU003, SU004, SU005 |
| CU014 | Hightouch: Personalizes billions of emails | Medium | SU002, SU003, SU004, SU005 |
| CU015 | UserEvidence: Storytelling engine scaled lean | Medium | SU002, SU003, SU004, SU005 |
| CR001 | Consent / publicity rights: High | Medium | SR001, SR002, SR003 |
| CR002 | AI disclosure / attribution: High | Medium | SR001, SR002, SR003 |
| CR003 | Privacy / data handling: Medium | Medium | SR001, SR002, SR003 |
| CR004 | IP / training data: Medium | Medium | SR001, SR002, SR003 |
| CR005 | Regulatory change: Medium | Medium | SR001, SR002, SR003 |
| CR006 | Quality failures: Medium | Medium | SR004, SR005, SR007 |
| CR007 | Platform dependence: High | Medium | SR004, SR005, SR007 |
| CR008 | Security incidents: Medium | Medium | SR004, SR005, SR007 |
| CR009 | Support burden: Medium | Medium | SR004, SR005, SR007 |
| CR010 | Brand/reputation: High | Medium | SR004, SR005, SR007 |
| CR011 | Browser platforms: High | Medium | SR001, SR004, SR009, SR010 |
| CR012 | AI / model providers: Medium | Medium | SR001, SR004, SR009, SR010 |
| CR013 | Acquisition capital: Medium | Medium | SR001, SR004, SR009, SR010 |
| CR014 | Channel partners: Medium | Medium | SR001, SR004, SR009, SR010 |
| CR015 | Customer concentration: Medium | Medium | SR001, SR004, SR009, SR010 |
| CV001 | Recommendation: Track | Medium | SV001, SV002, SV003 |
| CV002 | Confidence: Medium | Medium | SV001, SV002, SV003 |
| CV003 | Risk rating: High | Medium | SV001, SV002, SV003 |
| CV004 | Valuation stance: Expensive | Medium | SV001, SV002, SV003 |
| CV005 | Action: Wait for more proof | Medium | SV001, SV002, SV003 |
| CV006 | Thesis: Huge distribution and real product utility | Medium | SV001, SV004, SV006 |
| CV007 | Thesis: Agentic suite expansion could widen ARPU | Medium | SV001, SV004, SV006 |
| CV008 | Anti-thesis: Legal and consent risk could impair the brand | Medium | SV001, SV004, SV006 |
| CV009 | Anti-thesis: NRR and margin remain hidden | Medium | SV001, SV004, SV006 |
| CV010 | Anti-thesis: Feature commoditization is accelerating | Medium | SV001, SV004, SV006 |
| CV011 | Bull: Suite adoption and enterprise growth re-rate the business | Medium | SV001, SV002, SV006, SV007 |
| CV012 | Base: Good growth but still opaque metrics | Medium | SV001, SV002, SV006, SV007 |
| CV013 | Bear: Lawsuit / platform issues slow growth and compress valuation | Medium | SV001, SV002, SV006, SV007 |
| CV014 | Bull trigger: Public NRR / margin disclosure | Medium | SV001, SV002, SV006, SV007 |
| CV015 | Bear trigger: Prolonged adverse legal outcome | Medium | SV001, SV002, SV006, SV007 |