They Had The Brand. Then They Burned It For The Valuation.
Most brands don't fall apart because the product stops working. They fall apart when someone asks, "are we keeping up?" at the wrong moment and the answer turns into action before it turns into clarity.
Grammarly is a good example of that. 14 years of building something people trusted. A clear throughline — helping people communicate in their own voice, with confidence. Then ChatGPT arrived. The pressure to evolve kicked in. The response was expansion. New acquisitions. New leadership. A broader ambition.
On paper, the business still looks healthy. Year-on-year growth, though, has slowed over the same period the brand began to change. A complete rebrand gave Grammarly an identity crisis it hasn't recovered from. The brand feels less certain of what it is — trying to be everything, everywhere, all at once.
That gap between performance and perception is what this case study looks at and what should have been caught earlier.
A brand built to last is one that starts with a sustainable answer to a recurring problem. Max Lytvyn, Alex Shevchenko, and Dmytro Lider started Grammarly because they lived one such problem — 3 non-native English speakers who understood exactly how much psychological weight sits behind trying to sound credible in a language that isn't your strong suit.
Before Grammarly, they had built MyDropBox, a plagiarism detection tool for universities. When they sold it in 2008, they asked the more interesting question: why do students plagiarise in the first place? The answer wasn't laziness. It was difficulty. So they built something that made expression easier.
They bootstrapped for 8 years. No salary until 2011. No outside funding until 2017. By the time they raised $110M, they were already profitable with a million daily users. People loved Grammarly because they felt seen. Somebody had noticed their daily struggle, acknowledged it and fixed it.
When the gap between what a brand says it is and what the audience actually feels when they use it grows wide enough, no rebrand can close it. You have to go back to the roots.
Grammarly's roots were clear. 'Make communication feel less scary.' For people writing in a second language. For professionals who can't afford to sound sloppy. For students who have something to say but not always the confidence to say it cleanly. That was the throughline. And it was working.
GrammarlyGO, launched in 2023, crossed a line. The product was no longer making your writing better. It was writing for you. It replaced you. That's a different product promise wearing the same brand name. The original user — the one who wanted to sound like themselves, just more confidently — got told their voice was now optional.
There's also a deeper irony here. Superhuman's own brand philosophy is "AI is not the thing that's super, it's you" — but their product is moving in the exact opposite direction. They're saying humans first while building tools that make the human less necessary. That same tension is now showing up in Grammarly. It's become a credibility gap.
Then came Expert Review in 2025. An AI feature generating feedback under the names of real authors — living and dead — without their consent. It was impersonating writers. On a platform that writers had trusted for a decade. A class action lawsuit followed. The feature was pulled but the damage was done.
Consumer mindshare dropped from 16.7% to 11.4% in 2026. That number doesn't make headlines but signals a shift in how the brand is remembered. In the same window, year-on-year growth slowed from 40.7% in 2024 to a projected 19% in 2026 — an early sign that what the brand is known for is no longer keeping up with what it offers.
| Metric | 2024 | 2026 | Direction |
|---|---|---|---|
| Annualised Revenue | ~$252M | >$700M | ↑ Up |
| Daily Active Users | ~30M | ~40M | ↑ Up |
| B2B Share of New Revenue | ~30% | ~50% | ↑ Up |
| Enterprise Growth YoY | ~35% | ~53% | ↑ Up |
| Consumer Mindshare | 16.7% | 11.4% | ↓ Down |
| B2C Subscriber Retention | Moderate | High Churn | ↓ Down |
Hover over any year — the divergence is the whole argument
What's happening here isn't random — it's a shift in where the business is getting its momentum from.
Earlier, Grammarly grew because individuals picked it up first and then carried it into their workplaces. That created a loop where consumer usage and enterprise adoption fed each other.
Now the growth is being driven more from the top. Enterprise expansion is easier to push — through integrations, team-wide rollouts, and positioning that fits neatly into existing software stacks. That's why revenue scales faster even without a proportional increase in individual attachment to the product.
The consumer side behaves differently. It depends on consistent, voluntary usage. When that weakens, it doesn't collapse overnight — it fades. People use it less, stop renewing, and stop recommending it.
So what you're seeing isn't conflicting signals. It's a shift. Growth is being driven more by enterprise expansion, while the consumer side is gradually losing its pull.
As enterprise growth takes over and consumer pull weakens, the model becomes more dependent on top-down expansion. That creates 3 compounding risks:
Sales-led growth is expensive by design.
Bottom-up adoption is nearly free. A user discovers the product, gets hooked, recommends it and the company follows. Flip that model and suddenly you're paying for sales cycles, procurement approvals and integration costs just to get in the door. CAC climbs. Margins compress. The process is painfully slow.
Contracts aren't the same as customer loyalty.
Loyalty comes from habit and dependency. When that erodes, what's left? Copilot is embedded across the Microsoft 365 suite — Word, Outlook, Teams, Excel. Google Workspace AI runs across Docs and Gmail. Both are bundled with suites companies already pay for. Why would a CFO approve another $30–$40 per user every month for tools like Grammarly? That's the procurement conversation happening right now. And once you're in a feature comparison against something that's already in the stack, you've already lost the positioning battle. You're not selling value anymore — you're defending price.
Specialised tools are winning on depth.
Jasper owns marketing copy. DeepL owns translation. Hemingway owns editorial clarity. None of them are trying to be everything — and that's exactly why they're winning. As Grammarly broadens its promise, it gets harder to stand for anything. The wider the positioning, the thinner the conviction. And in a market where specialists are eating the edges, being the generalist is a vulnerability.
The solution was to go bigger. Acquire Coda. Acquire Superhuman Mail. Install a new CEO. Rename the entire parent company after a smaller product most of their users had never heard of. Build a new visual identity. "Hero," the evolved cursor mascot. Commission a custom typeface. Launch "Superhuman Go" as the all-purpose AI agent.
On a craft level, some of this is genuinely well executed. The visual system is cohesive. The mascot concept has internal logic. But a grand rebrand without intention is not a strategy. It's a signal that you've run out of one.
The CEO acknowledged publicly that most customers won't even use the new name — the same thing that happened when Twitter became X. You can change the sign on the door. You can't change what 40 million people say to each other.
The Expert Review lawsuit is still in motion. The trust wound with writers hasn't been addressed with a named public position or a visible change in product philosophy. It's been quietly archived. Audiences notice avoidance. Especially the ones you hurt.
Grammarly still has 40 million daily users, 96% Fortune 500 penetration and a brand name people use as a verb even when the product is called something else. The throughline isn't gone. It's buried. Here's how I'd dig it out — and why each move directly protects revenue.
Grammarly doesn't beat ChatGPT by generating better content. It beats every generative tool by being the one people actually trust with their real words. The shift: from "AI writing assistant" to "The fastest way to write clearly in your own voice." That's not just a tagline — it's a direct attack on every generative tool's core vulnerability. None of them can say it credibly. Microsoft and Google can't. Grammarly still can — but only if the product and messaging actually reflect it.
Rewrite homepage copy around clarity and confidence, not feature lists. Retrain the product UI to frame suggestions as refinements, not replacements. Campaign direction: Sound like yourself. Just better.
Users who feel the product amplifies their voice stay longer and churn less. Clear differentiation reduces acquisition confusion, lowering CAC. Higher trust at activation means more users reach the "aha" moment before they drop off.
The creative community didn't quietly forget Expert Review. They talked about it — and silence read as indifference. Indifference is what turns a trust problem into a retention problem. The fix isn't an apology. It's a clear, visible commitment to writer's consent built into how the product works. Something the creative community can point to and say: they actually changed something.
Surface a writer's consent toggle prominently in settings. Make it the lead message in re-engagement campaigns to lapsed creative users. Have the brand acknowledge it directly, in its own voice, without corporate hedging.
Lapsed creative users have high LTV. They're vocal, they recommend, and they're the exact people who later advocate for Grammarly Business inside companies. Winning them back is cheaper than finding new ones. Letting them stay gone compounds churn quietly.
The person who advocates for Grammarly Business inside their company is almost always someone who already loves it personally. That's how bottom-up adoption works. The individual emotional story drives the pipeline — the non-native English speaker closing her first international client, the founder sending his fundraising pitch without second-guessing every sentence. Those are the people who later sit in a meeting, someone asks what writing tools the team uses, and they say Grammarly without being asked.
Run these stories in paid social and on landing pages. Make onboarding feel personal, not feature-heavy. Design the free-to-paid conversion moment around a specific win the user just had — not a paywall.
Bottom-up adoption is lower CAC than top-down enterprise sales. Individual champions convert teams. Teams convert procurement. The emotional narrative at the individual level is the top of a revenue funnel most B2B dashboards don't measure until it's already thinning.
Naming the parent company after a smaller acquisition doesn't signal a stronger business. It signals an identity crisis. Coda and Superhuman Mail can coexist inside a product suite — but the parent brand needs to be built for that role. Either keep Grammarly at the top or create a new parent brand, one that earns its place rather than inheriting it from something smaller.
Run a naming and brand architecture audit. Map how the current structure appears across site, product and sales materials. Identify where confusion is costing deals or slowing onboarding. Fix the highest-friction points first.
Brand confusion lengthens enterprise sales cycles. Procurement teams google the company name. Investors read brand coherence as a signal of strategic clarity. A clean architecture makes every downstream commercial motion easier and faster.
Revenue and DAU look fine until they don't. NPS slipping in the consumer segment, creative communities turning hostile — those move months before any dashboard catches up. Are writers defending Grammarly publicly again? Is the consumer NPS recovering? Are individuals inside companies advocating for it without being asked? These are early indicators of whether the pipeline is filling or quietly draining at the bottom where no enterprise report shows it.
Build an internal signal dashboard that tracks sentiment, advocacy behaviour and consumer NPS alongside commercial metrics. Weight it in quarterly reviews. Treat a drop in individual advocacy as a leading indicator of future enterprise churn.
Brands that only track scale catch problems after they've compounded. The individual user base at the bottom of the funnel dries up slowly and invisibly. By the time it shows in revenue, the fix costs 10x more than it would have a year earlier.