Whenever I'm surprised by an investment announcement like this, I like to look the company name up on Google Trends. This curve will give you a sense of how rapidly user interest in Grammarly has increased the past few years: https://trends.google.com/trends/explore?date=all&q=grammarl...
It's no Snapchat or Instagram, but it's comparable to, say, Docker in magnitude and trajectory. I hear from a lot of writers through my own websites, and I hear them mention Grammarly more and more often lately.
That's not to say the $110M investment makes sense, just that Grammarly may be more popular than the community here realizes.
I could see why their daily user engagement would be really good but I wonder what their paid conversion rate is because I imagine it wouldn't be that great in the US.
The negative comments don't make any sense. It's hard for me to understand why a bunch of people who clearly don't understand the domain complain how more informed parties invest their money.
I use Grammarly daily. The subscription is cheap compared to hiring a proofreader, which I used to do. The market for ESL speakers is huge, and native speakers, too, can often benefit from improving their writing.
I can't make any general statements. I checked both tools against a document that was recently corrected by a human proofreader. Out of six errors found by the human, Grammarly identified three and proselint none.
"only half of the mass if accounted for by the stars", misspelling "if" -> "is"
"in the heath death scenario", misspelling "heath" -> "heat"
"The quantity follows Gaussian distribution.", missing article
It's about half as good as a non-specialist human proofreader (see my post above) for a fraction of the cost. While that's not sufficient for some documents, I consider it (in addition to my own eye for detail) good enough for things like client proposals and my academic work (grad student), not to mention all the real-time (chat, email) applications.
To those, who like me, felt a knee-jerk reaction to criticize:
If you are going to make a criticism, don't make a sarcastic or snarky comment.
Instead, make an argument about why there is no potential in a product that:
+ increases the quality & productivity of one of the most valuable and professional skills that's used in 99% of all jobs
+ has a potential customer base of over 1 billion people
+ has near-zero marginal costs
+ has increasing returns to scale due to its cost structure
+ has increasing returns to scale due to its product development being based on learning from user data
+ is based on a new machine learning paradigm that has recently made huge gains and has unknown potential
And just think about the value of writing. Some people make their entire living from writing (journalism, etc.). A large fraction of the population needs good writing skills in their job, just to share information and persuade others to act. If improving your writing - on items like your cover letter, your professional emails, your product - gives you a 1% boost to productivity and a 1% higher chance of a promotion, then this $110M is nothing compared to those percentages multiplied over the English-speaking population.
Of course, the whole idea could be a bust. But if you think so, make the argument why. Resist the temptation to post brief, snarky comments about how VCs are stupid.
Thank you for saying this. I would wager that a majority of people who comment can't think in these terms and resort to the easy way out - which is snark.
Just look at threads about Facebook, for example. It's usually just pure unsubstantiated hate and very little intelligent/interesting comments.
I'm skeptical not about an investment in them, but the size of the investment vs the size of the market that will pay a non trivial fee for it.
$12/month, but only if you prepay annually. It's $30 if you want month to month.
That makes you wonder if the high investment is then for something shady...they have indirect access to email content, for example. Or, something like AI hype.
- Requires complete access to all the data you type
- They don't own any platforms, and the product is potentially a core feature of a platform/service. For example if Google decides to roll their own grammar service and tightly integrate it with chrome, how would grammarly complete?
A "free" product (Chrome plugin). The checks for grammar aren't transient, per the privacy policy:
>User Content you transmit will be stored on our servers, which are located in the United States. When you use the Software on your computing device, User Content you save will be stored locally on that device and sync with our servers.
How well would adoption would be if more users realized (most) everything they type in a web browser is stored on a remote server? I also wonder how often employees have violated HIPAA or something similar via a careless Chrome extension installation like this one.
Maybe they have a solid security policy and won't use "User Content" for nefarious reasons, but this is one free product I'll avoid.
> I also wonder how often employees have violated HIPAA or something similar via a careless Chrome extension installation like this one.
If you are a HIPAA covered entity, you probably shouldn't be allowing anyone to have the ability to install any software, including browser extensions, on any workstation they can use to access PHI, and any centrally installed software should have a privacy and security review before being permitted.
Of course, I know that's often not true, but that's a high-level organizational failure.
I'm not versed in HIPAA or PII but I wonder if grammarly is stashing text typed into emails? How many business strategies, private conversations, and other such things, are they holding (and likely plan to parse with machine learning somewhere in their pitch deck... if they aren't doing it already)?
EDIT: apparently it's used with Gmail, so the answer is yes they are parsing business strategies, private conversations, and other personal things in their data pipeline with, quoted from a comment further down, "presumably a future play in adtech or AI".
I'd like to see data scrubbing/anonymisation disclosures become standard for services like this. Knowing there's a robust email or SSN scrubber would increase confidence, though for unstructured PII you have somewhat of a chicken and egg problem where you need training samples to build the robust anonymiser.
Grammarly advertises quite agressively, at least to my demographic. I pointed out to my brother, who installed the addon, that they necesarily must be making money off him, probably by selling his data. He said he didn't care. Taking a step back, why should he care? His data is completely free for him to hand out. If he gets an effective product in exchange, why shouldn't he take advantage of it?
This is already a problem in other parts of the world.
The Japanese government has rules on which IME (The software that sits between the alphabetical keyboard and the thousands of Japanese characters that are needed in normal text) public employees can use, since some of them draw on remote data and ML algorithms.
Baidu has a really good and popular Japanese IME which uses a large server side dataset and is problematic in official use for obvious reasons.
From Techcrunch [1]
"Grammarly has 6.9 million daily active users, most of whom use the service for free. The startup makes money on the users who pay $11.99 per month for help with sentence structure and vocabulary."
Unroll.me made some money from showing their users advertisements. Many of their users assumed that this was the service's only source of revenue. They also sold their user's data [1], prompting backlash from those users.
How can you be sure that a service, like Grammarly, isn't selling it's user's data, or will do so in the future?
Honestly this seems pretty valuable. Non-native English speakers can get the tone wrong in writing pretty often, and that can sometimes make the difference between whether you get a job or whether you get properly recognized for your work.
I love both Grammarly and Hemmingwayapp, which I use both daily. Also, I think there is a huge market for these type of apps in China. If I were assured there are no GFW issues with servers located in China, I'd buy it for all of my customer facing employees in our China office.
Professional proofreading involves more than grammar. Those two products are mostly used with Microsoft Word, still the primary document exchange format in publishing.
I wonder how many of the investors didn't read the latest Microsoft blog before tipping in their funds. Sure Word isn't a web browser, but it's install base is massive and it is still very relevant, even if it isn't web-based or written by a bunch of hipsters.
And I suspect (although don't know) that word online also got this functionality, or will at some stage in the future.
My bet, this is $110m up in flames in a few years' time.
As a non-native English writer, I use Grammarly every day. It's a good product. In fact, it gave corrections for this small message thrice. It's really particular about commas!
But when using with long articles it's grammar check isn't perfect. I hope they get better. Best of luck Grammarly!
ATTN: Grammarly only works on desktop computers (not smartphones or tablet devices). Please re-open this email on a supported device to use the features mentioned here.
They have lots of work to do. I see why they needed the money!
I don't contest the 300M+ pre-money valuation given the enormous potential market and the likelihood of locking in paying users for life, but can someone explain to me why they need 110M?
If they're going to be as profitable as their valuation implies, why give away so much of the company? There's no way they need that many engineers. Is it just for a massive marketing campaign?
I feel most people don't realize Grammarly stores everything on a remote server. Not sounding like an illiterate in a company email is nice, but how many confidential emails have been written in there? Probably too many.
They are supposedly profitable, but yeah, the market size seems limited. It's a freemium product. The upgrade is $30/month if you pay month to month, or $12/month if you prepay annually. Seems high for unbundled nature of it. Maybe someone is hoping Google buys them to put it in Google Docs, Chrome, Gmail, etc?
> The founders of Grammarly had eschewed venture capital since setting up the company in 2009. Grammarly said its business is profitable.
It's hard for us to say without knowing the actual details and business metrics, but it's quite silly for you to dismiss their 8 year old profitable business without knowing anything.
It's no Snapchat or Instagram, but it's comparable to, say, Docker in magnitude and trajectory. I hear from a lot of writers through my own websites, and I hear them mention Grammarly more and more often lately.
That's not to say the $110M investment makes sense, just that Grammarly may be more popular than the community here realizes.