How Grammarly Grew to 6.9M Daily Users in 9 Years | Hacker News Reader

How Grammarly Grew to 6.9M Daily Users in 9 Years

producthabits.com

74 pointsallenleee47 comments

It's a shame its product sucks so bad. I see compliments here, but every test I've done comes up terribly short. Perhaps the bar is set rather low because people are comparing it to Microsoft Word's built-in tools, which are so incompetent they're easy to improve upon. And I say this a linguist and lexicographer who has written three language-related books. Many copyeditors I know, too, find themselves having to clean up Grammarly's work when an author uses it on an MS to decrease the amount of work/money needed for a copyeditor.

Do you have any examples of how terrible it can be?

I have been forced to use Grammarly in the past. If you are structuring a sentence loaded with terms specific to a field, it will very likely flag that sentence as being too wordy. Combine this with Canadian English, and you end up with a low signal-to-noise ratio.

This struck me as funny because "a sentence loaded with terms specific to a field" probably is too wordy.

Although it may be appropriate given the context and target audience.

I get what you are saying though.

> And I say this a linguist and lexicographer who has written three language-related books.

I know this is an easy mistake to make but in this context the irony is palpable.

What's the mistake?

It should be "I say this _as_ a linguist..."

I despise this software, but I think that it's intended for ESL users.

>And I say this a linguist and lexicographer who has written three language-related books.

Sounds like you're exactly the opposite of the target audience.

Besides, Grammarly would have suggested a missing "as".

> Sounds like you're exactly the opposite of the target audience.

Seriously. The magic for the mass market in a product like Grammarly is not handling all cases correctly, all the time, with lots of complexity; it's handling most cases cases correctly, most of the time, with minimal complexity.

I find tools like Grammarly, Hemingway App & MS Word are best used to highlight items you should take a 2nd look at after you've finished writing. Sometimes they point out flaws & other times I disagree with them.

I would hope none of their authors argue that any of those tools could replace a professional copywriter.

even more terrible is not using it and coming across as an ass in your comment

Grammarly is a great tool. I never forget to check my article with it before publishing. If you had asked me a few years ago how much would anyone pay for a Grammar-checking tool, I would have guessed it would be no more that $3-4 / month. Grammarly charges $29 a month. Of course, its value can be measured with the improved communication but $29 _is_ a lot. It's amazing that it convinced people that it's worth it.

High price point is the only reason why I haven't gotten around purchasing its premium plan even though I think it would very beneficial for me.

$30/month is a lot, I agree with you there, but there is also the quarterly billing for $20/month, and the annual one for $12/month. I highly suspect that the monthly billing price is a decoy price to make the $20/month look cheaper.

Personally, I've been a premium user for more than a year now, and it has become an invaluable tool for me, especially as a non-native English speaker. $20/month is not too cheap, but I feel like it's worth it for the boost in professionalism that all my writing gets.

That's a good point regarding the $29/month being a decoy price. By the way for anyone curious, that concept I think is known as anchor pricing or price anchoring, if you want to read up on it some more.

If English is your second language, then I highly recommend their premium plan. Both my wife and I love it and how much it helps us in writing better English and getting better at it (in terms of that 'last mile' learning which could take up to a decade for most people to get perfect).

This is how Grammarly's premium edition would have corrected your post:

> Grammarly is a great tool. I never forget to check my article ~~with~~/on (Incorrect correction, think 'with' makes more sense here) it before publishing. If you had asked me a few years ago how much would anyone pay for a Grammar-checking tool, I would have guessed it would be no more ~~that~~/than (Probably a typo on your end, rather than a grammatical mistake) $3-4 / month. Grammarly charges $29 a month. Of course, its value can be measured with the improved communication but $29 _is_ a lot. It's amazing that it convinced people that it's worth it.

>~~High~~/The high (probably the most common mistake non-native English speakers do) price point is the only reason why I haven't gotten around purchasing its premium plan even though I think it would very beneficial for me.

Honestly, this doesn't read like much of a recommendation to me.

I think a native English speaker would probably write "I always check my article with it before publishing" - reversing the negation avoids the awkward "never forget to".

Meanwhile, that last sentence reads like perfectly reasonable conversational English to me: omitting the "the" emphasises the "High price point" part.

It seems that Grammarly will help you avoid the very common ESL type English errors (missing definite article, verb-subject agreement etc etc), which makes it worthwhile for many people, but it isn't going to get you over that final 'native speaker' grade hump.

They often have a 55% off discount, reducing the cost to $5 a month (If you make one payment of $60). English is my second language, this is sometimes useful. I may switch to premium.

So funny, if the MS of old saw someone was charging $29 a month for a feature, not a product, they would not sit idly by.

Perhaps this can serve as future guidance for those who soon look to encroach on Google or FB turf who may be afraid to act due to anti-trust enforcement.

This article completely leaves off any mention of Adwords. Grammarly absolutely crushed it with Adwords on the content network. There was so much cheap inventory they were able to buy such as on dictionary sites.

There is a bit of a Y Combinator attitude against buying advertising to build a startup. If it ain't viral, then the product isn't good enough. Unfortunately, a grammar checker isn't exactly something you tell your friends about, but it is quite cheap to tell the world with cheap display ads on dictionary site garbage inventory. Advertising is a really good option when there is cheap inventory in the niche, and a profitable CPA can be established.

I'm also surprised that wasn't mentioned as they advertise a lot on YouTube too, they have over 200,000,000 combined views on their YouTube adverts. Paid advertising is a major component of their strategy.

The first rule of paid advertising, is don't talk about paid advertising. Customers don't like to think they "fell" for an ad, and the advertising is often secret sauce. It is pretty easy to duplicate a technical product, and actually reasonably easy to copy the advertising. The big difference is most tech/product guys look down on advertising. Also many make the mistake of pricing lower than the competitor which means lower margins to buy advertising, losing every ad auction, and never getting off the ground.

They were also very aggressive with their youtube ads this year. For an extended amount of time this year every second ad I saw on YouTube was them (I live in Sweden).

Was going to mention this. The first thing that came to mind when I saw the title of the submission was:

Lots and lots of advertising

I almost can't go a day on YouTube without seeing a Grammarly ad. And yes, certainly coming across it a lot on the content network too.

Good luck to them if it works. The premium version is relatively inexpensive, so it would be interesting to know (not as if we ever would!) how much on average they spend to get a premium customer.

It may take them far more than the $29.95 they charge a month, so they could be out of pocket for months on that customer. But then, the customers who pay up front for a year, plus all the other less expensive lead sources, more than likely make up the difference and then some.

I haven't been through their funnel so I don't know if they have any upsells, cross sells...etc. On the surface it appears that they basically only have one single product (even though they also have business and education versions of that product).

I do worry about businesses that don't really diversify their offerings, since it's more of a funnel than an actual business. But for many companies, I guess the goal is to make as much money as quickly as possible then cash out, rather than create something that will really be around for the long-term.

Anyway, good luck to them (again). But they're probably desperately hoping Google doesn't add similar functionality to Gmail/Chrome, and that Microsoft doesn't expand their grammar checking functionality in Word!

As a Grammarly user, the thing that's been most surprising is that there's no option for an API to integrate with other tools. I guess with the growth they've experienced, there's not a lot of pressure to expand it, but it seems like a world of opportunity. I'm sure there are some good reasons, like the editing experience or API abuse, but their tool simply isn't the best overall writing experience.

To be able to use Grammarly within Sublime text or other editors would be incredible. As it stands, because you're forced to copy and past content over into their editor, the workflow is the biggest drawback. It's really handy in textareas on the web, but I've struggled to integrate it into my writing workflow because of the copy/paste process. Writing mainly in Markdown doesn't make it any more elegant either.

There is an open source engine offering similar features that might be a better candidate for other products to integrate with - https://www.languagetool.org/

I've been meaning to have a play with it as a potential replacement for Grammarly before the special offer subscription of that is due to renew in a couple of months (though if I keep the special offer rate it is probably just as well for me to stick with Grammarly).

It is Java based which may put off a lot of projects from including it directly in their desktop or mobile products, but they could host their own instance and have their applications submit data to that. Assuming the extra admin (keeping the service running & up-to-date, and monitoring/enforcing API-key use if they don't want the service to become de-facto public once other developers notice it exists) and hosting costs would not be too high, of course.

When I was in a coding bootcamp, one of the other people in my cohort was able to get some access to an API for Grammarly, and we worked on a toy project to use the API to 'score' medium blogs:

http://grandma.space/#/

Site functionality is pretty much broken now, but it's been almost two years since it went online, so I'm surprised they haven't offered a public API by now.

I don't know if it's on purpose on not, but I agree with lack of a usable API.

API focused businesses don't own the customer.

https://www.accenture.com/us-en/insight-retail-customer-expe...

I assume Grammarly, via the plugin, can read every piece of data you use on the web. Emails, etc.

Are they upfront about their plans to use (or not) this data?

You (others) probably clicked "accept ToS" without fully reading them. Oh well. If its free then you're the product.

Yes, everything you type is sent to their servers AFAIK.

I would guess more than just what you type. Probably everything in text and textarea elements. So, if you reply to an email, it's all the text from the original sender in addition to what you type.

I didn't see it that way, this is rather scary.

A little-known fact is that Grammarly is a Common Lisp app: https://tech.grammarly.com/blog/running-lisp-in-production

They use clojure as well (instead?).

https://tech.grammarly.com/blog/building-etl-pipelines-with-...

https://www.grammarly.com/jobs/engineering/software-engineer...

While I love the detailed growth article, I've been disappointed by Grammarly recently.

It used to do a really good job fixing my notes, but I feel like it's detection ability dropped off a cliff recently. I've had some pieces with pretty obvious mistakes that it has previously caught not get highlighted on a regular basis.

I saw the same thing happen with FoxType's V1 product, and then they pivoted away from it and eventually closed.

I wonder if there is some issue in aggregating people's writing mistakes into AI that lowers its value over time?

I love the weekly report. You are more productive than x% of Grammarly users. You are more accurate than y%. X always makes me feel good, Y not so much.

This week I got 80% for accuracy! It's usually... lower.

If it didn't invade privacy, I wish I could know who I was competing against. I consistently get "You used more unique words than x% of Grammarly users." where x is always 95% or more, but I didn't think my vocabulary was top 5% territory.

Hm, I was half-expecting to see "by starting from 2500 users and managing 10% monthly growth" :)

But:

> By the time Grammarly transitioned to freemium, it was already profitable with millions of users—and could fund a freemium plan to drive even more new user acquisition.

It might be an apples to oranges comparison but does 6.9 million DAU seem low? My expectations for DAU were set by working in the game industry. Both for Zynga (back in the CityVille days) and for some mobile game companies. If by DAU they mean "people who pay $11 or more per month" then that does sound decent. But if that number is free and paid it seems a little low over all. Of course if you look at the stats for F2P, web, or mobile games that launched in 2008 maybe 6.9 million is still a good number for their DAU.

It's useless to me as I write in markdown in an app, not plain text in a web browser. How can this be useful? I don't really care about my Facebook posts, I care about my blog posts.

It works with MS office as well as having their own editor you can use. For unsupported apps, I just cut and paste it into their editor.

It's interesting that they started off selling to universities. This is the same strategy that Qualtrics used, where they started with Universities then moved into enterprise:

http://www.businessinsider.com/qualtrics-raises-180-million-...

I liked it and it improved my writing. I stopped using it after I noticed that it was leaving css classes in wysiwyg editors that were showing up on websites that I had edited.

Probably with their catchy, absolutely non-cringey video clips.

Used it for a month and got annoyed at pay2play popups. Chrome spellchecker will do.

Isn’t this a patent minefield?

Reply on news.ycombinator.com