You've got just shy of 5000 users, but how active are they? How many daily/weekly/monthly active users do you have?
If you've got a decent number of active users then those are people that can potentially become paying users. So ditch the unlimited free version. Yes - ditch it.
Look at your weekly active users and see how many words they each translate. Presumably there is some kind of distribution, where 80% of users translate <100 words per week, then you have some power users that translate 1000 per week. Start with them and start telling them they need to pay to keep using Readlang once they have translated 80% of the words that they usually translate. A nice pop-up with a simple call-to-action asking them to pay. A:B test the pop-up message. You've then got dynamic pricing that is based on individuals' use habits.
Start with the power users and work backwards towards your regular users. See what happens as you get closer towards the 100 words/week guys.
If completely removing the free option is too aggressive for you, maybe once they reach the word limit they can dismiss the "pay to continue" pop-up but then it comes back every 5 words they translate after that. Try both strategies - experiment - see what works best.
You've only made $1k so far. I don't think there is any harm in being a bit more aggressive. You've build a good product - now is the time to test monetisation rather than adding features.
On the CRM side, you have done a little bit with Mailchimp, and that the reminders are opt-in. Make them opt-out.
Ditch Mailchimp. Make a $50 investment in customer.io - that will allow you to do CRM against 5000 users. Hit them with one email to all (which will use up half your 10k email allowance) with the most compelling email you can draft. Then use your remaining 5k emails to hit those that responded to your first email regularly and based on their actions (customer.io is really good when you get that right). /edit - just noticed extra emails on customer.io are really cheap so it's the user number limitation you really pay for. In that case send as many emails as you need to do some very active CRM experimentation for a whole month.
Reminds users that they have a text they have just started but not looked at again.
Remind them they need to do flashcards everyday
Remember - if nobody complains about the price of your product, it's probably too cheap.
Lastly, make an iPhone app that works offline and allows translation of single words via an offline dictionary and includes the flash-cards which then syncs with your account when back online. I'd pay for that! ;)
Rough number of users signing in at the moment:
~110 / day
~360 / week
~900 / month
Google Analytics RETURNING visitors per month:
~1000 returning uniques
~3800 returning visits
28 min avg visit duration
I'll have a serious think about your pricing advice. I'm plan on increasing the price but removing the free plan scares me. I'm sure it will increase revenue in short term, but don't you think it would seriously harm growth? My current user base was formed largely due to people blogging and sharing on forums and social media. Isn't there a risk this would dry up without a very usable free plan?
I completely agree about sending more emails. It's been on my todo list for a while, I already have Amazon SES set up so I'm planning to roll my own solution to send lifecycle emails to boost engagement. ($50 / month sounds too much, although now I'm tempted to try the free plan just to get an idea of what I'd be missing.)
iPhone app - one thing at a time! I want to polish and optimise the current site first, but I agree an iOS app would be awesome.
Note: Readlang works well on mobiles as a web-app, on both Android and iOS you can bookmark it as a "homescreen app" and it then runs in full screen mode similar to a native app.
wrt to CRM. Try to consider the $50 as an investment. You only need to add X paying subscribers in one month to get a ROI. If you get them, then anything after that is profit. If you don't, then you have learned a nice lesson for only $50.
wrt to pricing - just experiment - take my least aggressive suggestion - allow users to dismiss the notice. Or if you can, split test it.
I think one of the other commenters here is right - you can probably build more of a sense of community around it. Perhaps it can be the place where people really recommend good books for language learners. You already have the voting etc. That then maybe leads you in to the publisher deals, which I don't think would be as difficult as you imagine.
2nd I travel alot - so I'm often not able to access a data connection, or I'm on a plane, or I'm on the London Underground. This is probably not uncommon for people that are learning languages. Even when on a train 3G connection is pretty inconsistent. Commuting is when I'm most likely to use Readlang. I'd say the captive audience on London Underground everyday is worth it alone.
I've already discussed this offline with OP, so I know that part of the issue is that you can never get the sentence translating power of Google Translate offline, but single word translations (plus the flashcards) might be possible.
You only have to look at the Duolingo community to see how eagerly anticipated the offline ability was. Same for Memrise.