The very first thing I saw was a clear sentence explaining exactly what Twilert was and why I should care. If I were the exploring type, I could scroll through a lot more, but the big "Try it now" button was inviting, and it delivered just what it promised.
The UI for trying out a search was actually fun, and the live preview feedback was fast and confidence-building. I knew where I was and where I was going, thanks to the breadcumbs at the top, and every step delivered a clear value exchange as it led me down the funnel.
By the time I reached the point where I could authenticate with Twitter and start using the service, I'd been shown what Twilert was, how it worked, and what I could expect. You hooked me in.
Expertly done. I'm going to use this as an example for others who want to know how to lead people from awareness to engagement smoothly and quickly.
However, the 'filter' endpoint gives you access to all tweets matching to your search, up to a certain limit. Unless you're tracking very commonly used words or brands like 'selfie' or 'ipad', you're likely to get all tweets you are interested in without ever reaching the limit.
cf. https://dev.twitter.com/docs/faq#6861
Also, the API limits are per authenticated user, not per app.
It should be noted, though, that the basic functionality that Twilert provides -- email alerts when a particular word or phrase appears in a tweet -- is pretty trivial to roll on your own using the Twitter API.
1. Paid approach: use some authorized firehose data resellers. That's expensive, but if there are already thousands of users, that will pay off.
2. Free approach: use poll-based solution to search for the keyword every X seconds/minutes. When new tweet appears in the search results after the last recorded tweet - that's your new tweet. This way it shouldn't miss anything and you don't need firehose. Of course, to overcome request throttling limits, you'll need to use multiple different credentials.
aside from the results being random it's an interesting idea. Better still if it tracked sentiment and charted it over time for me.