Comment by Silhouette | Hacker News Reader

Comment by Silhouette | Hacker News Reader

Parent Full thread

I normally don't send more than 2 emails a week for the first one month and then this number decreases. I'm wondering what others do here.

FWIW, I'd have killfiled you long before you reached that one month mark if you were sending that many e-mails just because I signed up for something (other than intentionally joining a high traffic mailing list, obviously).

Given that basically everyone is going to send mail when someone first signs up and for subsequent events like charging a renewal, I think you have to be careful how many more things you send out unsolicited. Some use of lifecycle e-mails is reasonable, for example if you're offering a free trial and want to remind someone after a few days. Depending on what you're offering and who your market are, maybe the acceptable limit goes up or down quite a bit. But some of the things I've seen advocated on HN, even by otherwise reputable people, are way over the line into spamming IMHO.

I use a simple benchmark: if someone in my family signed up and got whatever mails I was thinking of sending out, would they mind? If so, I probably don't want to send it.

Actually, two emails a month is the upper limit, which we use for actively engaged new signups. With 2 emails a week in this category, we were able to get about 4 times more engagement and time spent on the app than single email a week.

Also, we measure when people use the emails we sent and it's generally of no use to send more emails, after two consecutive emails not being clicked on - so one can stop there.

I don't see why one should not send two or even more emails, when the user wants them. Users have understood our product better and it helped achieve what we define as activation, much faster.

On the other hand, If you were a normal trial signup who just registered and never actually used the product, we'd email you thrice in the first month (including the first email) and then once every 3/6 months. We're also thinking of sending some informative blog posts once in a while. This system isn't old enough yet to give significant data about the effectiveness of the mailers sent after the first month.

Thanks for clarifying, and for sharing some interesting data points. Again FWIW, what you've described there seems a lot more reasonable to me than the raw 2/week rate without as much context did.

I think the key thing that didn't come across before, at least as I read your previous post, was the "when the user wants them" qualifier. If you're sending material that is useful for genuine users and carefully timed for when it's likely to be of interest to them, that's great.

The ones that wind me up are the businesses that have your e-mail address because you tried their site for ten minutes or made a one-off purchase from their on-line store and then think they're welcome to send you their latest "special offers" every few days for the next several weeks with a bit of trivial customisation based on what you tried/bought before.

I agree and I think everyone would - we've all seen such businesses and no likes getting spammed.

Although I wonder if it makes sense financially to do that. It would be great to see if someone has data on this. I'm not really sure if one can actually improve conversions or whatever they want to improve by sending emails that way.

Reply on news.ycombinator.com