I am working on the onboarding process for an enterprise webapp. One of the onboarding elements is a series of emails to help the user get up and running and effectively using the application. My personal experience has been that if a company pesters me too much after signing up, I tue out and just start deleting emails. Is there any evidence that helps point to an effective post sign up email frequency (i.e. daily, weekly, ...)?
I don't provide any evidence or proof of best practice, but I've noticed a good thing in TripAdvisor. I signed up over a year ago, didn't think much about it and got one or two e-mail. Since I didn't care, the e-mail stopped coming.
However, I started posting stuff to TripAdvisor after our latest combined trip of Iceland and New York (just for the record :)), TripAdvisor began to interact with me. There are e-mails of "hey, your latest review got a helpful vote" and "Did you know your review of X has 500 readers" and all of this kind of positive e-mail.
Great work by TripAdvisor, but it takes a lot of work to set up all these rules. But with a little web analytics, a SMTP service and a design logic you're well on your way.
Advice? Follow your users frequency. If they log in 4 times a week, you can send 2 e-mails. If they don't, let them be.
Generally, I'd go with the 'make sure they're still alive' principle:
Send them an email when they register. That can be a confirmation email or just a welcome email.
Send them more when they interact with the site, as Benny mentions. But not too many - lots of emails are annoying. If you really have to send lots, provide an option to turn them off.
Loss of interaction
If they stop interacting, stop emailing - until a point. After a couple of months (or whatever), you might want to send them another email saying 'Hey, are you still alive? Or don't you want your account?' Because if they no longer need the account and they confirm this, you can delete it from your servers and save some space.
My single biggest recommendation would be to think less about the emails themselves, and more about what you want them to get people doing.
For example, if it's impossible to get value out of your application without taking three initial steps (say, creating a project, uploading your logo, and sharing it with a client), then I would look to time the emails in accordance with when each of those might reasonably be accomplished, and only send the second after the first is done, and the third after the second, etc.
Remember, you're sending out emails to get people back into the app (with a purposeful activity), not just saying "hi."
Of course, that recommendation is trumped by an even bigger one, which is to set up anything and iterate based on how it performs. Fortunately, you can go even lower-fi than a typical A/B test simply by measuring opens/clicks/unsubscribes (not the most scientific approach, but a good first step!).