What is a good open rate for your email newsletter?

My long term average open rates are 50.6%, 54.2%, 52.5% and 58.1% respectively - that's for weekly, curated news. They certainly drop a little over time and I've discovered two of the key reasons so far: people skipping some weeks due to e-mail overload and increasingly strict spam filtering.

What's a little annoying about this at first is that e-mail advertising is typically sold by CPM. So let's say I have 20k subscribers on a list and I can sell for $20 CPM (not unrealistic at all in e-mail).. that's $400. But that's $400 whether my open rate is 20% or 60%.

The most eagle eyed advertisers will ask for historical click rates and open rates but you'd be surprised how many don't. So in a way I'd rather charge per click but this doesn't seem to be accepted practice in e-mail right now.

Last, I've not found the word "free" to be too much of a problem. It's pretty common in mails and doesn't appear to be much of a trigger unless you all-caps it or add exclamation marks (e.g. FREE!!)

What absolutely massacred one issue of a newsletter, though, was the "Do You Really Want to Be Making This Much Money When You're 50?" post that did the rounds on HN recently. I included it and that issue's open rate fell by over half! Why? It's about as spammy a title as you could get and set off a ton of bells. So now I do even more spam analysis with each issue I send ;-)

BTW, I'm always happy to answer questions about the e-mail newsletter game, so feel free. I could probably ramble here all day about it. It's fun!

My training list (need a better name for that) gets between 40 and 60% or so per mailing, long term average is 50%.

BCC typically gets about 40% on its "monthly" emails, although those are more like "semi-annually" these days -- didn't even send the Halloween one this year.

Clients' numbers are clients' numbers but, anecdotally, if you could wave a magic wand and get 40% from their lists they are interested in what you are selling and would like to subscribe to your newsletter.

Clicks are a much more reliable metric of engagement because a click is always a click. Email clients can't hide them, and people rarely click through by accident.

If you engineer your email to be read all in one sitting (no clicking), then there is another option, which is to do what print publications do--pay someone to survey your readership and provide a statistical estimate of your active readership. This also gives you the opportunity to learn more about them, which might enhance your value proposition for advertisers--if you can show that the "right people" read your newsletter.

I think open-rate is a terrible and meaningless metric unless there is a filter specifying @yahoo, @gmail, and other email clients that force you to explicitly open the emails. Even with this metric added in, there is still phone email clients and email settings where delete automatically sends the person to the next email.

Even after filtering the data open-rate is still pretty meaningless compared to conversion. Conversion is very easy number to measure, but extremely difficult to increase. I'd rather have a 10% open-rate with 25% conversion than 50% open-rate with 1% conversion. A high conversion-rate allows the company to discover their customers and their behavior. Increasing conversion is an art-form. Increasing open-rate is difficult as well, but it should be secondary to increasing conversion.

If you go with Beta, Weibull, or Gamma distributions, mean and standard deviation are the wrong vocabulary to use. Obviously you can't tell just by looking at the plot, but it's an easy test to show definitively if it's normally distributed or not. The difference here is meaningful, depending on how much you want to read into the distribution of open rates.

One important thing to note is that "open rate" is a very fuzzy metric. It usually works by seeing if the recipient loaded images in the email... But aside from iPad and iPhone, pretty much all email clients do not load images by default. If your message is mostly text, many people won't bother loading images and the "open rate" you're seeing reported is going to be a dramatic undercount. Perversely, crummy messages sent to the same list that are unreadable without the images loaded will get a higher reported open rate even though fewer people may actually be reading them.

(edit) Or do they simply send a link to webpage and then count hits?

Anyone who tries to answer it definitively is probably not going to give a good answer. You need to get context out of the situation before you can answer it well. Just like you wouldn't throw the ball to first base if you could try to turn a double play, you wouldn't expect similar open rates between a list made up of young tech savvy users and a list made up of senior citizens. You wouldn't expect similar open rates between a list that you have to opt-in to on a preferences page and a list that you are automatically signed up to when you register for a product.

Funny thing is, most people do make that claim. They are the same people that say their penis is bigger than it actually is.

I think there's too many variables to say this is too small and that's too big (I'm talking about email open rates, stay with me here). Any information you get here probably isn't even worth considering.

Don't believe the hype.