- Published by
- Posted on
- 5 Comments
By Zach and Elizabeth
A YC founder asked me this past summer what kind of stats he should expect for his email newsletter. I thought I would share our learnings publicly since we see a lot of data in our ad network for email newsletters. This is a 3 part series on open rate, subscribe/unsubscribe rate, and click-through-rate. Note: this is only data for email newsletters, not transactional emails. Moreover, the email newsletters in our network tend to be focused on content as opposed commercial-focused. This first post is about open rate.
Here is a plot of open rate for the newsletters in our ad network*. We filter/reject a lot of publishers to keep the quality high in our network, so this plot is essentially a bell curve of the open rates of high quality email newsletters.
Based on just eyeballing the skew of this graph, I’d say that 20% open rate (over which the vast majority of our publishers lie) is roughly what you should be shooting for. Anything above that is amazing.
We were curious, though, whether a larger list size affects open rate? Is it harder to retain a good open rate with a larger subscriber base?
We found there is no correlation between list size and open rate. Open rates don’t actually get worse (or better) as you increase in size. Disclaimer: we didn’t have enough data to include our largest publishers (100k+ subscribers), so we can’t say with any certainty that this continues to hold at the very large list levels.
Now, a few entrepreneurs have asked me if they should benchmark themselves to these email stats broken down by industry. Unfortunately, I don’t have enough data around that study to say whether you should seriously look at those numbers. For example, if we look at the Agriculture and Food Services category, these newsletters have an average open rate of ~24%. For simplicity, let’s pretend there were only 2 newsletters in this category. This could mean that one has a 25% open rate and another has a 23% open rate. It could also mean that one of them has an 47% open rate and the other has a 1% open rate. In other words, we don’t know how different the open rates are.
That said, if we use the standard deviation of publishers in our network to figure out how big of a difference the open rates are, we find a normalized stdev = 0.23, which is quite high (relative to 1). What this means in plain English is that the vast majority of our network is spread out by a lot and are not closely centered around average. With that in mind, I would not worry about the industry averages — just try to keep your open rate above 20%.
Now, if you are just starting to send email newsletters, you may find that at first you’ll have an amazing open rate, but as you send more email newsletters, your open rate will probably drop. Keep an eye on this — this is the number that tells you that people actually care about your emails. This is important because once your open rate starts dropping, it will continue to drop in a downward spiral. Even people who are opening may eventually start seeing your emails go to spam, because of other people on your list’s lack of engagement.
Assuming that you are sending to people who have legitimately subscribed to your list, low open rates start happening for two reasons:
1) People don’t want to read your emails anymore
This is obvious, but how to combat this? Unfortunately, limiting your emails or unsubscribing people from your emails is the best way to save your open rate from going to the pits.
2) Your emails are getting stuck in spam filters
There are certain key words that will often send your emails to spam. Check out: 100 Spam Trigger Words & Phrases To Avoid. For example, we see a number of people use the word “Free” in their newsletters, and this can often send emails to spam. If you are offering a product or service for free, try using phrases like “complimentary” or “at no charge”. Also, check out Mailchimp’s explanation on spam and spam filters.
Next week, we’ll publish some stats and insights on click-through-rates and unsubscribe rates.
*to the extent that we can measure — and across a given publisher’s lifetime in our network. We’re not able to measure for every publisher in our network.