Subject line assistance offers two new beta features we're trialling with a select group of customers.
The inclusion of subject line rating and similar previous subjects aims to help you make a more informed decision regarding what subject line to use with your campaign.
Every organization has different goals and is working with its own unique set of subscribers, so both of these tools provide insights based on your subscribers? historical engagement with your campaigns.
Subject line rating
As you work on your new campaign subject, our subject line rating system will perform real-time checks against a machine-learning-based model, trained with your set of previous subject lines and their open rates. "Good one" indicates the subject is likely to produce a good open rate by your historical standards, while "Revise it" means the subject could do with more work.
No predictive ranking system is perfect. There are many variables e.g. seasonality, varying recipient lists, the from name and email, time of send, natural statistical fluctuation, and though the model is trained against your sending history, a good rating is no guarantee of a great open rate.
Our system is only intended to provide additional information, not to replace your judgement as an email marketer.
Similar previous subjects
To compliment the subject rating system, you can also click on the down arrow to expand the subject assistance tooling to get a sense of how your campaigns with similar subject lines have performed in the past.
It will show you a maximum of the three most similar previous subject lines, along with their open rates. This will help you to get a sense of how well that kind of subject has worked to engage your recipients in the past, or whether you might actually be breaking new ground.
Again, this is not intended to replace your judgement as a marketer, but to supplement your instincts with some qualitative data and gain extra confidence in your choices.
Frequently asked questions
Why can't I see the subject line tools in my account?
Subject line assistance tooling is a beta feature that we're trialling with a small, randomized group of customers. If the feature proves to be sufficiently useful we will seek to expand the rollout to more customers.
How do your subject-line ratings actually work? What do "Good one" and "Revise it" mean?
Our ratings system's answer is an educated guess about how well your subject line will engage your recipients, relative to your body of previous campaigns. The answer comes by running the subject line through a complex machine-learning model trained independently on your campaign history.
If the system determines that it's likely to fall in the top 25% of campaigns, the response is "Good one", otherwise we respond with "Revise it".
Your subject line rating system said "Good one", but I sent it and the open rate really wasn't very good.
Prediction systems always come with a degree of uncertainty, so we can't promise a perfect correlation between the rating and eventual open rate. There is so much which varies from campaign to campaign e.g. the recipient lists, the way the content affects inbox placement, seasonal and time-of-day effects, and so on.
All this means that some campaigns will have results which don't perfectly match the initial rating, so be careful with your expectations. A subject-line rating is not intended to replace your sense of what your recipients will respond to, but is one piece of information to help you in making that decision.
The rating system says "Revise it" for a subject line which got a really good open rate in the past. Shouldn't it say "Good one"?
See previous answer.
I sent a campaign yesterday with a really similar subject, why doesn't it show up in the "Similar subject lines" section?
After sending it can take many hours for your recipients to see and open your campaign, so open rates can be very misleading in those early hours and days. To avoid confusion, new campaigns won't be included in the "similar previous subjects" results set until 7 days have passed.
We're using the exact subject line we used in a small campaign last month, why doesn't it show up in the "similar previous subjects" set?
Small campaigns can have oddly high or low open rates, due to natural statistical variation, or because they were test campaigns that you and your colleagues were interacting with.
To avoid misleading conclusions, we don't include a campaign in the results set unless it was sent to at least 50 people.
How does your system determine that a previous campaign has a "similar" subject line?
Subject line similarity is determined on the basis of "linguistic adjacency". This goes behind a simple synonym check, looking for related concepts and topic areas, mimicking the way humans process and connect language.