In order for SharpSpring and our partners to be successful, we need to ensure that every email sent lands in the intended recipients’ inboxes. In order to do this, we strive to keep our sending pools of the highest quality and to keep a good reputation with ISPs like Google and Microsoft, who ultimately decide if any email should be delivered or not.
Our responsibility as a marketing automation platform and Email Service Provider is to make sure that the pools our customers send from are full of quality senders, since high bounces and spam complaints can hurt deliverability for all the other good senders on the platform. In order to deliver the highest level of deliverability to our customers, SharpSpring has developed a new list sampling tool on the back end of the application. This limits the potential for a bad email or a bad list to damage the reputations of good senders.
What is Sampling?
Sampling refers to the process of sending an email job to a portion of a list's recipients, waiting for a short time, and then evaluating statistics from that sample before sending the email to the rest of the list. We try to achieve a balance of delivering the email as fast as possible while also taking the necessary time to gather enough data to confidently determine if an email is potentially damaging to your sender reputation.
How Does It Work?
SharpSpring initiates an analysis with the following parameters to keep the best interests of our partners and clients in mind:
- Limit the sampling period to 20 minutes. Timely email sends are essential to clients, and there is no reason to sample if it results in dramatically delayed sends.
- Set sampling volume to at least 5,000 sends. We decided to set a sampling floor, as sends under a certain volume have limited effect on reputation, and without a larger sample size, it was difficult to obtain any meaningful data.
- Use the Spam For Unsubscribe Reason parameter. Due to the inconsistent nature of third-party spam complaints, we decided to use our own feedback loops in order to determine if recipients deem the email to be spam or not.
Our goal was to determine if we could prevent sends that fell in the bottom three percent - the ones that are truly damaging to reputations - before they had a chance to fully affect our senders. However, we had to be extremely sensitive to the risk of false positives, since any good sends we catch will be a missed opportunity to better our sending pools.
What Did We Find?
After detailed analysis, we found an approach that allows us to catch a substantial number of bottom three percent of sends and gives us 95% confidence that everything we caught would fall in the bottom 10% of a send with a volume of over 5,000. While there will alway be some risk of an outlier being caught, we are confident that only the widest of outlier would be wrongfully caught in our rejection parameters. With these findings, we have elected to maintain these parameters moving forward.
How Does This Affect You?
As a SharpSpring partner or client, you will see very little impact from this on your business. For larger email jobs over 5,000, you should expect a 20-minute delivery delay while we gather data for our list sampling. If there is an issue with your email or your list that falls within our rejection parameters, the admins of your SharpSpring instance will receive an email notification that the email has been stopped due ot exceeding the thresholds of our list sampling.
SharpSpring will still return the stats from sampling sends to the email jobs area as usual, so you can review the results and adjust your send as necessary. This can include changing the language in your email or reviewing your list quality. As always, you can reach out to our Deliver Support team or our Compliance team for any questions you might have. We are happy to provide advice on how to improve your emails. Once you have revised your email and fixed any issues, we'll invite you to resend it.
We are extremely excited about this new list sampling functionality, and we're looking forward to monitoring and tweaking it as we move forward. We understand how frustrating it is when a single bad send can have such a huge effect on your reputation status and how long it can take to make your way back up to 'good sender' status. List sampling is meant to mitigate that risk, and you can be confident that if a job is rejected, the tool has potentially saved you a huge headache and a ton of working winning back your sending reputation.
Our Compliance and Delivery Support teams will continue gathering list sampling data so we can adjust our system to minimize risk to our community pools and ensure you can confidently deliver content to your clients.