Product Update: Insights in 3 Clicks with Machine Learning

Product Update: Surface Contextual Insights in 3 Clicks or Less with SocketLabs Spotlight’s New Machine Learning (ML) Feature 

Machine Learning Features Now Available in SocketLabs Spotlight Email Analytics

We’re curious: How well can you track the performance of the many — hundreds, thousands! — of senders on your platform? Is it easy to proactively identify senders who might have email delivery problems, or is it more like trying to save a princess from a castle full of obstacles? 

With SocketLabs Spotlight’s newest feature set, our proprietary machine learning provides not only big-picture views for immediate visibility into problematic senders, but it also gives you already-calculated analysis on opportunities lost due to the nature of the sender’s problem in less than five clicks.

If you’re not a big blog reader, don’t worry. Here’s the TL;DR: Just watch this bad boy in action.

Go from 40,000 feet… 

Before we start showing off these machine learning (ML) capabilities, let’s look at the big picture we’re able to provide now that we’re powered by Snowflake

Imagine what you can do with a unified view of all your senders’ performance on one simple dashboard. Instead of navigating to individual accounts, spending 40 minutes pulling reports and analyzing them line by line to find an outlier — or worse, having to send another custom request to the dev team for that report you need — you can use SocketLabs Spotlight to sort all your subaccounts by the urgency of their needs

Looking at this view from StreamMonitor – which is but one click away after you log in — we can see this email service provider (ESP) has 861 customers, and 72 of those accounts have problems with either a High and Medium level of urgency.  

Where to start? Well, we can sort these subaccounts by issue count, urgency, or sending volume to see which accounts are experiencing the most urgent and impactful delivery issues. 

Looks like Donut Planet has 14 issues, but is only Medium urgency. This means those issues could cause performance damage, but not imminently. Let’s look for a bigger fish to fry first. 

Such as Dealz 4 You, which has a slightly lower issue count, but a High urgency indicator, suggesting this account is heading south and the impact to delivery has become critical. This is the best place to start our day to protect our platform’s (and customer’s) sender reputation.  

Typically, when you have hundreds of customers (who may have 100’s of sending domains or subaccounts of their own), it’s simply not possible to get an overview like this. You’d need to build a custom data warehouse to ingest all the data, which takes years and lots of money to do. Then you’d have to sit a business intelligence tool on top of it to actually analyze the data in a way that’s meaningful. Oh, let’s not forget, you need to get this internal project prioritized on the product roadmap over features that benefit the customers. NBD.  

That’s a lot of investment into an internal reporting feature that, let’s be honest: probably isn’t what you asked for and isn’t giving you nearly enough insight…if you can even trust what you’re seeing. 

Or you’ll need to run an individual query or report for each sender. Who has time for that? You certainly don’t. Which is why Spotlight guides you to the senders with the most impactful issues within a few clicks, so you can focus on what matters most: helping them fix the problem! 

…Down into to a Narrower View of Subaccount Performance… 

Alright, it’s time to jump into the machine-learning layer! Finally! 

Once you’ve identified a sender in need of a closer look, you can go into the next level of detail — right on the same dashboard – by simply clicking on the subaccount you’d like to learn more about. 

Understand what’s driving a sender’s issue…no spreadsheets required. 

Other email analytics platforms can give you raw data that can indicate a problem with hygiene. But at SocketLabs, you’ll see clearly there is a problem with list hygiene without even thinking about raw data, because our machine learning algorithms surfaced nine issues related to hygiene alone. 

Notice, too, in the lower right corner: This sender is considered High urgency because SocketLabs Spotlight’s detected a significant and unexpected change in performance. 

SocketLabs Spotlight surfaces anomalies in individual performance. This insight is not derived from an industry benchmark or best guess. You know just as well as we do: this is huge, folks. 

This sender’s unique performance during the last few sends is significantly lower than their previous performance. This is contextual information you can use to alert your sender of a growing issue. 

Spotlight can even tell you which provider(s) a sender is having issues with, so you don’t have to spend time digging through logs to figure out which mailbox providers (MBPs) are destinations of concern. 

Dive into diagnostic investigations…right within the UI. 

Speaking of logs, though…we know some of you enjoy digging into the details. It’s your happy place. We’ve got you covered there, too.  

Within just a few clicks, you can access a variety of reports, directly within the UI:  

  • Delivered & Failed Message Summary reports 
  • Complaints report, visualizing the top Mx’s where spam complaints have been reported for better trend-spotting. 
  • Engagement Tracking
  • Message Detail, zooming in to see what’s happened with each individual email that’s been sent – including the failure reason you need to request mitigation with a provider, and so much more. Bye, bye, Kibana! We’ll miss you (…sort of, not really).  

We should also mention there are near-endless ways to slice and dice your data.  

Reports can be sorted by Mx / mailbox provider, From domain, DKIM domain, IP address, MailingID, Hourly, SuperTags, and the external provider (such as SendGrid or Mailgun) if you use multiple. 

You can also filter by StreamScore, issue urgency, and volume to ensure you’re looking at the exact data set you need at any given moment. 

Got a report you’d like to share or a set of customers you need to check regularly? We feel that. All reports can be saved for later and/or shared with stakeholders. 

Automate Actions to Proactively Manage (and Protect!) Sender Reputation 

You can also use this information operationally. Noticing you have a handful of senders who are struggling with the same issue, or are performing worse than usual and you’re growing concerned about collateral damage? Gather like-senders and route them into an IP or pool with senders with similar performance so you can better protect your best senders from performance issues caused by others who are struggling. 

For even faster implementation, our Rule Engine tool lets you set thresholds based on things like StreamScore so you can automate re-routing when someone drops below your ideal score for the IP or pool’s reputation. You don’t even have to manually monitor for those anomalies and instead can set-and-relax rules to ensure anything approaching problematic behavior is routing appropriately to avoid damage proactively. 

…to Instant Impact Analysis at a Granular View. 

Remember how we said you can get to deep detail in just three clicks? Here we are. 

Say goodbye to tedious math and human error, compounded by the need to pull disparate reports on different sends at different times. Spotlight’s new machine learning feature instantly surfaces key insights you can bring to your sender to illustrate the severity of their issues. 

Each issue will be brought to your attention clearly, so you can see the amount of lost opportunity at every turn, for every one of your senders. These aren’t generalizations or bulk numbers for you to dig into deeper or investigate exactly where those bounces came from… 

It’s all right here for you to either pass on to your sender to justify the need to act, or take action yourself, depending on the severity of the issue you’re seeing. 

Let’s Shape the Future of Email Analytics, Together 

Remember when we asked what you thought you’d see in email in 2025? Overwhelmingly, people shared how they were ready to use AI in a way that made email performance easier to understand and optimize.  

We were ready, too, so why wait until 2025? 

One of the best parts of this machine learning feature of StreamMonitor is how flexible it is, which opens the door to so many more potential views, reports, and dashboards we can build to bring you exactly what it is you need to help your senders succeed. 

We’re currently looking for partners to help us collaborate on what those views are; interested in helping us make the future of email analytics even more powerful?  

If you send via SendGrid or Mailgun, or you want to switch to SocketLabs email, let’s talk and see how we can build it together. 

Rather jump right in? You can try Spotlight analytics free for 30 days. Tell us what works, what needs work, and what features we can build to make your senders (and your role!) more effective.  

Table of Contents