- Developer Hub
- SparkPost API
- Free Tools for Email Teams and Developers
- Slack Channel
- User Guides & Migrations
- Submit a Ticket
- SparkPost Academy
- Email Deliverability Resources
- Email Explained
- White Papers & Guides
- Webinars & Videos
- SparkPost vs. SendGrid
- Customers Stories
- Contact Us
We typically think of segmentation as a function of marketing. Marketers are the ones worried that one email’s content is relevant to a 25 year old woman in San Francisco while another is more appropriate for a 55 year old man in New York. While making content relevant to your audience is absolutely critical to long term marketing success, that’s not the type of segmentation I’m talking about here.
Deliverability experts say that if an individual hasn’t opened or clicked on your email in a certain amount of time, it is time to adjust the segment they are categorized into. You might want to send them a different type of message, not the same message you send to people who express interest by actively engaging with your emails. In fact, it might be time to take recipients who haven’t engaged in the past 6 months off your active list entirely.
Using SparkPost’s webhooks to get subscriber-level engagement data, you can update your database of record – or customer relationship management (CRM) system if you have one – as recipients open and click. As time of last engagement gets older, you can modify your content to those recipients with additional incentives to engage. Once a recipient reaches 6 months from the last open or click, you can send a “win-back” message. We’ve all gotten these: “Dear Customer, we want you back. But if you don’t want to hear from us again, please let us know.” Honoring these opt-out requests is important: it reduces the number of messages you send to recipients who may not want them. And removing those that don’t respond one way or the other from your active mailing list helps ensure you don’t hit spam traps and hurt your deliverability with ISPs.
To start, a simple engagement-based segmentation strategy might look like this:
- New recipients who signed up within the past 2 weeks: welcome messages, information to get started with your product or service, and then ongoing, regular contact. The type of content to send is of course specific to your product or service and ideally you know something about your customers to know what they would like. For example, existing paid customers might want more in depth information on HOW to use your product vs. prospects would like to understand WHY they should.
- Recipients who’ve engaged in the past 6 months: consistent, ongoing communication is key. This is where a content strategy, and understanding your audience is really important. Your goal is likely to convert recipients to becoming paying customers (if they’re not already.) And provide relevant, timely content to existing customer to keep them coming back.
- Recipients who haven’t engaged in 6 months or more: the goal is win them back. Or give them the chance to opt-out.
Another behavior-based strategy is segmenting based on time of day. If you know your customers engage with your email at a certain time of day (typical spikes are first thing in the morning in their time zone, lunchtime, and evenings), then striving to be at the top of their inbox at those times is likely to boost engagement.
The SparkPost user interface enables you to look at your data in 15-minute increments. This is in contrast to some other providers that show day as the smallest level of granularity. This allows you to see any obvious spikes in engagement and then using your webhooks to segment recipients into transmissions that send at those times based on when those recipients engage. For example, in the data below, we see a slight uptick in opens and clicks around noon. Using webhooks to understand which recipients are opening and clicking at that time, putting them into their own segment, and sending to them around noon will likely boost engagement of that group. Experimenting with different sending times for other recipients might yield another timing-based segment.
To learn more:
- Consuming SparkPost Webhook Events with Loggly – Part 1 of 3https://www.sparkpost.com/blog/consuming-sparkpost-webhook-events-loggly-part-1-3
- Creating Webhooks https://www.sparkpost.com/docs/create-webhooks
The Transactional Email News Digest
In this week’s transactional email news we’ve got a lot of helpful information on best practices, the pitfalls of using transactional email for promotional purposes, and some ideas on how to solve the problem of messages that show up in multiple places. Let’s start things off with our friend Chris Penn, VP of marketing technology at SHIFT Communications.
Chris makes the case that when it comes to emails that people are expecting – transactional notifications, newsletters, etc. – it’s smarter to use straightforward, utilitarian or even boring subject lines. Why? In a word, brand. You want recipients of, say, your newsletter about transactional email to recognize it when it shows up in their inbox. Changing up your “From:” line or “Subject:” line might seem like a way to grab attention, but over time you’re likely to see engagement rates slip. And as usual, what works for one business might not work for another, so do test and test again!
Guillermo Rauch, who’s CTO and co-founder of LearnBoost, and blogs at DevThought, proposes a solution to a problem that’s becoming more common as the popularity of social media and mobile apps grow. Say you get a new friend request on Facebook. The site generates a message that you’ll see on your timeline, and also (depending on your notification settings) sends an email that you’ll see in your email client. This is basically the same message, but it’s in two places. Let’s say you get the emailed message first. Wouldn’t it be preferable to have the message on the site go away, since you’ve already seen it in your email client? Guillermo discusses a code fix to make this possible.
List segmentation is basically the ancient practice of audience targeting applied to the specific technical parameters of email marketing. Common approaches use demographic data, such as segmenting lists by gender or age, or geographic location and so forth. Michael Linthorst at Econsultancy shares some thoughts on how marketers can boost email marketing ROI by segmenting their lists on more arcane data such as customer shipping habits, or their level of social media engagement.
Jim Davidson of Bronto Software hosted a webinar earlier this month focused on how merchants can use transactional messaging strategies to boost customer engagement during the post-purchase period. Jim advises that post-purchase is the crucial opportunity for cementing loyalty, when interactions with the brand can either convert customers into brand advocates and spur additional purchases — or sour the relationship irreparably. Read the whole thing on the MarketLive blog for specific how-to’s and customer case studies.
Finally this week, Return Path’s Kelly Molloy relates her long history with mistaken identity and mistakenly getting automated messages for people who share her first and last name. It would be funny except for the fact that there’s usually no easy or automated way to correct these kinds of problems when transactional email gets sent to the wrong recipient. By way of Spamhaus, Kelly sets out some best practices that all transactional senders should adhere to avoid these kinds of problems, and make it easy for mistakes to be remedied quickly if and when they do happen.
Learn more about leveraging on transactional email to deliver better customer engagement and ROI with our Transactional Messaging Best Practices eBook!SparkPost © 2018 All Rights Reserved