With over 30 million users and 220 to 260 million emails sent per month, Bayt.com is the biggest job website in the Middle East. With a frequency of two to three emails per day, it falls into the category of medium to heavy senders.
Nemanja helped us to increase our email output significantly. With his systematic revamp of our entire email program, on both infrastructural and campaign levels, we successfully introduced new automation, segmentation and scheduling procedures. This gave us a significant boost in our inbox placement rate, now exceeding 90%, without us having to purchase expensive certifications. Our open rate increased by 140%, which in turn enabled us to increase our traffic from email by over 150% since last year.
Omar TahboubGeneral Manager at Bayt.com
Bayt.com is a complex system that encompasses not only the core product, offering jobs to job seekers and relevant CVs to employers, but also a dozen other B2C and B2B products. Several departments rely heavily on email and have been using the email system to dispatch their own campaigns, sometimes in parallel. As one of the crucial channels of company communication, email needs to serve all these requirements; while achieving its own KPIs and proving its value to the business. Another dimension of the challenge lies in the fact that Bayt.com has a complex subscription model, where a single user can be subscribed to multiple options, some related to the core of the product, and some not. A good example of this is 3rd party offers, where the company sells advertising and uses external email HTML; provided by various advertising agencies and clients.
The company was struggling with deliverability, more precisely with delivering to Microsoft domains. Decreasing engagement was also an issue, with a documented relative drop comparing to earlier years. Because of its business model, which doesn’t rely directly on extracting a dollar value from email engagements, Bayt.com has a budget limitation for using premium sending platforms like MailChimp, SendGrid and others. The company used its own email campaign application and an open source MTA, strong in terms of pure throughput, but not agile enough for today’s large-scale email program. Another reason why companies like Bayt.com steer clear of premium cloud sending solutions is a very strong need for user data security.
In Bayt.com’s case, the goal was clear: to increase the output of email as a channel and create long-term improvements that will keep the system stable. The output is measured initially in traffic coming to the site from email, but also in qualitative actions that users take when landing on the site: job applications, CV modifications etc.
Naturally, approaching a complex system like this requires running a very diverse set of activities and jumping from one to another in a rather hectic way. However, with the 9 Steps of Email Method, this process always proves to be an iterative one. Switching from documenting to template design, and from there on to deliverability metrics and back to the user’s personal drawing board does not pose a problem. In the first 2 weeks, a comprehensive plan was devised which included the introduction of new company procedures, redefined user journeys, improved back-end email logic, and a full revamp of the campaign creation and sending infrastructure.
The plan included short term, mid-term and long term improvements that essentially turned the email channel around and allowed for such huge increases in email metrics and KPIs.
Following the 9 Steps of Email Method, we explored the Value Proposition of the email channel, set goals and started documenting the Flow, i.e. what the users were supposed to do VS what the reports were actually telling us. After that, we started creating user personas based on the expected user lifecycle stage and moved onto drafting the onboarding, retention and win-back strategies. One of the obvious job seeker engagement drivers is the quality of jobs we recommend them via email. With this in mind, we went a level deeper to understand and tweak the recommendation engine logic and aligned it with user expectations. Some automated campaigns were stopped and new ones introduced.
In the meanwhile, we had to tackle deliverability issues and find the underlying problem. For this, we used ReturnPath‘s inbox placement monitoring module (Inbox Monitor) for all manually sent campaigns, as well as some of the biggest automatic ones. We soon discovered that the deliverability issues were mostly coming from attempts to deliver to Microsoft network, because of (now not so recent) changes in the SRD vote policy that Microsoft had adopted. This required thorough segmentation and a separate approach. Since several different teams were using email to communicate to the same, albeit big, database, we had to implement standards and procedures. We streamlined the dispatch process by introducing a series of checks that every team had to undergo before sending their email out. Starting from the day of sending and the targeted audience, over detailed QA of the HTML using Litmus, to the final approval by the content team, we’ve reduced the frequency of manual campaigns to a campaign per user once every 2 days.
In parallel with these two processes, we went through a change in the sending infrastructure. We moved from an open source MTA to PowerMTA, which in turn enabled us to fine-tune sending and bounce-reading rules and process them adequately. As soon as these changes took place, we started to do a much better job in list hygiene which resulted in a lower number of negative SRD votes.
However, it took around 6 months to solve issues with Microsoft and to start improving the reputation again. Technical issues were complemented by improvements in segmentation and frequency, and both of them went hand in hand with the constant optimizations of the email copy, design, and the HTML code.
By implementing the full round of the 9 Steps of Email Method, we were able to reach above 90% inbox placement rate, without leasing expensive infrastructure or certifications. Optimizations in the campaign and segmentation layers gave us an increase of over 140% in the open rate. Along with the improved deliverability, the traffic from email went up by around 150% from 2016 to 2017.
Learn how to do grow your business using Marketing Automation
The course that you’ve been looking for. For free. Forever.
An end-to-end course that will teach you how to plan, execute and grow your business using marketing automation. Ideal for business owners, product/project and marketing managers who can’t spend days on end trying to find the right information online.