MailMum

As 2020 started we started to do marketing like setup a newsletter and writing relevant content for the target group. The domain was new, so we needed some content to gain some reputation on all search engines.

By creating content we started to identify more pains of other email admins and the things they were searching for. We started to research the market again based on the pains we have found and how our solution could help them.

In this time we reevaluated the product over and over again. It simply worked and got us and our email customers a lot of value.

The software development struggled a little bit, as we did more research to stop coding less relevant things and create a bigger picture on values we can provide.

Finished the first MVP which can handle basic analytics and blacklisting.

Updated our servers to use MailMum blacklist instead of internal cidr_map based blacklist directly on servers. By this change we got a lot of data which gained a lot of insights regarding internal mail traffic of our servers.

As we expected a lot (40-60%) of blacklisted hosts and networks from our manually built blacklist are not relevant anymore or never hit our servers again. This matched the insights we built based on our email server logs.

As the analytics also showed us, we could predict spammy hosts by their traffic peaks and especially networks by their well distributed delivery behaviour.

Identifying on spam and attacks specialized providers became a game changer too. We could see them, blacklist and monitor their action through different data centers around the world. A first "can be reputation" idea came up.

A name, domain, logo, and a draft website - this is the way MailMum was born out of an internal project.

To reduce the time to manage the blacklists through DevOps tools directly on servers and restart the email services after every change, we built our blacklists, which could diversify servers and this way and was fast enough to answer to handle all the requests without getting high system load and slowdown mail processing. At this point, every admin had his instance he could manage by using his favorite scripting language.

Building our blacklists based on gathered information through hitting junk our inboxes, spam traps, and analyzing logs, we started to develop our blacklists based directly on servers. With about 50 entries, we got better then the public blacklists were using and could stop most junk. These blacklists became a big success because we blocked most of the same unwanted emails for our users, which they have seen hundreds or thousands of times. For receivers, they were all the same. If you are an email admin, you know what we are talking about.

But as our built list grew, we need to find a solution to check, update, and extend the entries with less manual work. With time the records got stale, or IP ranges were used for valid traffic; we need to figure out how to manage better and see what we can do to improve this.

This success and need for more information was the point we started to build a service, where we could see, update and improve our blacklists, which were at this point, our best line of defense against mass junk and unwanted emails - especially the recurring ones.

MailMum was born out of a need to face problems with unwanted emails called spam or junk. As filters like Spamassassin were good, but spammers are human, their job is to figure out how to bypass them. They will try and will find ways to reach your inbox.

But even before we were using filters, blacklists were the most common defense against unwanted emails. Like most starting email admins, we used public blacklists first and were very happy, that a lot of junk arriving these days as waves from single senders could be blocked, reducing the amount of junk and traffic to our servers.

But as soon as blacklists started to list IPs essential for our customers and us, we faced the problems as we think every admin met, who used or still using public blacklists.

But as the internet started to grow and more and more servers and services with virtual machines and later called cloud computing, grew even faster, more and more blacklists began to miss a lot of spam hitting our systems. That’s why we’ve started and tried to understand and build our first blacklisting and whitelisting solution, to deliver better quality for our customers.

By whitelisting essential services, we could bypass blacklisting for our customers' important IPs and email services on public blacklists.