Most Email Service Providers (ESPs) offer their customers a choice of using either a shared IP pool or a dedicated IP address for their sending, with many charging a premium for the dedicated IP (a justifiable cost since operating a dedicated IP incurs the ESP additional costs and complexities). A common question from the users of these ESPs is whether they should opt to spend the extra money on a dedicated IP and whether it will help their deliverability.
What is the difference between a shared pool and a dedicated IP?
In order to make things easier for new customers and those with smaller volumes, ESPs will pool together a set of IP addresses used by all customers assigned to the pool. In addition, these ESPs will often operate multiple IP pools, and assign customers to a given pool using logic they have developed based on factors such as the age of the customer’s account, the customer’s overall sending reputation, and the type of mail the customer sends.
For larger senders and those that want better isolation of reputation, many ESPs also offer dedicated IPs, where one or more IP addresses are allocated exclusively for a single customer. Often the customer will have tools available to control which IPs are used for any given traffic they send, but those IPs will only ever be used by the customer.
Choosing between dedicated and shared IPs
The difference between using dedicated IPs or shared IPs is a lot like choosing whether to join a choir or be a soloist; when you’re first starting out, there are a lot of benefits to singing in a choir: your relative inexperience is hidden by the rest of the choir, you don’t have to worry about whether you can project your voice well enough to fill an auditorium, you don’t have to worry about booking events or venues based on your own reputation, and if you need to stop and take a breath you don’t have to worry that the audience will leave.
On the other hand, if you’re a good experienced singer you may want to step out of the choir and start a solo career: you don’t have to deal with inexperienced singers making the performance worse by singing off-key around you, you can start to build your own reputation as an artist instead of hoping the choir you belong to is popular enough, and you can get more of the credit for a good performance instead of having to share it with a bunch of poorer singers.
Let’s explore how this analogy maps to real-world factors in deciding on dedicated vs shared IPs.
Sending reputation in a shared pool vs dedicated IPs
While the filters used by Mailbox Providers (MBPs) can use an immense number of data points to attach reputation to a message including the URLs in the links included in the message, two of the most prominent data points are the calculated reputations of the domain and the IP address used to send the message. Both of these contribute to the decision on whether to send your message to the inbox or the spam folder (along with many other factors).
While your domain reputation is the same with either approach and has an increasing importance over the years, your IP reputation is of course tied to the IP addresses used to send your mail. The impact of using shared IPs depends a lot on what IP reputation that pool has, which is affected by the other tenants on the same shared pool. If the overall pool of senders follows good practices, the IP reputation should in turn be good. A good IP reputation in a pool can help you, especially in the early stage of your sending, especially if it’s higher than what your practices would otherwise warrant. The inverse is also true: if you have good practices but find yourself sharing a pool with senders who have poor practices, you could find that the IP reputation of the pool is hindering you.
With a dedicated IP you write your own outcome, for better or worse. If you have good sending practices, you won’t have to worry that the other tenants are bringing your reputation down, but if you slip up and make mistakes or bad decisions, the impact will be undiluted and you’ll bear the full brunt of the consequences.
It’s all about volume
Similar to when a soloist has to have a voice trained to reach the back of an auditorium vs. a choir where their combined voices fill the space, sending volume is important when deciding between dedicated vs. shared IPs.
The filters employed by the MBPs are based on pattern recognition, and one of those patterns has to do with the flow of mail. While some receivers are more sensitive to it than others, your best results will always come from sending a steady and continuous stream of mail. For a service provider that hosts a large number of tenants, this is relatively straightforward since there’s always traffic flowing through the servers, but for a single sender with a dedicated IP, it’s far more likely that you’ll be sending more intermittently.
This is not to say that unless you send a consistent amount hour by hour that you’ll end up in the spam folder, just that in the mesh of variables that come together consistent send volume is considered a plus by many filtering systems.
You have to get noticed
Remember how if you want a solo career you need to be able to book venues based on your reputation alone? Well, to have a reputation, the venues need to have heard of you, and more than once. We all hear about various things over the course of our day, and we can’t possibly remember them all, so it’s repetition that makes things stick because after we’ve heard something enough times it’s probably worth remembering.
It’s the same with MBPs: they receive so much mail that they have to track absolutely massive amounts of data to do an effective job of filtering mail, and just like our short-term vs long-term memory, one tactic they use is selecting what data to keep track of based on how often they see it.
What this means is that if you’re not sending a sufficient volume of mail, some filtering systems won’t bother remembering you from send to send. While some malicious senders will try to take advantage of this as part of a snowshoe spam effort in an attempt to stay under the radar of filtering systems, the takeaway for legitimate senders is that if you don’t send enough traffic on a regular basis, you’ll always send with a blank reputation at best, and be treated like a snowshoe spammer at worst.
If you’re not sending at least 10,000 messages a day to each of the major MBPs, you’re likely doing yourself a disservice by using a dedicated IP address and would be better off on a quality shared pool where the IP volume is sufficient to allow for an IP reputation to build.
Dedicated IPs are still part of a pool
Often dedicated IPs are pitched to senders as a way to isolate their sending reputation, often in response to concerns about deliverability, and it’s certainly true that for a sender with good practices and appropriate sending volumes, the use of a dedicated IP address can help the situation by separating them out from the other senders and their mixed-bag of sending practices, but there can be a catch: the dedicated IPs are also part of the service provider’s pool of IPs that they have to work with and often are in the same subnet as the shared pool IPs.
For example, say the ESP owns 10.0.25.x as their IP range, giving them 254 IP addresses to work with for their customers. They decide that 10.0.25.2 through 10.0.25.100 are for shared pools, with the remaining IPs assigned to their dedicated IP customers. When an MBP sees a lot of spam and other unwanted traffic coming from the shared pools, they may eventually decide that they don’t want anything from the full IP range, blocking all traffic from the 10.0.25.x network.
While a dedicated IP can potentially help improve your sending results, it can’t overcome the challenges of a service provider that is too loose with their compliance. If you’re sending practices are good, but you’re getting poor results on shared IPs, you may not necessarily want a dedicated IP with that provider as the reputation of their entire IP range may be compromised by their pool.
Dedicated IPs come with other advantages
So far I’ve talked about how dedicated IPs come with risks, and require a certain degree of volume to be a good idea, but if you have good practices, send regularly and with sufficient volume, they are indeed a good idea. Not only do you get to stand alone on the basis of your own good practices, but certain tools and benefits come along with dedicated IPs.
- You can sign up for Microsoft SNDS to receive additional data about your sending reputation at Microsoft domains.
- You can get meaningful IP Reputation information from Google Postmaster Tools.
- You can see your SenderScore.
- You can sign up for certification programs such as Validity Sender Certification and the SuretyMail Good Senders List.
As with many things, the answer to the question of whether you should use a dedicated IP address to send email is “it depends”.
If you follow good sending practices, send consistent volumes of mail each day in the tens of thousands of messages or more, and are on a provider that doesn’t have their entire IP range’s reputation compromised by lax compliance policies, you can certainly benefit from the isolation of your IP reputation and the other benefits that come with using dedicated IPs.