The email service provider (ESP) buyer’s journey—the process of shopping for a new ESP—usually feels a little bit like this…
With hundreds of ESPs to choose from, the buying process can be overwhelming. Throw in a couple T-Rex’s with laser guns on top of it and things get totally out of control.
If you’re lucky, it’s a process you only have to experience once during your career. For others, it can feel more like Groundhog Day, as it’s not uncommon for marketers or product teams to switch ESPs every two years. So, regardless if you’re shopping for your first ESP or looking to switch, I want to help.
FULL DISCLOSURE: Before arriving at 250ok, I previously worked at three ESPs: Aprimo (acquired by Teradata), StrongView, and, most recently, SparkPost.
To help you jump-start your list of questions for potential ESPs, I sent five questions of my own to a small group of fellow email nerds for feedback.
Below are some of their responses (because I only have so much space) and they are placed in no particular order:
What’s the one question most senders fail to ask when shopping for an ESP?
Dan Deneweth (Senior Director, Strategic Services [Deliverability], Oracle Marketing Cloud): “What does the ESP’s deliverability onboarding process look like? Will your team help us with our IP and domain warm up and ramp up?”
Mike Hillyer (Director of Product Marketing SparkPost): “I think the most commonly missed question is ‘will changing ESPs truly affect the outcomes we are looking to change?’ This often happens when a sender is facing issues related to practices and falsely believes (or is led to believe) that simply changing ESPs can fix things.”
Len Shneyder (VP of Industry Relations, SendGrid): “By now most companies understand the basic concept of deliverability—the assumption is that if I pay someone else to send email on my behalf it should be delivered to the inbox. However, what senders don’t ask is how the culture of an ESP they work with will help them become better senders; or to put it another way, what kind of resources exist to help me help myself? What many senders don’t realize is that good deliverability starts with good opt-in practices, matures into engaging content and finishes with segmented frequency and cadence. Customers that understand what to look for know to ask about the kinds of available resources their ESP can deliver. This can help measure their effectiveness, inform their strategy for success and set them on a path to achieving happy customers and revenue that is up and to the right. Some of the things to look for include the kind of reporting available to the customer, the size of the deliverability team, a professional services team and the kind of engagements they can carry out on an ongoing and one-off basis. Does the company provide help on strategy and creative? Is there technology and tooling at the disposal of the aforementioned team to thoroughly analyze every point along the journey from the sending infrastructure to the MTA and the decisions leading up hitting send. Small nuances like this may go unnoticed until a problem arises, so it’s important to understand every angle involved in working with your chosen ESP partner.
Laura Atkins (Founding Partner, Word to the Wise): “How does the ESP handle authentication? Do they have the ability to provide custom SPF and sign with customer domains for DKIM? If they don’t right now, when will they have it? Is it an additional cost or is it included? With the rise of DMARC, alignment is becoming a critical factor in getting into the inbox, and some ESPs are still behind the curve. Others charge extra for it. Senders should be striving to make all their mail align, and some ESPs don’t make that easy.”
Roger Barnette (CEO, MessageGears): “This is a hard one. I would flip the question a bit and wonder if senders are asking themselves why they’re shopping for an ESP in the first place.”
What are the top considerations when shopping for an ESP?
Jordie van Rijn (Founder of www.emailvendorselection.com): “When shown a demo, be aware that it is the job of the salesperson to show the best version of the software as possible. Skipping over things or not digging into your questions is a red flag for you to investigate deeper. Another good one is ‘we can’t do it now, but it is on the roadmap.’ One way to avoid that is by asking them to do a demo with your use case, email, and data-model. Pretty hard to make excuses that way.”
Roger Barnetter: “For large senders with in-house data living behind a firewall, all initial considerations have to revolve around access to data. How can data be accessed? How fast can the data be utilized and updated? How does the ESP handle PII/data privacy concerns? After that, usability and support are important. Senders need tools they can actually use and need to know someone will be there to pick up the phone and help when they need it.”
Len Shneyder: “That really depends on the stage that a company is in. Younger companies may only require transactional email services so an API driven email platform may be just the thing. However, as companies mature they embrace and rely on the power of email to drive brand awareness and other aspects of the business. At this point, a company may bring on another platform that begins a vicious cycle of messaging sprawl. At the high end a company could be working with 6 ESPs and suddenly realize that the use cases are so customized and unique that they have to build their own and they are back to looking for the best email infrastructure they can find to simply power the layer above. To put it neatly: can you grow with your ESP or is the ESP you’ve chosen a one trick pony? Or equal importance is the partner network and tools that neatly plug into your ESP. Maturing companies will connect to commerce engines and other technology that has an output through email. It’s important to look as far ahead as possible when choosing an ESP and ensure that the cloud service is architected in such a way as to be accessible and extensible. Again, it depends on what you as a sender need. Are you looking for a super sophisticated front end to minimize your reliance on IT? Do you require professional and strategic service around your marketing calendar? Do you need a production team to build your templates on an ongoing basis? ESPs come in all flavors, sizes, and shapes—the right one is out there you just have to do a little digging.”
Dan Deneweth: “My top four:
David Baker (Cofounder and Chief Strategy & Operations Officer, Cordial): “Today, it’s speed and flexibility. We believe the old processes are broken and fraught with opportunities to make mistakes, so organizations typically over-process to reduce errors, thus impacting speed and efficiency. We feel there are new ways to do things that may: require process changes in how you produce content; how you manage data; how you manage personalization; and how you operate testing and optimization. We have to remember that many of the core technologies were architected a decade ago aren’t agile or flexible, and require labor and workarounds, and customization that contradicts agility as the business changes. We know optimization is the catalyst, and data is a feed, and real-time is the chain that pulls it all together, so force real-time views of data, performance and optimization as key criteria when choosing a new partner.”
John Caldwell (Principal and Founder, Red Pill Email): “Does the platform do what you need it to do (objective) and is the platform what I expect it to be (subjective). Demos are cool, fun, and often exciting, but while the sales team is showing you all of their cool demo-ware, how much of it supports your business today and your plans for tomorrow? Ideas and innovations are fine and it gives you something to look forward to, but if the platform can’t support the way you make money today then it doesn’t matter what it can do tomorrow.”
Should senders use multiple ESPs?
David Baker: “I’d not recommend it unless you are a global organization where there is no centralized control and in region requirements that differ dramatically. You are minimizing the value of engagement by separating transactional from triggers and promotional/batch. You also lose the aggregate value of ‘seeing’ what’s happening across your customer base, market and people that may be in different markets but exhibit like-behaviors. It’s just as hard to manage at scale without a mission and the mission shouldn’t be to Band-Aid your existing ESP, as you’ll get fractional value from each. I would only use multiple vendors if you plan to switch to help ease the transition, or you are so big you aren’t centralized to reap the benefits.”
Len Shneyder: “My advice would be to avoid messaging sprawl. Messaging sprawl can lead to a dangerous over messaging of clients. Let me give you an example: the marketing team uses one ESP, the partner and loyalty team uses another, a third is used by customer service, and legal/compliance uses some in-house tool. It’s nearly impossible to know when or how often one of these groups sends a communication. The likelihood that someone could hear from all four groups in the course of a single day can be quite high. Additionally, there is the question of efficacy: how do you analyze performance and understand the impact of all that messaging? You wind up burdening yourself with building a BI system, getting all the data into it to analyze, synching things like bounces, suppression lists etc. across platforms. You see where I’m going with this? Choosing the right ESP is important, it’s your bridge, a conduit, a critical line of communication to your customers and it is highly measurable and should feed your understanding of your company’s performance.”
Mike Hillyer: “Using multiple ESPs is like a frequent traveler staying in multiple hotel chains: most don’t do enough volume to build reputation across both. If you’re spreading yourself across multiple ESPs, odds are you will not only lack the volume per IP to build reputation with the internet service providers (ISPs), you’ll also put yourself at risk of being identified as a snowshoe spammer.”
Dan Deneweth: “Using multiples ESPs is no replacement for a proactive, well thought out deliverability sending strategy. However, there are use cases where sending from more than one ESP fits into a company’s business goals. For example, if your company has both B2B and B2C divisions, this could warrant using different ESPs to cater to each market. Keep in mind that there are higher costs associated with using multiple ESPs. Splitting your volume across ESPs will usually increase your CPM costs since CPM rates typically drop with higher sending volumes. There are also labor costs associated with learning and training staff on a second ESP platform. These labor costs can be easily overlooked, but could become significant over time.”
Roger Barnette: “Not as a long-term solution. If you’re evaluating or testing various solutions out (running a proof of concept, for example), that’s one thing. Some senders will also send transactional, triggered, and marketing mailings from different ESPs, but that can get messy. Ideally, one ESP handles it all.”
John Caldwell: “It depends; some don’t like all of their eggs in one basket, and that can be a legitimate concern for some, but thinking that you need a “marketing” vendor and a “transactional” vendor is silly. Gasoline and diesel are different fuels, but email is email is email. Promotional and transactional messages are both electronic messages that follow the same RFC specs. Some think that they might save a few bucks by using a platform marketed as ‘transactional” (typically not full-featured, and just a fancy relay with some reporting), but the reality is that by splitting your message volumes you’re splitting your buying power; and where the added volume of transactional messaging could bring down your overall CPM, you end up paying more for your promotional to save a little on transactional. Another problem with this is leveraging interactions and learning across both types of messages (i.e., if a subscriber clicks a link in a transactional message that could trigger a promotional message). Since promotional messages come out of a different platform, the two platforms somehow need to synchronize—and that’s usually through the user’s backend systems, usually with some level of built-in delay (batch file transfer). And then the pennies you saved in a separate ‘transactional’ platform are eaten—and then some—with additional backend programming and maintenance. When we see a user with multiple platforms they either have multiple autonomous business units working independently of each other; have gained additional platforms through business acquisition and haven’t migrated all to a primary platform where they can leverage CPM buying power; or it’s a political decision (e.g., the product team doesn’t want to use what the marketing team is using).”
Why do senders jump from one ESP to another?
Roger Barnette: “Lack of innovation, especially from the big players, coupled with bad, impersonal service is frustrating senders who are under an increasing amount of pressure to deliver great experiences to their customers. Data is key, and access to that data is very important, and large senders are beginning to understand that the traditional model of syncing data back and forth from an in-house database to an ESP’s database just to send email isn’t very effective.”
Len Shneyder: “There are a number of reasons. One could be that they simply outgrew the capabilities of their current provider. Or they realize that their provider is working with companies that may generate numerous blacklisting and the lack of a white hat culture at the ESP is affecting performance. A common scenario is that something the customer does (e.g., let’s try to re-engage opted out users) destroys the sender’s reputation and ability to deliver email. The normal (and I’ve seen this) response is to blame the ESP, decide that they are incompetent and jump ship to another ESP. Did that solve the problem? No. The problem was how the customer perceived the effect of sending to unengaged and opted out users, not the ESPs software/infrastructure of deliverability practices for that matter. The important thing to keep in mind is that sending email at scale is hard—email, as a channel and ecosystem, has developed a powerful mechanism for establishing trust and a long tail memory for brands, domains and IPs. Simply changing ESPs doesn’t solve the problem if the root cause is a lack of best practices. There are definite reasons why changing ESPs is totally warranted, but not following best practices isn’t one of them.”
David Baker: “We’ve been in this business for two decades, and been on both ends of the marketing side, the key reasons people switch are:
Mike Hillyer: “It’s all too common to see senders jump ESPs because poor practices lead to poor results, and they are drawn into the false promise that simply bringing their poor practices to another vendor will magically make their results improve.”
Dan Deneweth: “Senders often think they will get better deliverability just by switching ESPs. The ESP sales teams often foster this false expectation, saying their platform will give you better deliverability. The truth is most modern, large ESPs provide an equally strong environment for good deliverability.”
Laura Atkins: “There are a couple reasons I’ve heard over and over again but they all kinda boil down to the same reason: The sender is unhappy with the service provided by the vendor. Why they are unhappy can vary. Sometimes, the sender has just outgrown their current ESP and needs more or different features. This is one of the good reasons to switch.
In other cases, the ESP has some systemic problem that’s causing all their customer’s sending reputation to be tarnished. This is an ugly spiral I’ve seen a couple times. The ESP lets in some questionable customers and they end up getting blocked. But the revenue from the questionable customers is vital for the business, so they can’t shut them down. Better customers start leaving because they can. The company becomes even more dependent on the revenue from the questionable senders, and more good customers leave, and so on. Eventually, the company is stuck with bad customers and poor delivery. It’s ugly, but I’ve seen it happen, repeatedly.
The final reason senders jump from ESP to ESP is because they have bad practices and their ESPs tell them ‘no more.’ So, they shop around to try and find someone that will let them send their problematic mail.”
John Caldwell: “Senders jump ESPs because they didn’t take a measured business approach to selecting the right platform. They are swayed by nice sales people and cool demos and pay more attention to those things than they do their own functional specifications. A lot of times this happens because they don’t know what their functional specifications are when they go shopping. They buy based on price and personality and not on how the platform can support their current needs with the ability to scale to their future needs. Rinse and repeat. There are legitimate reasons to change vendors and companies do it all of the time, but if you do your homework there’s no reason that you can’t find a platform that you can comfortably stay with for five or more years.”
Is deliverability dramatically different from one ESP to another?
Len Shneyder: “It can be. Some ESPs like to tout the highest deliverability in the industry. That’s fine, I’m sure that every ESP has brands and customers with engagement so high that they achieve 99% inbox all of the time. However, those same ESPs most likely have customers with deliverability rates around 50% and those are casually not reported or excluded from the data set as outliers. This lack of transparency makes it rather difficult to judge an ESP as a whole. There’s a surprising amount of public data on how white hat or black hat an ESP is, it takes some digging but it can be found. Additionally, and this is anecdotal, ESP size is probably a good measure. If an ESP was incapable of delivering email then their customers would go elsewhere. It’s important to read between the lines, not to get hung up on the marketing, to find a team that you like to work with, talk to customers using the product who can reference its efficacy and the team’s responsiveness and determine if you can grow with the company. This is key—if you’re building a business that’s built to last then you want to work with companies that share your DNA and understand that a partnership is a long-term investment in the success of your product and customers.”
Dan Deneweth: “ESPs provide a strong environment for deliverability, but deliverability success is mainly the result of a sender’s mailing strategies. While good ESPs will help senders with their mailing strategies, deliverability results are less about the ESP platform and more about the sender’s mailing practices. Sender’s who switch ESPs hoping for better deliverability, without changing their mailing practices, will be disappointed.”
Laura Atkins: “Yes and no. There are clusters of ESPs. Most of the ‘big name’ ESPs have similar deliverability. They’re well managed, they police their customers, they enforce standards, mail gets through. Then there’s a different cluster of ESPs, mostly smaller, that are stuck with a few dirty customers and can’t get out. These are often very cheap, but their overall delivery is bad. A case from a few years ago, an ESP had all their customer mail and their corporate mail going to bulk at Gmail. Their reputation was so bad, that everything was filtered. It happens. But, overall, the major players in the space invest enough in compliance that their reputation doesn’t factor into delivery.
Which actually brings up a question that’s never asked and probably should be! ESPs often tell me they want to be as clean or respected as this-ESP or that-ESP. They’ll do whatever needs to be done. Then I tell them how large the compliance teams are at the place they want to be like. Better ESPs invest money in compliance and deliverability. Ask how big their compliance team is and what tools they have to identify and address delivery problems. That shouldn’t be a hard question to answer if it is there’s a sign the ESP isn’t investing in deliverability.”
Roger Barnette: “It really depends on how tightly they monitor and regulate their senders, how they educate their customers, and who they work with for monitoring and troubleshooting.”
Jordie van Rijn: “Yes, deliverability can vary. All ESPs can say they are the best at delivering email and they do. But most of the time a sender with deliverability problems will bring his trouble along with him if he is having trouble. And then he is having trouble at the next ESP as well!”
Mike Hillyer: “Only when it’s paired with the people and tools that help improve a sender’s practices. Poor practices will produce poor results on any ESP.”
John Caldwell: “Deliverability has more to do with the sender than the platform with a few exceptions. Typically, users on a shared IP address are only as good as the worst sender. There are always exceptions to the rule, but we’re looking at the fat part of the Bell Curve here. Along those same lines, if the platform is known to be a bit loose in the quality of their users then that platform may develop some reputation challenges to their entire IP block, which could even impact their users with dedicated IP addresses. But when it all shakes out, a sender’s own practices are the biggest factor in that their ability to deliver to the inbox.”
David Baker: “No. The semantics of deliverability are the same, what is different is the level of support and forensics provided. Everyone does monitoring, but it varies so widely by company given the type of mailers they have on their roster. Sadly, this is the least-funded service element in all ESPs, and typically you don’t get the level of support or proactive management of your reputation until there are issues. This is definitely something marketers and/or brands should dig much deeper into and check references as a good healthy relationship here will ensure sustainability of performance long-term. It’s far too easy for vendors to say, ‘We got it covered.’ I knew an industry-leading ESP that touted great deliverability and they had literally people managing hundreds of clients! As you can imagine, everything was an auto-response and it took forever to get any meaningful resolution. Yet again, the key to deliverability is not the process, monitoring or remediation, it’s the synergy between the client and the deliverability-people proactively addressing things which are known to cause problems and managing that level of risk with business pressures. This is an art form that only the most ‘green’ to our space promise and under-deliver. Experience goes a long way.”
Do you still have some questions that need to be answered? Feel free to contact me at info [at] 250ok.com and I will do my best to help you think through your requirements. In the meantime, here are some additional resources to help you on your journey.