How to Change ESP's Part 1
Email is central to many companies, so transitioning to a new email service provider (ESP) is a large project that will touch many departments at the company. This is Part I of a guide of how to change ESPs, from beginning to end, from the perspective of a brand-side marketer who was heavily involved in managing one ESP change and lightly involved in another.
This guide will be published in two parts. Part I will cover the processes and tasks around deciding to leave your ESP and choosing a new one, and Part II will cover what to do once you’ve chosen your new ESP.
A note on vocabulary – many companies that facilitate sending mass email are now known as marketing automation companies rather than ESPs. I refer to them as ESPs throughout this article because the primary focus is on the email sending capabilities of these companies.
The Decision to Change ESPs
The project starts with a decision to change ESPs, and there needs to be a solid business reason to change. Your current ESP needs to be limiting your core business in a severe enough manner so that even adding other tools won’t help you.
For example, one reason we decided to change ESPs several years ago was so we could put all of our subscribers in one database to get an idea of subscriber activity level across multiple brands of newsletters. A large reason we decided to change ESPs more recently was the ability to host our lead generation forms and landing pages in the ESP. Our IT department built a home-grown platform in 2008, but it had gotten clunky and outdated, and because no one who built it still worked at the company, fixing features that broke was like finding a needle in a haystack, and then trying to replace that needle with a tool no one knew how to use.
Another reason we decided to change ESPs earlier this year was getting access to better reporting. We want to start sending more emails targeted to specific demographics, so we needed a way to both track the number of emails our subscribers receive from us and systematically limit the number of emails any one person can receive from us in one week.
Other examples of good business reasons to change ESPs are new features that you can exploit to bring in more revenue, integrations that will allow you to consolidate the number of vendors you work with, the ability to track activity of your email subscribers across other platforms, and really anything that could make a significant impact on your business.
An example of a (usually) bad business reason to change ESPs is deliverability. This topic is broad enough and important enough for its own blog post, but for the purpose of this discussion, let’s suffice it to say that most of the time poor deliverability is caused by your own sending and list hygiene purposes, not an ESP.
You should also consider the effect the transition period will have on all aspects of your business. First, your email metrics will drop. You’ll be sending from a new IP address, have a new link tracking domain, and you may be sending from different email domains as well. Your email is going to look different to ISPs and email filters, and any work your subscribers did to whitelist your emails and move them to their main inbox will be undone. As long as you’re a responsible sender, this will be temporary pain, but it can be acute.
Depending on the reputation of your old ESP vs. your new ESP, you may lose subscribers as well. In one ESP transition where my company moved from a B-level ESP to an A-level ESP, we lost 20% of our “subscribers” because the A-level ESP was better equipped to recognize invalid emails.
Second, there is the impact the project will have on your employees. In my experience, this project completed sidelined two people for a full two months, and there were two additional months of heavy involvement for 5+ people.
If you’ve identified a solid business reason to change ESPs and you’re ready for your metrics to drop temporarily, potentially lose subscribers, and have multiple employees only focus on the ESP transition, you’re ready to start your search for a new ESP.
Create Your Requirements Document
Choosing the right ESP to move to is the most important part of the project, and it should not be a rushed process. We spent about four months creating our requirements document and RFP and then seven months meeting with the vendors who answered our RFP and testing out the finalists in a sandbox demo account. Looking back at my calendar, my first meeting about creating the requirements document to change ESPs was in August 2013. We signed the contract with the new ESP in December 2014, and our first live send out of the new ESP was on March 16, 2015. There was definitely some dead time between August 2013 and March 16, 2015 when my department and myself weren’t working on the ESP transition, but that should give you an idea of the time it takes to do this project correctly.
The first step is to create your requirements document. The resulting RFP from our requirements document was 700+ lines long in Excel. To be fair, ours was probably a bit overkill, but because we needed to find a vendor that could replicate all of the capabilities around forms and landing pages that our home-grown system had, we had to be thorough.
When you set out to create your requirements, make sure to document every workflow that relates to your ESP. How do you set up a batch-and-blast email? What sorts of automated messages do you have? How do you set up forms and landing pages? How do you query your database for particular subscriber demographics? What data can you not live without? After you have all that, think through them all again and tease out all of your edge cases. Is there one important client who insists their ads look a particular way or they have to use a different set of values for “industry” on a form?
It’s likely that you’ll have multiple departments and even multiple business units whose duties touch your ESP, so once you have all of their requirements, you need to categorize them into some sort of level of priority. Something that saves one person five minutes every month is not as important to have as something that saves an entire department hours every day.
We started out by assigning each requirement a score of 1, 2, or 3, with 3 being a high-priority item and 1 being something we could probably live without. We then took all of the items we had marked as a 3 and created a list of showstoppers out of those, which was a list of absolute must-haves. It becomes much easier to vet prospective vendors when you have a list of showstoppers that all relevant departments agree on.
Armed with your showstoppers and the rest of your requirements, it’s time to review each vendor’s response to your RFP. This is where the showstoppers and our priority categories came in handy. We reviewed the showstoppers first and disqualified any vendors who couldn’t handle those. We then reviewed each priority category. The meetings we had to review the RFPs were still long, but having the priority categories made them as efficient as possible.
We sent our RFP to eight vendors and from their responses, we selected four to meet with in person. We decided at the beginning of the process to give our current ESP a fair trial to make sure we were using their product to the fullest and they hadn’t rolled out new features we weren’t aware of, so we included them in the RFP process, and they made the cut to the final four. Including them in the RFP process obviously signaled to them that we were shopping our business to other vendors, so before including your current ESP in the RFP process, make sure you’re not in a situation where this will irreparably damage your relationship with your current ESP. You need to keep a positive enough relationship with them to get you – and your business – through the transition.
Vetting Your Short List of Contenders
Each of the final four vendors got a full-day onsite meeting to demo their product. In these introductory meetings, we had a VP from our company, a director from another business unit, and 2-3 other employees from the marketing departments who work in the trenches of the ESP every day. We didn’t want to take up too many people’s days with these meetings, but it’s important to have at least one person from upper management and a representative sample of the marketers who use the ESP daily to get a good sense of whether that particular vendor would work for the business.
The onsite meetings should be less of a presentation on the importance of email and marketing automation and more of a live walk through of your use cases. Asking “Show me how to do XYZ requirement” is an excellent opportunity to identify where the “solution” the ESP has is complicated and not scalable for your business. At a minimum, you should ask the ESP for the demo to include your showstoppers, and if you have time, include your level 3 priorities as well. This is a much more meaningful use of time than the generic presentation and demo the ESP gives for prospective clients.
After the onsite demos, we selected two vendors as finalists. To make the final decision, we had each ESP set up a sandbox/test account where we could actually get in and work through our use cases ourselves. We had one more round of demos that everyone whose duties included working with our ESP attended, and the finalists explained how to recreate our use cases in their product.
We created sandbox accounts for about 10 marketers and gave everyone about a month to recreate their typical workflows in each ESP. We also set up weekly office hours with each ESP where we could ask questions and get refreshers on each platform as we worked through our requirements.
I highly recommend using a sandbox account as a part of the decision process. It’s one thing to watch product experts zip around their ESP and demonstrate how easy it is to set up everything you need, but it’s a completely different story to have all of your marketers stumble around the product themselves. It’s also worth spending quality time in this step to make sure the finalists have all of the functionality you need – it’s much better to find out about disqualifiers in this stage instead of three months into sending from the new ESP when you’re locked into a long-term contract.
Going into the sandboxes, I felt equally positive toward each of the finalists, and after one day in each sandbox, a clear winner emerged. The rest of the sandbox testers felt the same, so we were lucky enough to have a unanimous decision.
Part II of this guide to changing ESPs will cover the steps to take once you’ve chosen your new ESP.