"It is almost ridiculous how much money you can save by being a self-sender. At Zeeto, our ROI has doubled since we went down the path of self-sending – even though we have experienced each and every one of the cons outlined"
All right, email marketers. Who of you out there hasn’t said (while usually shaking your fist at the sky) “I’m spending HOW much money and putting up with WHAT from my ESP?! I’m going to do it myself!!!”
The temptation to dump your ESP and take things into your own hands can be strong. So strong, in fact, that we took our ESP-led system and – in a fit of rage, hubris and adventure – took our 40 million email per month program and became our very own self-sender.
For those of you who are considering the move, I thought I’d put together a list of pros (there are some awesome ones) and cons (there are some awesome ones.) There’s really not a ton of literature on the subject…mostly because your ESP friends will abandon you (“You just want to be a SPAMMER!!!”) and the send engines are not, frankly, a fountain of great advice (first CON!)
The first piece of advice is a word of warning…becoming a self-sender is NOT for the faint of heart. If YOU are not technical – or have a solid team of developers who are – you should stop reading right now. Which brings us to our first CON (actually, the second…) of the evening.
CON –Sending email is a highly technical undertaking. You not only have to know HTML, you have to know (or have people who know) CSS/Javascript (for writing your front end), SQL (for getting the data), database structure (it’s not easy pulling millions of records from a relational database,) Dev Ops (for setting up PTR/DNS/DKIM/SPM/etc), server management (cause your database will get big) as well as a host of unexpected other development issues.
If you enjoy technology, setting up your own email infrastructure might be kinda fun. Then again, dev projects that are “kinda fun” can easily go off the rails…which brings us to our next con.
CON – You can’t yell at developers like you can marketers – Building a brand new, highly technical product can be challenging. The temptation to lash out at your developers when they say things like “that wasn’t in the spec” is enormously tempting. The problem is that when you yell at devs, they don’t respond. It’s a bit like bopping a turtle on the nose. The turtle won’t suddenly get on the right course – it will simply pull in its head and refuse to move. Then you’re hosed. In fact, we’ve taken to calling this “turtleing.”
When you have turtleing, you have problems…which leads us to our next con.
CON – Stuff you’re used to will be missing – What’s that? You want to run a trigger campaign? That’s going to need to be developed. Reporting? Maybe some basics…but that’s more of a front end responsibility. At which point the front end guy will tell you it’s a back end problem. You want to do time series analysis on your data? Going to have to go to your BI for that one. The things you often take for granted in your email system simply are not present in many large-scale email system. You’re either going to have to build it or (next con coming up…)
CON – Email front-end systems that scale are EXPENSIVE – Remember how you wanted to get away from your ESP because they charged you a ridiculous CPM to send email? Guess what! Commercial-grade front-end systems charge you (drum roll please) a CPM! In fact, the price you pay for the front end of your system (should you choose to buy it) can rapidly exceed the price you pay for the rest of your entire email infrastructure. Some ESPs will un-bundle their front end and sell it to you…but the prices are not cheap.
What’s that? You still want to be a self-sender? Here’s another con for you…
CON – Deliverability is a gigantic pain – You know how you love to make fun of the deliverability people of your ESP? Guess what – they have a terrible job. The mailbox providers aren’t exactly welcoming ESPs with open arms, never mind self-senders. In fact, unless you’re a giant brand, they probably won’t even talk to you. If there are any people there to actually talk to, that is. You know how many (expletive deleted) Comcast blocks I’ve asked to remove – even though Comcast users have our second highest open rate of all mailbox providers? You will be on an island, forced to fend for yourself. Delivery people are like Col. Jessup – you WANT them on the (fire)wall! You NEED them on that wall!
What’s that? Still standing? Here’s one more –
CON – Less time for marketing – when you manage your own platform, you will spend a lot of time managing your platform. Which means that you spend less time figuring out how to generate more and better results for your organization. ESPs have entire teams of people thinking about how to build a better mousetrap. You won’t.
While all these are big cons…they are not the biggest. To me, here’s the biggest one…
CON – No Plausible Deniability – My great friend Chris Smith (aka - @chrissmithnyc) told me – when I talked to him about building my own system – “don’t do it! You need a throat to choke!” In the polite world of marketing, what he meant was that you need someone you can hold accountable when things go wrong. ESPs are great at this – they usually fall all over themselves to try and protect their business by getting in the way of your speeding train. Choking internal throats does not work (see “turtleing”.) Without the shield of an ESP, YOU become the throat that gets choked. Every delay in development, sub-par campaign, and/or deliverability issue gets tossed in your lap whether you know the answer or not. It’s not a comfortable place to be. Plausible Deniability means you get to toss those grenades right back at your ESP.
At this point you might be saying to yourself…”why the heck would I ever want to be a self-sender? Why would I do this to myself?!?!”
There’s one reason that drives almost all self-senders –
PRO – You will save GOBS of money! – It is almost ridiculous how much money you can save by being a self-sender. At Zeeto, our ROI has doubled since we went down the path of self-sending – even though we have experienced each and every one of the cons outlined above. Yes, I spend more on devs and hardware than I ever thought (or budgeted for.) But even given those expenses, we’ve still seen a stunning improvement in our bottom line. For large senders, this can mean hundreds of thousands or even millions of dollars of annual savings. Taken over a three year planning horizon, the savings can be astronomical.
While that’s the reason many self-senders choose this path, for me it’s a slightly different reason…
PRO – Freedom – When you’re a self-sender, you pretty much have carte-blanche to do whatever you want. Want to send more? Fine, go ahead. It won’t cost you more. Want to set up dedicated IPs per mailbox provider? Be my guest. Want to see what happens when you add a dozen IPs? Sure. Want to tightly integrate your internal data with email data? It’s way easier when everything is in the same database. Want to stop being lectured about best practices? Nobody will lecture you (mainly because nobody will talk to you.) It’s an incredible playpen of discovery. YOU get to choose what you develop. YOU get to choose how you deploy. YOU get to discover all the holes in the ecosystem. YOU get to learn for yourself to drop the dogma and focus on results.
Becoming a self-sender is extremely challenging. But if you’re up for a couple of years of pulling your hair out, the rewards can be spectacular. Otherwise, call your ESP and say "thank you" and "I'm sorry for being such a jerk."