The New Atlanta Billionaires Behind the Unlikely Tech Unicorn MailChimp
forbes.com
forbes.com
Mailchimp, named after their most popular e-card character, launched in 2001 and remained a side project for several years, earning a few thousand dollars a month. Then in 2007, when it hit 10,000 users, the two decided to commit full-time.
A unicorn isn’t unlikely. It’s imaginary. It doesn’t exist.
But like "cloud" it's been gobbled up by people who didn't know what it meant and has been bent into a parody of itself.
(not to imply at all that a "unicorn firm" is as profound an idea as not caring how your packets are routed but rather connecting to a resource on the network)
I can't make out the difference between all its competitors - Aweber, ConstantContact, iContact. But I can figure out MailChimp from a mile away.
Hardly. Only the first part of that statement is true. It's $10 a month AND UP... and it goes up pretty fast: https://mailchimp.com/pricing/
We switched to phplist + Amazon SES several years back after calculating what our MailChimp bill would be for our list size, and are glad we did.
There's also the consideration of "should all emails live in one platform?" It may very well be that high-volume, low-impact transactional mail should absolutely go through something like SES, whereas higher value emails go through an actual marketing ESP like Mailchimp or something more robust.
It isn't a black and white situation.
Has anyone migrated off of Mandrill in the past few years? Where’d you go, and how was the transition?
One potential pitfall is keeping your blocked-email list synced between ESPs. You definitely don't want to repeatedly send to an address that has bounced or reported your email as spam, as that's a good way to get yourself into an ISP's spam filter for all its recipients.
When using multiple ESPs, the best approach is maintaining and enforcing your own block list. But if you're relying on your ESPs' own block list functionality to prevent additional sends to invalid/complaining addresses, you'll either need to sync their block lists, or somehow partition things so that the same recipient email always goes through the same ESP.
Nearly all transactional ESPs offer webhook notifications on bounces and complaints, but not all have an API where you can add an email to their internal block list.
If this had been a simple mistake, fine (all ESPs have occasional issues), but to my knowledge SendGrid has never acknowledged this as an operational error. In fact, at the time SendGrid representatives repeatedly insisted the change was made intentionally for its (paying) customers' benefit.
(That said, SendGrid had some really nice tracking and segmentation dashboards, which I do miss at my new ESP.)
That said, we have a duplicate of Mandrill’s sending infrastructure at SES and use it enough that we know it is ready for migration. We have a lot of domains (agency.) Keeping track of domain sending security for two providers is occasionally challenging, but we could move everything to solely use SES in about a week if needed.
It is reliable and costs $1 per 10,000 emails sent.
As for where to go -- Sendgrid and Sparkpost are popular offerings in the same space:
http://blog.mandrill.com/important-changes-to-mandrill.html
https://sendgrid.com/blog/how-to-migrate-from-mandrill-to-sendgrid/
https://www.sparkpost.com/migration-guides/mandrill/
Also I mean "absorption" because at some point Mailchimp required Mandrill users to start using Mailchimp credentials to log in. Even though they were the same company ultimately, the transition felt like an extremely sloppy acquisition!
Currently on Postmark, and have found them to be reliable, transparent about the very rare problem on their end, and helpful with problems on my end.
Following Mandrill’s policy changes, I started maintaining a package that simplifies moving Django projects between a number of full-featured transactional ESPs. Even if you’re not on Django, you may find the list of ESPs it supports helpful. [2]
Napkin math time. Assuming 30% profit (very achievable for a tech biz that size), thats $90M a founder per year (if they want). I would take that over the risk of trying to get to a $1B sale any day of the week. There aren't very many industries that can achieve something like this. What a time to be alive.
Note - I understand their are so many assumptions here, but I think my point still remains.
And if Walmart had the assumed 30% margins, they probably could be a $2.5 trillion company. Their revenue is more than 2.5 times their almost-$1 trillion competitor.
Facebook bought instragam for $1 billion even though there was no significant revenue or income. But the potential for growth of instagram itself and the photo sharing social media space was immense.
With established companies like walmart in establish spaces like retail, you know what you are getting. With tech unicorns, it's a bit of a gamble.
Look at marketo for a good comp: 21x EV/TTV - http://tomtunguz.com/marketo-vista-ma/
how about Bloomberg
Aren't their services which use our AWS SES account (with automatic setup) and offer same features as MailChimp but much cheaper?
Are there any cheaper solutions out there that have similar feature sets (including split testing campaigns), and are designed to scale to list sizes greater that 1m?
What's order of magnitude is it?
Hard to beat $0.10 per 1,000 emails...
SES is for all intents and purposes a "dumb pipe" that will send email, and provide APIs for things. AWS Pinpoint is a step up from that, but I personally found some features lacking at this point aside from basic usage compared against more full-featured ESPs.
There's a whole range of products catering to different market segments though. Mailchimp is what I'd consider more in the SMB tier in many ways, and then you go all the way up to things like ExactTarget, Responsys, etc. at the Enterprise level. Even then there's considerations around pricing models that depends on whether you're more of a low-volume, high-touch B2B business, or high-volume B2C for example.
Definitely helpful to demo several providers and compare them against a business/technical needs assessment matrix, of which pricing is probably just one factor (and perhaps not even the most important).
It is so, so, so very difficult to get my users to stop clicking on phishing links when they, correctly, point out that legitimate links in real e-mails look virtually identical. Why does
www.ecommercesite.example/products/brown-boots/buytoday
have to turn into
ecommercesite19.ie.randomdomain.otherstuff.xd/lists/email/4910/598gjweo5g8er7485hwog8u3eo8whfo8wc2o38fh38f/totallynotphishing/9384gjh34fgoiu34hgffh/noreallywepromise/?utm_stuff=2928&utm_things=morewords&utm_whyareyoustillreadingthis=lolmoney
At the absolute minimum, these e-mail campaigns should only contain click-tracking links that originate under the recognizable domain of the sender.
emails.ecommercesite.example/products/brown-boots/buytoday/4918ac7
would be so much easier to understand and use as examples when showing my users.
If so, and they decided to further support integration with senders like MailChimp by auto-creating subdomains as needed, and it worked perfectly, that would make it all easier. But anything less than perfect, and users (ie small business owners) would be hopelessly lost.
Regardless, whoever handles the registration should have an interface to add records. For example, you'll need to add MX records to use Google or Office365 for your mail service. You'd also need to add TXT/CNAME records to prove ownership of the domain for Google Webmaster Tools, etc.
Here's Google's documentation for Wix users: https://support.google.com/a/answer/6149037?hl=en
[Edit] Just read the Wix instructions. Looks like they made it easy and offered presets for Google Suite users. They could do the same for Mailchimp. But they do allow arbitrary records. https://support.wix.com/en/article/adding-dns-records-in-you...
So I'm not sure regular people would even have a clue on how to do that.
If you create the record, they'd make the links use your domain: https://mail.your.biz/destination?unique=123abc
If you don't, they can do what they do today: https://something2-x1-really-sketchy01.mailchimp.com/m0re-sk...
Also, there is precedent for service vendors telling companies how to configure the appropriate DNS records (with a variation for each common domain registrar). Example: https://support.google.com/a/answer/33353?hl=en
Do you have have a suggestion as to how to fix it because everything I am thinking becomes horribly complicated?
If you were designing DNS and URLs from scratch to minimize phishing, you would treat "www.example.com" and "example.com" identically.
(Which is why it is a mostly pointless suggestion.)
- generate an unique UTM (or something like that) identifier (UUID)
- add identifiers to the links
- let mailchimp customers connect their Google Analytics with mailchimp
- pull in data from GA to mailchimp to generate click reports
No, simplifying their links would work. Use an url shortener, e.g. https://business-name.mailchimp.com/campaign-name-abc123.
The only advantage that routing clicks through MailChimp has, is clicktracking. Clicktracking can easily - trivially - be handled serverside. Just embed per-recipient unique ids in the url.
If MC is worried about customers lying about #clicks, they coukd easily make all urls point to A.com/mailchimp?mailid=xxx&linknum=yyy.
That forces A.com to run the one thing MC sends them to sort out the clicks... which of course does the clicktracking and hides the useful info from their customers.
Case in point: google.com
The real reason they don't do this is because it's a more involved set-up from the customer's side. As people mentioned, enterprise-oriented mail services will do this kind of thing but MailChimp is a long tail solution.
The processing power would still be needed on mailchimp's side. All the customer (A.com) does, is add a tiny redirecting script on their site.
It's similar to how one fingerprinter for android works.Get your client to forward you the user details. User only sees the client (so nothing suspicious, no 3rd party to be blocked), while encryption ensures that your client has to forward you the data - they cannot parse it themselves.
Maybe you just need to invest some money in better spam filters, and user education.
I think we need go further: what we need to do is kill the idea that email is an good vector for notifications inside of big organizations (or maybe even outside of them). I should have a dedicated app with a whitelist based system that official notifications go through. This could be not only a security improvement, but also a stab at better UX, better productivity, and avoiding notification fatigue.
Wouldn't a CNAME cover it? It's the way that gmail.example.com gets routed to google's servers.
Better user education is a non-starter.
Mailchimp is pretty far from enterprise and is heavily geared towards small businesses and click tracking links aren't that big of an issue for small businesses.
False. Source: Work with small business owners daily. Email click tracking has been a feature they use for years, and now rely on to help guide content decisions.
For small biz the order of priority usually goes:
1. collect email addresses
2. start a newsletter
3. publish regularly
4. start to look at open and click rates to get a better understanding of your audience
5. use that data to influence content and send rate and acquisition methods
Unfortunately most don't get past 1/2 :(
I feel your pain - it's challenging to have users be vigilant by checking URLs when these things happen.
The share that goes into your yearly salary, bonuses, raises, and benefits (like the 401k match you disregard).
They owe you nothing else, and each employee is free to ask for more or leave for greener pastures if they are unhappy.
Also shares keep on producing income after you have left a company.
If I was making $120k per year working at MailChimp the last thing I would want is 30% or more of my net worth tied up in private shares. What employee wants that much exposure? It’s a private tech company with no plans for liquidity potentially ever.
Nobody is forcing people to work there.
The idea was some things are easier than other (changing jobs is likely easier than divorce)
In my mind it says the founders truly only care about getting their pay day in a very tangible way.
I’m no fan of any X-as-a-Service company, but I trust a bootstrapped and privately owned one a fucking lot more than I trust one built on VC dollars.
(disclaimer: I work for a successful privately owned company)
Maybe they won't get as dedicated personnel but at the same time if they have kept the company owned by themselves to this point I guess it's a valid choice in continuing doing so.
Is it greedy or not, well to me it shows that the founders are shrewd for sure but.. Are they doing this for the benefit of the company or just to maximize their own profits? Hard to tell. I feel giving up equity isn't such a big deal and it would make me more satisfied having my best employees part of the company. Paying only bonuses sounds a bit cheapskate.
Stock options are the cheapskate way to go. Lotto tickets for the newbies.
And vesting isn't always a bad option, don't you agree? In the case of many successful start-ups like Facebook and Google you'd have earned and probably still will earn a lot more from the stock than from any bonuses they could have given you.
Vesting is bad for everyone. The company buying yours gets to keep you around but you are the least motivated employee. You get to waste a few years as they figure out how to replace you. That's the best case.
Usually you leave and are given the option to buy the shares within 30 days at some price without a way to cash out.
Will these shares go up? perhaps but not really. What happens is new investors money will reduce your percentage. Your shares will be put in a non-protected group while the founders/investors take preferred shares with 10x the voting power.
If they gave you cash you buy google shares you sell when you want and you have the choice to put the money anywhere.
If the founders and investors want to screw you over I'm sure they can and will do so, nothing much you can do with that. If they're concerned with their employee happiness I doubt they'd do it just for pure greed. However as you said the lottery-ticket will probably keep the best employees around much longer than they'd otherwise so it's probably not all bad? Not everything can be valued by its monetary value only.
And also I'd argue that the hottest start-ups aren't for sale in stock markets or at least they are already highly priced when they enter them. It's a lottery-ticket for sure but as with all companies you can never be 100% sure how they're going to end up.
The cool thing about solely owned companies is that they can think long term, as in > 1 year. CEO's of an externally funded company don't have that luxury.
It enables a company to do what's best for the company/employees/customers in the long-term vs. only needing to be focused on the short term or some unreasonable expectation of an overpriced exit.
Please consider not painting all such companies with the same brush.
Cory Doctorow asked about this once:
https://twitter.com/doctorow/status/641642822286753792
And they basically said no. I wonder if now that GDPR is a thing if they have to provide this information?