Earlier this week, Dropbox announced it is shutting down Mailbox, adding to the long list of promising yet failed email products. Why do so few email services make it to long-term viability?

1. Building a good email product is hard

Mailbox had a great mobile app. That alone got Dropbox interested enough to spend $100 million on an acquisition. However, Mailbox never reached the same level of execution with its desktop app. The recent reviews are unforgiving: “Hopelessly unstable. Lovely interface, but ultimately unusable.”

How did a team that clearly knew what it was doing have such a hard time delivering a good email experience on a new platform? Because email is just plain hard!

2. Users’ expectations are extremely high

When it comes to software in general, people forgive a lot of things: bugs in their Twitter client, missing features in their CRM, ugly interface in their invoicing tool (this one is almost expected), and so on. As long as the core experience is fine, it’ll do.

Except for email. The problem with email is that we are all email experts, since we all spend a lot of time sending, answering, and managing emails, and have been doing so for many years now.

Here is a list of what users expect from day one in their email client:

  • High reliability: It has to work all the time. Mailbox did not. Emails wouldn’t send, etc.

  • A huge number of core features: composer, attachments, tags, drafts, archive, trash, conversation threading, contact management, etc.

  • Instant onboarding: Most people already have several email clients. If they don’t get the value of your product and how it works within five minutes, they’ll just go back to their previous solution and forget about you. Forever.

  • High security/confidentiality levels: People don’t care if their emails are used by Google to push ads or if the NSA spies on them. But early adopters tend to care if a young startup is dealing with a really private part of their life.

And here is what users expect from day two:

  • Multi-platform: IMAP, Gmail, Yahoo, Exchange …  Users will want to be able to connect every possible email provider

  • Multi-device: People check their emails everywhere and on all kinds of devices. Computer, tablet, smartphone, web or native, you better have a dedicated client for each one.

3. An email client is no monkey work

Even if you’ve assembled a team of incredibly good back-end and front-end engineers to deliver on the expectations, the next step is to train them on the technical specifics of email.

Emails are still transmitted in a decades-old format, patched again and again over the years to accommodate new features. You’ll need to master Base64, Quoted-Printable, UTF-8, MIME, Multipart, etc. just so you can display your first email correctly.

All-in-all, shipping a minimum viable product is a long, complex, and painful process. And that’s before you’ve even gotten to the first stable version. To get there, you need even more insanely talented people to take it to the next level, both from a design (UX/UI) and technical standpoint. Mailbox had 12 engineers, one of the best designers (Hi Tim!), yet did not manage to get to that first stable version.

4. Long-term sustainability is elusive

Gmail is free. Mail.app is free. Yahoo Mail is free. And they are far from terrible products. People are hardly ready to spare even a few dollars to get access to a new email client. Shifting from consumer centric to business centric products, Dropbox had no hope of monetizing Mailbox one day, let alone make it profitable.

Given the competition, merely staying alive selling an email app is almost impossible. Sparrow tried it but ultimately sold to Google. If you want to make a dent in the B2C market, your only option is as a free product, but then your only prospect is an acquisition. But getting there might be the ultimate curse.

Once you get acquired, you lose two things. The first one is the incentive of equity: You can no longer recruit top talent by offering shares of the company. The second one is that your original mission gets diluted in the acquirer’s primary goal. When your company’s only product is free, your success metric is the number of users. Before the acquisition, your mission to build the best product possible so you can make even more users happy. But after acquisition, your objective is to help the parent company make money; that’s a mission that’s harder to sell to both current and prospective employees. Overall, selling the company almost certainly equates to a diminished motivation of the team and a lower attractivity of the company among top talent. This alone probably explains why so many product acquisitions end up going nowhere.

5. The future of email is not in a new UI

A lot of startups are tackling the email client problem by renewing the user interface. The idea is to improve the user experience with beautiful, polished interfaces (Mailbox was definitely one of them). Very few apps are actually trying to solve the “email problem” through something other than the email interface, but Mailbox’s farewell address — emailed by the Mailbox Team to all their users on December 7 — suggests that there is much to be done in this direction:

“As we’ve increased our focus on collaboration, we realized there’s only so much an email app can do to fundamentally improve email. We’ve come to believe that the best way for us to improve people’s productivity going forward is to streamline the workflows that generate so much email in the first place.”

In other words, to unclutter the email inbox, you don’t need to fix the interface, you need people to send fewer emails. Email is used for everything: personal communication, collaboration, file storage, customer support, calendaring, to-do lists, and more. If, for each one of those “areas”, a service could replace the need to use emails, it would negate the need for new email interfaces. It’s a hot topic right now, and many great companies are pushing in the right direction.

  • For team collaboration: Yammer, Asana, Trello, Basecamp, Notes, etc.

  • For file storage and communication: Box, Google Drive, Dropbox, etc.

  • For team communication: Slack, HipChat etc.

For better or worse, there is a type of email that those services won’t get rid of: external communications. When two people from two different organizations need to communicate online, they default to the most common protocol. It’s been this way for about half a century and it’s where email really shines, so it’s not going away anytime soon, but it could still use fresh ideas.

Mathilde Collin is cofounder and CEO of Front, a multiplayer version of Gmail.