“Salesforce looks like garbage, let’s find something else or build our own CRM.” — A conversation happening right now, inside of a startup

2015 was a strange year for me. I gained an appreciation for terrible-looking software. Functional but ugly.

When I joined imgix in May of 2015, we needed to make some quick decisions1 around a few key pieces of software for our growing team. One of those questions was: What CRM would we use?

At LayerVault, we spent a lot of time agonizing over the tools we chose. We wanted each tool to be as beautiful as the one we were building. It often paralyzed us into not purchasing software that we desparately needed.

Standards have changed little in the last decade. Having a sensible interface with a smooth learning curve is still not a requirement for developing successful software.2 Harsh truth.

Salesforce is merely a service that illuminates the others: its looks is dated, it has years of design debt, but it’s worth every damn dollar.3 Using it feels like checking your email on your parents’ Windows XP machine over a holiday break.4 After stumbling around in the dark for some time, it eventually clicks. You generate your first report after giving the system some information and woah. You see the value in the result, and you forget about the system’s dated look. Over time, you get used to it.

It feels a little strange now using software on a daily basis that would draw the ire of all of Designer News.

The Big Redesign

It’s for these reasons the Big Redesign is an outward indicator of internal flailing. A company eagerly overhauling and re-overhauling its design language is headed for the dumpster.

The Big Redesign is an implicit agreement that understanding customer needs within the current system is hopeless, so just drop a bomb on it and start from scratch. I’ve been there. The Big Redesign has the right intension, but never produces the right result.

In a weird way, it would be nearly impossible for Salesforce to update elements of its design without alienating a large chunk of its users. When making sweeping design changes, companies should back-of-the-napkin plan on losing or at least frustrating 20% of their users. Damnit, where’d they move that button?

When users log into your product-market fitted app, they are there to complete a task not to learn a new design language.

When is this line of reasoning wrong?

There’s one exception that may prove this curmudgeoned line of thinking wrong: Slack. Slack took the time early on to add a level of visual polish and sophistication usually not seen in a product at such an early stage. Is that a driving reason for its success to date, or is that just camouflage for some other problem they elegantly solved? Would Slack still have shattered SaaS growth records were it to look like Salesforce?

The Great Slackening still should be taken as the exception.5 The rest of the startup world is still processing it. Mimicking their decisions outright probably won’t work for your young company. Put another way: “X but designed better” is not an effective recipe for success.

Not an excuse for stagnation

This post is not an excuse for eschewing all design improvements in your app. Rather, it should convince of you of two things:

  1. A visually-pleasing design is still not a requirement for a successful business as of 2016, and
  2. Design improvements should be made incrementally.

Make a functional but ugly company, and focus on revenue growth. Incrementally improve your design language along the way. The first version will be a cute screenshot to reflect upon while you ring the opening bell at Nasdaq, or whatever your end-game success story may be.

In short, lose your edge.

Special thanks to X, X, X, and X for reading early drafts of this post.

  1. Most of this post will rub folks the wrong way who are not incrementalists. Over the years, I’ve learned I am very much an incrementalist as opposed to a completionist. I recommend the Rands in Repose “Incrementalists & Completionists” post for a primer on the matter. ↩

  2. I’ve driven the purchasing decisions for several services within imgix where we couldn’t sign up or even get a trial account. I had never made a SaaS purchasing decisions like this before, but still have yet to be totally bitten by this practice. Instead of getting a demo account, you’re usually treated to a Google Hangout or WebEx with a sales rep driving a test account. This prevents prospects from self-qualifying themselves out of the sales process due to complicated setup procedures. ↩

  3. I shouldn’t crap on Salesforce too much, they are extending an honest effort with the Lightning Design System. Still, this is not generally available on most Salesforce installs without some extra configuration, which tells me this updated look is likely stuck in internal political limbo. ↩

  4. My parents don’t run Windows XP, but you know what I’m talking about. ↩

  5. Slack is such an outlier in SaaS that it’s about as effective as using Apple as an example. Everything about it is over-generalized and rose-colored. ↩