Why We Fail Blog

The word innovation gets thrown around a lot, but rare is a company who sets an audacious goal, such as recovering a rocket by landing it back on earth, and then works through years of expensive failures to reach success. Innovation is not an exaggeration here, and so SpaceX is completely comfortable editing together all their failed attempts for us to see. What other “innovative” company would do this?

Zoom has been my go-to online meeting tool for years because it’s easier to use than competitors. So I was surprised recently when they screwed up so many privacy and security issues. It reminded me of some of the case studies I wrote about in my book, Why We Fail. My theory: our standards for privacy and security increased significantly in the past several years, particularly due to companies such as Apple who made it a strategic issue, and Zoom missed this shift. It looks like they’re moving swiftly to remedy the situation, but hopefully they don’t throw out the baby (usability) with the bathwater (dirty privacy practices). My video explains more.

In my book Why We Fail I included a case study of BMW’s iDrive in-car software (here’s a condensed version published by Fast Company). I’ve been following the progress of iDrive over the years. My working theory was that iDrive undoubtably hurt sales, but BMW’s cars are successful in so many other ways they still sell, and as long as they can gradually improve iDrive over time then the customer suffering will eventually subside.

How software made me loathe my luxury car by Mikkel Høgh is a sobering update that makes me think the software ecosystem for iDrive owners has only become worse. I sometimes travel for work and have the opportunity to rent and try in-car software from many manufacturers, and I’ve come to the same conclusion as Mikkel: CarPlay is wonderful. I plug in my phone and it just works. The car I own, a Mazda, is well-designed with fine if not great in-car software; CarPlay is the single thing I wish it had.

So perhaps iDrive’s gradual improvement has been too gradual, and CarPlay is poised to subsume the in-car software market?

HP goes up against the iPad Pro with its $599 Chromebook x2

The Chromebook x2 seems to have a lot of potential, but there are some big questions — and not just about whether the hardware is as good as it looks. The real open question is whether Chrome OS is cut out to work on a tablet. Google has been overhauling the operating system to work better with touchscreens for a couple years now, but it’s still very much a desktop system. (It’s based on the Chrome desktop browser and its display of desktop websites, after all.) That’s likely to limit how useful it is, especially in comparison to an iPad, which was designed for touch from the ground up.

Wait, what? Is it 2018 or 2008? Ten years ago it was just starting to become clear that Nokia’s Symbian mobile operating system, a system designed to work with buttons that never made sense with touch screens, was going to get killed by the touchscreen-native iOS and Android. Fast forward 10 years and it looks like HP and Acer are doing the same thing by (mis)adapting Chrome. Why? And who is going to pay $600 for this? Especially when great Android tablets exist for around $500?

This is a fantastic book and should be in every designer & PMs’ collection. https://t.co/UZ9ymWNSu4

— Jared Spool (@jmspool) December 15, 2017

Here’s a story about a website I created several years ago that I had considered a success and now, upon reflection, see as a partial failure.

When it comes to measuring the success of information architecture, I usually think of clearly measurable criteria, such as findability. I ask someone to find something in a website and measure if they can and how long it takes. Easy.

But what if IA goes deeper than that?

For several years I did pro-bono work for my kids’ preschool. It started one day when the school’s website disappeared. That’s right, it just disappeared. And it was just a few weeks before their big registration period. Apparently a non-profit that received a grant to make websites for schools had built a reasonably nice site for free. But when the grant ran out, the organization disappeared. And the hosting account and all the website files disappeared too. All the school was left with was their domain name.

My wife heard about this and volunteered me to save the day. I was used to these sort of client-driven fire drills in the consulting world, so I was able to conduct some geurilla research, gather assets, and build a site based on a purchased template in a little over a week. Success, right?

Months afterwards, the school scraped together about $16,000 to hire a local agency to build them another site. While my ego stung a bit, this decision made sense: they wanted more functions and a CMS which my site didn’t have. Fair enough. Unfortunately the agency-built site had several “coming soon” sections that were never filled in, used an oddball CMS that was difficult to use, and overall was hard to maintain.

My version didn’t have these problems. I built a one-page site that was long but one could easily find everything by scrolling down. There wasn’t any superfluous content or navigation. But the biggest problem with it, and why the school replaced it, wasn’t the information architecture’s performance. It was about how the brand was projected by that single page. It didn’t look like a “real” website; it didn’t look “normal.” Normal websites have navigation along the top and maybe the side. The pages aren’t more than 2 screens long (with the exception of articles). And while the school was happy I came to the rescue and gave them something when they needed it, ultimately they didn’t want a site that projected an image that was outside their perception of convention.

So that’s a lesson I learned: IA is also brand, and brand matters.

Victor Lombardi is the design director at CapitalOne, and the author of Why We Fail: Real Stories and Practical Lessons from Experience Design Failures. He helped turn around a failing media business at Fox Mobile Group through the development of a new web platform and mobile apps. He walks the walk by developing his own product, Nickel, with the goal of making personal financial planning accessible to everyone. Follow him on Twitter or buy his book. 

If you’re starting up a new company or product, it wouldn’t hurt to scan these 204 Startup Failure Post-Mortems to learn what not to do from startups who fizzled.

One in particular, struck me in a twisted sort of way: this postmortem of SunEdison’s backruptcy:

SunEdison at its core is a boring construction company, that earns the trust of its institutional investors by being boring and managing risks … [but the company’s senior executives] didn’t want to be boring, they wanted to be a technology company.

Victor Lombardi is the design director at CapitalOne, and the author of Why We Fail: Real Stories and Practical Lessons from Experience Design Failures. He helped turn around a failing media business at Fox Mobile Group through the development of a new web platform and mobile apps. He walks the walk by developing his own product, Nickel, with the goal of making personal financial planning accessible to everyone. Follow him on Twitter or buy his book. 

Tom Peters’ perspective on the enterprise is as true today as it was 10 years ago: too much talk, too little do. If you don’t do, you don’t fail, and if you don’t fail, you don’t learn (not a lot, anyway)…

Etsy has created a guide for facilitators investigating why accidents happened there: Debriefing Facilitation Guide: Leading Groups at Etsy to Learn From Accidents (PDF)

Notably, they jettison the old model that places blame on people who forgot to do something, and instead focus on learning how they can improve and making changes:

Tim Hartford talks about following the “God Complex” of designing with a great engineer vs taking a genetics-inspired approach of generating many random variations, testing them, and repeating the process over dozens of generations until you have something that works very well.

It’s a difference of trying to avoid anything that looks like failure and embracing trail and error as part of the design process.