A few months ago we were hours away from launching a fairly large web program, and I received a panicked call from my client: “We need to delay the launch. The website is broken!”
Apparently, his boss had taken a look at the site, and from his description of how it was working and what was on his screen, it did indeed seem pretty broken. Which was odd, because it had undergone some pretty robust and rigorous testing.
“Are you with your boss now? Do you happen to know what browser he is using?” Turns out, he was using IE 7, on Windows NT. Frankly, I was kinda surprised that it even loaded.
My client wanted to delay the launch until we could “guarantee 100% that the site is working flawlessly”. But as Warren Beatty and, the rest of us found out on Sunday night, flawless is an ambitious goal. Even when you have backup systems in place, sometimes those backups need backups.
Monday morning had a chorus of people wondering “Why would they have 2 sets of envelopes!?! That’s a recipe for disaster”. But just imagine the reaction if the single briefcase was hit by a bus or abducted by squid-like aliens trying to stack the deck: “Why would they only have 1 set of envelopes!?! That’s a recipe for disaster”.
Even the most thoroughly tested, well-crafted plan has potential flaws. It takes creativity, experience, third-party objectivity, and perhaps a bit of humility — or at least a lack of hubris — to identify and mitigate them. Yes, a backup briefcase of envelopes is probably a good idea. But maybe you should lock them up with the stage manager once they safely arrived at the theatre. And while we’re at it, perhaps we should lock up the accountants’ phone with it.
But don’t get comfortable thinking that your plan is 100% fail safe. Make sure you have plans for disaster recovery, a communications and action plan, and constant vigilance to recognize and minimize new risks that might arise. And also recognize that some level of risk is inherent in any activity, and any “100% Risk-Free Guarantee” is probably disingenuous.
In the end, I could not assure my client that their site was working 100% flawlessly. In fact, I could guarantee that it wasn’t. Some users are going to be using outdated browsers, (unfortunately, a disproportionate number of them seem to be the ones signing the cheques), browsers and social networks update their code, and the thousands of technologies that run the internet are constantly in flux. Any of those variables can impact performance and experience.
What we could do was quantify that risk, establish baselines of what we could deliver, provide parameters to define what risk is manageable, mitigate that risk as much as possible, and have several backup plans waiting in the wings.
Assuming - or promising - that any project is 100% flawless is the real recipe for disaster.