Turning mistakes and ‘meh’ into a growth plan for 2018

When I first started working for SSD Nodes, Matt, the founder and CEO, made an offhand comment that’s resonated with me. I’m paraphrasing, but here’s the gist:

We’re a small company, so we can take risks and make significant changes, and we won’t get called out for it.

Unlike, for example, Dropbox, which was universally panned with the dreaded “hipster” label when they released their bold new design language. No tech blog is going to write an article based on the changes we make. If users even notice, they’ll likely move on and get back to building things without making a fuss.

Matt’s point was that we could make meaningful changes to how we present our company and our products while staying in stealth mode. Aside from a few celebratory tweets, that’s how we approached overhauling our brand throughout 2017: quietly and without fear.

Here’s the thing, though: We’re not such a small company anymore. 2017 was a very good year for us.

Does that mean we’re going to abandon the strategies that got us this far? Nonsense. But we are going to approach them with a little bit more order.

Be unafraid to push what we don’t love

Aiming for perfection is a surefire way to get absolutely nothing done.

Since September 2017, we’ve made hundreds of commits to change every single one of our pages, both with small tweaks and massive redesigns. I’ve hated most of it.

The process began with a complete redesign (and restructuring of the underlying code) in September that was meant to modernize our brand and create a bright, welcoming environment. I picked some bright colors and organic shapes, filed a pull request, and Matt gave it a thumbs-up in the form of merging my new code into master.

The old SSD Nodes hero. No bueno.

Like I said, I hated it.

But this redesign was nevertheless an essential step forward—it was quick and dirty but precisely what we needed at the time. It established the foundation for all of our growth strategies in the last quarter of 2017 and brought years of rapidly-developed code into a more orderly form.

Maybe most importantly, the change reestablished communication between us and some users who had gone quiet over the years, recharging their confidence in us and ensuring their subscription doesn’t lapse without them even knowing. Never underestimate the power of a little disruption to get a once-silent customer talking.

This idea will remain the core of how we continue to build and market this company: perfection is for the people who don’t expect to go anywhere.


Move faster than data

One problem is that our fast-paced, iterative model doesn’t mix well with traditional web analytics and techniques like A/B testing. How can we control for a single variable when any given page is always changing? We’ve grown a lot, but we more than a day to collect the right amount of data.

I keep thinking, why wait for data when I know that a new iteration is better than the old one? As designers, creators, or developers, we damn powerful machines for intuitive analytics.

In 2018, we’re going to remember that our collective guts have gotten us a long way and that we shouldn’t start ignoring them now. We need to be brave enough, and trust in our knowledge about this marketing and our user’s desires, to make changes without having a graph give us the green light.

That’s not to say we’re going to eliminate analytics from our toolbox—much to the contrary.

Isolate analytics

In that final 2017 marketing meeting, we agreed on a new approach to analytics and iteration: we would leverage paid advertising, which offers isolated insight into the success of copy and designed assets, as our “playground.”

If we want to experiment with an entirely new visual style or a different angle on how we approach copywriting about our services, we can commandeer an existing paid advertising campaign, try out the new assets, and see how things work.

We’ll probably use this approach primarily for better understanding how to speak to our wildly different types of users. We have some users who have never used a Linux terminal before and plenty others who have decades of sysadmin experience—why would we talk to them the same way?

With any luck, isolating analytics will help us understand how each type of user reacts to the unique ways we talk about our product and its features. If nothing else, it’ll be a fun testbed for us to try new things with even less risk than usual.

Annoy people

I don’t love what Dropbox did with that that redesign I mentioned earlier, but I do love their ambition.

The bold and disliked Dropbox redesign.

In fact, the thing that annoys me most is that this page hijacks my scrollbar.

The point remains: Even as a massively popular company, Dropbox acted without fear of annoying its current users, alienating potential new customers, and well, pissing just about everyone off. At some point, for reasons we might never get insight into, Dropbox employees decided that a bold redesign was necessary to the future they wanted to build for themselves.

I would like, as much as possible, to be bold enough in our changes that we happen to annoy people here and there—not enough that they leave us but just enough that they start paying attention to us again.

Because we sell virtual private servers, a blinking cursor in a faraway server (with some fantastic hardware behind it, of course), it’s crucial for us to remind our users that we’re still here.

Indeed, we are here, and we’re growing fast. In 2017, we doubled our number of users, and we’re now competing with some of the biggest names in cloud hosting. In our final weekly marketing meeting of 2017, Matt said that he wants us to take even more risks in 2018.

I was taken aback for a moment—annoyed, even, as though our 2017 efforts weren’t enough—until I realized that, instead, they had been proven themselves to the point where they’ve become part of our company culture.

About a week later, I pushed yet another major overhaul of the site to our repository, changes you can see live on our site now:

The new SSD Nodes hero.

And then, the next day, I opened a new branch and started to think about how I could do it all over again, but better.