TL;DR: We’ve always tried to put user experience first, even when that gets in the way of making more money. Browsers have changed, we’ve become a security company, and we’ve shifted our business to rely on paying security customers, so we’re turning off the ads in our free DNS service to make that service even better.
Since this is a longer post, I’ve also published this story on Medium, where it’s slightly easier to read.
I started OpenDNS to create a DNS service that was better than what was available. The year was 2005 and when it came to DNS, ISPs were pretty much your only option. And ISP-provided DNS was far from good. It was a cost center for them, relegated to the old servers gathering dust in the corner. They had no incentive to invest in this critical piece of infrastructure, despite it influencing so much of your Internet connection’s performance. ISP-provided DNS was slow, unreliable and did nothing to protect you online, despite being (in hindsight, thanks to OpenDNS) an obvious vector for security enforcement.
DNS was an un-policed conduit for the spread of phishing attacks, malware and other nefarious stuff. All these bad things needed DNS to propagate or execute their attack. The mission for OpenDNS was to change that and give people everywhere around the world a better DNS — one that’s safer, faster and more reliable.
This was a fantastically fun endeavor — reimagining something fundamental that everyone on the Internet uses and unrestrained by concerns about what was or wasn’t possible or had been done in the past. We automatically blocked phishing sites, gave administrators the power to filter unsavory content if appropriate (on school, library or corporate networks, for example) and boosted speed and reliability by making smarter caching decisions behind the scenes. We opened up the black box of DNS and allowed it to be a platform for services, unlocking so much more of its potential. Tens of millions of people switched to OpenDNS, but we needed to make money, too — we’ve always been a revenue-focused company.
When we originally thought about OpenDNS, we really thought of what we were doing as navigation. We were helping people navigate the Internet. We thought this was big. Really big. Bigger than search. In our minds, first you had the browser, and the browser ruled everything — that was Netscape. Then you had Yahoo! They didn’t have a browser, but they had a homepage and a portal. Everyone needed a portal back then, but Yahoo!’s portal was the king. Then Google came out and said you didn’t need a portal, you just needed a search box. You would just go to Google and look for stuff. So we thought that navigation was the next big thing. We could take your address bar, and instead of filling it with complicated URLs, we could instead turn it into a navigation box, and you could type in whatever you wanted and we’d figure it out and take you there, and when we couldn’t, we’d show you search results. Remember, this was years before Chrome and the omnibox existed.
It made all the sense in the world to provide this navigation service for free to users and to support it through ads whenever we showed search results. We struck a deal with Yahoo! to serve the search results and ads, and with that the OpenDNS Guide was born. It worked well, and we became a profitable company quickly.
The way the Guide works is this: If you’re one of the 50 million people using OpenDNS’ free DNS service and you attempt to visit a website that isn’t resolving (the site’s having technical difficulties or you mistyped its URL), we take you to the Guide. We parse what you typed in the address bar and show you relevant search results in a hopefully helpful attempt to get you back on your merry way. Alongside those search results are ads. Around the time we launched OpenDNS it was one of the most trafficked pages on the Internet. Today, for a variety of reasons, you rarely see it, but back then it was extremely common. The Guide appeared when you’d try to load a webpage that didn’t exist, instead of your browser’s “Server not found” error message.
It worked great for several years, but as it happens with technology, the state of the art evolved at a fast clip. Most notably, there were dramatic shifts in the way technology enabled us to navigate the Internet. Enter the omnibox, known to most of you as the Chrome browser today. It built on our idea and turned the address bar into a search field, accommodating behavior that happened anyway and made it so you could type just a phrase into the address bar and be taken to a page of Google search results.
We knew, and know, the omnibox provides the superior user experience: it was built into the browser as code, and had more smarts than using DNS alone allowed, providing a better user experience. For a while, we tried to improve our experience, but it was a losing battle. We also realized it meant fighting Google on the only strategic battlefield that matters to them. Google’s only real battlefield is anything that gets between the user and the advertisement. That’s not a fight we wanted to be in.
In that same period of time, we also invented and brought to market a new enterprise security service. One that only we — with those 50M geographically-distributed users — could. It’s called Umbrella and it uses a combination of machine learning and big data from our DNS traffic to predict and spot threats across the Internet based on analyzing our own data, along with billions of other datapoints from around the Internet. Our security labs are full of researchers who build algorithms to understand how malware spreads, and who build classifiers and anomaly detection systems to identify as-of-yet unknown Internet threats. This predictive nature of our service means we have the capability to identify threats other security vendors don’t. And because it’s based on DNS, it’s delivered as a service, and it’s not perimeter-dependent like most existing security solutions. Again, we’re seeing sign-ups in droves, but this time we’re being paid for it. It felt good to again reimagine something and build it, and see customers reap much value from it.
And I’ll say that the relationship between us and our customers is stronger, because we are building a service for the people who are paying for it. We’re not in some weird three-way handshake that is difficult to balance where we want a great user experience on the one hand, but we get paid by advertisers on the other hand. The cognitive dissonance was getting unwieldy. And frankly, nobody loves ads. At best, ads are a thing you have to do, and you make them as good as you can, when there isn’t a better business model. And for consumer services, there rarely is anything that performs better than ads.
Our enterprise service contributes the vast majority of our revenue now, but not all of it. Those ads from the Guide still account for a meaningful amount of money on our balance sheet, on the order of millions of dollars annually. That put us at a crossroads where we were conflicted by the meaningful amount of revenue from the ads on the Guide and the truth that, today, they lessen our DNS service’s user experience.
We first asked ourselves, “Why do people choose our DNS service?” The answers took me back to that place in 2005 when we set out to do something I saw as noble and important. Make the Internet better by giving people a superior option in DNS. They choose it because it’s safer, faster and more reliable — all 50 million+ of them. And the Guide, originally designed as a way to keep people moving forward, now rarely shows up, and when it does, we know it isn’t the right experience. The guide delivers an experience that’s not inline with what they expect or want. So it has to go.
There’s also the elephant in the room: ads and security don’t mix. It’s clear to us that they are fundamentally incompatible. Text ads and banners alike, they’re all vectors for the spread of malware. We’re a security company first, trusted and relied on by Fortune 100 organizations to protect their people, data and networks. Anything that weakens our security offering by introducing vulnerabilities is a conflict. As we’ve become more and more of a security company, it was clear ads couldn’t stay.
So we made the call: The Guide will go away on June 6th. While the math in the short term makes this appear to be a bad business decision, it’s the long-term impact that makes the decision the correct one. And OpenDNS is playing the long game. We all are here. This experience is one of the only reasons people cite to not use OpenDNS. And today more than 2% of all of the world’s Internet users rely on our DNS service to navigate the Internet safely and reliably. My expectation is that with this reason eliminated, we’ll welcome (and welcome back) waves of happy OpenDNS users. In order to build a sustainable, growing business, a delightful customer experience is a must. And sometimes revenue and experience conflict. When they do, and when we can, we want to put user experience first.
On June 6th, the OpenDNS Guide will cease to exist. If OpenDNS users type a phrase that isn’t a website address into their browser address bar, they’ll get whatever experience the creators of their browser intended. A native, unaltered navigation experience. Ultimately, I feel great about the decision, guided by delivering — first and foremost — a fantastic user experience. And my hope is that this new OpenDNS experience will be met with satisfaction by our users, whose loyalty we are incredibly thankful for. Internally, we know this is the right call, and it allows us to focus all of our energy on building world-class DNS and security services.