Last week, we taught you a new way to think about analytics. Today, we’re going to introduce you to the first category of tools: real-time dashboards. They’re the most addictive and misunderstood way to learn about your users.
Real-time dashboards get a bum rap for being too addictive. They’re the hottie at the bar. Your team will gawk, whisper and daydream…
But real-time dashboards aren’t all sex appeal. They’re one of the best ways to keep everyone on your team in the loop and in high spirits.
A good dashboard keeps your entire team aware of how many people are on your site right now, where they came from, and which pages they’re looking at. It’s like looking over all your visitors’ shoulders simultaneously, and it’s guaranteed to give you a better gut understanding of your business.
But there’s a little-known secret that makes dashboards even more important: they naturally make you think of experiments.
Last week we talked about how the trick to analytics is to constantly run small experiments on your business. Real-time dashboards help you come up with those experiments.
How’s it work?
Looking at your dashboard every now and then gives you a sense for what’s normal. When something unexpected happens, you’ll wander past the dashboard and start thinking…
Why are so many people just sitting on the pricing page?
Pretty quickly, you and your team will formulate a hypothesis:
People might be confused by our pricing.
Which leads to an experiment:
Let’s A/B test an easier-to-understand pricing page!
That’s exactly what happened to us last month at Segment, and the experiment we thought up increased our signup conversion by 4x.
That’s the secret to dashboards: they prompt you to ask questions and think of new ways to improve your business. Here’s how that happens:
- Walk by your dashboard on your way to lunch.
- Woah, what’s going on with X?
- Hypothesize about what to improve.
- Test your hypothesis with an experiment.
To sweeten the deal, dashboards are so much fun that your team will want to look at them. Good questions and hypotheses will start bubbling up from everyone on the team.
Just tell me about traffic spikes!
Lots of people argue that dashboards are too distracting, and that they “just want a traffic spike alert system!” It’s a nice idea, but even the smartest spike detector won’t catch the nuance you need:
Most of your traffic starts coming from Facebook. Is that remarkable or typical? This happened to us for the first time yesterday.
Traffic switches from one of your blog articles to another, but overall volume declines. This happens to us every night as the US goes to bed and Russia wakes up to read about PHP.
A tiny trickle of visitors from a totally new source appear. Is it remarkable? It depends on the source! Last week John Resig linked to us in a technical article. We saw it early and were able to reply to him before the post saw major traffic. It was a relatively small, but important source to us.
Instead of trying to predict all the unexpected, interesting scenarios that could occur, just put a dashboard in front of your team. Humans are excellent pattern matchers. (If you don’t believe us, watch Tony Haile’s talk at the Mashable Media Summit.)
To show you what we mean, we’ve collected a few different stories about how other companies use their real-time dashboards to ask questions and run experiments.
Case Study: Al Jazeera
In January 2011, Al Jazeera’s initial coverage of the Egyptian Revolution became an international fixation. Editors watching their Chartbeat dashboard saw a sudden flood of traffic to the English-language page covering Hosni Mubarak’s resignation.
“Without Chartbeat”, Nanabhay said, “we wouldn’t have known the world was coming to us for this story.”
But they did have a real-time dashboard and they knew what was happening, so they decided to run an experiment.
They sent additional reporters to Egypt to increase their coverage of the drama unfolding there. The experiment was a stunning success. Their unintentional hit became a series of high-quality pieces covering each development in the revolution. The coverage was extremely popular and became widely watched and discussed online.
Their dashboard helped them convert a temporary sensation into a sustained 600% increase in viewership.
PS. If you’re interested in the story, check out this cool infographic by Chartbeat.
Case Study: LaunchRock
LaunchRock helps you get users. One way they do this is through their Discover network. If your LaunchRock page starts getting traction, LaunchRock will promote it on Twitter and in their email newsletter.
Every launch page is an experiment - which pages will be popular? Using their real-time dashboard, LaunchRock can see which pages are going viral. Popular launch pages are likely to be significantly more interesting to people in the Discover network, so LaunchRock uses their dashboard to decide which pages to promote.
By using their real-time dashboard to find landing pages to promote, LaunchRock has significantly increased engagement in their Discover network and drives mores signups to launch pages.
They’ve also used their dashboard to validate another hypothesis: teams who carefully cultivate a relationship with their initial signup list get a huge boost when they launch a new product.
The hypothesis came from noticing a pattern on their dashboard: teams would email a new product launch page to signups from their first product. The traffic to the new product would start with returning visitors, but quickly expand to a torrent of new visitors. The initial burst of traffic would drive a strong viral loop:
Case Study: Hacker Buddy
When someone important mentions you on Twitter or posts a glowing (or scathing) review of your company, you want to be ready to respond. The internet moves quickly, and all the hoopla will be over in a couple hours. It’s important to be opportunistic and capture all these new inbound visitors.
Sometimes you need to take an active role in the discussion about your product happening on Twitter or Reddit, but other times you just need to make sure your service doesn’t crash!
That’s what Hackerbuddy managed to prevent. A tweet from British luminary Stephen Fry (3.4m followers) suddenly drove a flood of traffic to HackerBuddy. They noticed the traffic spike on their real-time dashboard, so they hopped on Heroku and scaled up their servers to accommodate.
You can imagine the excitement in the air at Hackerbuddy HQ.
When you’re handed a golden opportunity to capture so many users at once, be like Hackerbuddy and use your real-time dashboard to respond immediately.
Real-time dashboards help you quickly see if something unexpected is happening. If it is, you’ll start asking questions, you’ll generate an experiment, and you’ll be on your way to improving your business. It’s a virtuous cycle.
To help you choose the right dashboard, we’ve put together a shortlist of the best options for real-time dashboards.
GoSquared and Chartbeat both have gorgeous interfaces. They’re designed to be displayed prominently on the wall, and show you everything you need in a simple, real-time dashboard. We highly recommend both of them.
The most important thing is that you get a dashboard. So that’s your homework. Install a dashboard.