During the delightful vendor panel at the Emetrics Summit in London a question was asked that I’ll paraphrase as: “What is unique about your tool’s UI for end users?”.
There were the usual set of answers: “We have great UI.”, “Would you rather have a good UI or accurate data?”, “We focus a lot on improving UI.”, “What?"… You get the idea.
Having spent a few years on the Practitioner side and having worked with tons of users of web analytics products I have to say that in the Web Analytics world there are few things more important than a good User Interface (UI).
Here is the main reason for that:
There are few decision making environments that are quite as data rich and information poor, containing complex definitions and metrics which are defined differently depending on which vendor you have! If things could have been made perfectly confusing then we have been there done that.
On the other hand we have Users of web analytics data and reports who are not sophisticated, they want simple easy to understand that they can translate into action. You can’t substitute the need to understand and do analysis to gain ground-breaking insights, but there is no reason for even the Newbie User to not understand what happened on the site last week at a high level.
Does UI matter? You bet you last dollar it does!
(And if you don't believe me just ask my friend John Marshall, CEO of Clicktracks!! :))
Leading a User Research team for the last few years gave me a great appreciation for how simple Design, Information Architecture, User Interface, enhancements can radically change the ability of a User to complete their task.
Specifically for web analytics good User Interfaces can help by:
- Instilling confidence in the User that they can trust the data they are looking at.
(by expert use of buttons and colors and layout and pretty exports of data etc).
- Improving discoverability of data
(today most tools do their best to provide tons of data and then “hiding” it effectively so the User has to dig and dig and dig to find anything helpful, or worse they don’t know what to look for in which case the whole exercise is a missed opportunity).
- Providing context
(why should we have “single metric” reports or individual isolated slices? why not surface contextual metrics automatically to help provide critical context to Users? example: here is the trend for visitors but look what happened to time on site and bounce rates and conversion rate and …. , and here is the key, why not everywhere?)
- “Short-circuit thinking”
(new phrase I coined the other day, tools should be so do their best to not require Users to think up to a certain level, they should be so Customer Centric that they should “talk” and the Users can simply “listen” to the data without having to think, most tools force you to think even for primitive metrics / data – “filter this and apply that and then mix a whiskey and press this button to get your Visits by Search Keywords”, if you make people think for simple things they won’t use the tool / report / metric / data).
The reason most tools are where they are in terms of poor UI are there because we, Web Analytics, started in a world of rich data and then we had to figure out how to present it to the Users. Often we have not stopped to say, “let us think of what people actually want and see if we can present the data in a meaningful way”.
Let me illustrate some of the specific principles I had mentioned above with a tool I have come to love a lot, specifically isn terms of its UI. Measure Map is a alpha stage product that is now owned by Google (please see Diclaimers & Disclosures).
Go ahead and click on the image above. You have to admit it is damn pretty (ignore my ugly arrow marks). When was the last time you logged into a web analytics application and it looked so yummy?
Big bold use of text and colors. It clearly illustrates what time frame you are looking at (#1).
There is a easy to understand timeline that is “smart”: it helps you choose the time-frame for your report and at the same time shows you a trend of Visitors (#2).
Next up the principles of discoverability and providing context (#3). Your core KPI’s are listed clearly and provide the context you need about what is going on in the website. I discover that every single one of my posts in the last nine months has been read during the last two weeks (!).
Finally the principle of short-circuit thinking (#4). Rather than giving me a spam of numbers that I have to think through I get a quick summary that informs me of the performance and helps me understand what is going on.
Perhaps you’ll say that it does not give all the answers, it does not. But it does get me going on asking the right questions.
But it does not stop there, I am curious and dive deeper…..
Notice there is no breakdown of “New” vs. “Returning” visitors that you might normally find. Rather a simple text: “22% of those visitors had been to your blog before”. Simple, anyone can understand. And as if it was not enough you get the raw numbers (1,782) and context (90 more than…). You’ll see the principle of providing context in action again at the bottom with other metrics to help you understand your visitor metric, by providing you a intelligent summary.
It is all in Ajax so even simple tasks are made intuitive. Want to change the time frame? Simply drag the slider on the timeline and let go…..
Your graph / report is automatically refreshed, you never leave the page and you never click ok. The principles of instilling confidence and short-circuit thinking.
We are all used to dashboards and graphs that have bunch of lines / tables, some of them are even pretty, but take a look at this one…..
Notice the application of principles 1, 2, 3, and 4. It shows you the trend and you can easily manipulate the timeline for summary stats. But it is also a “smart graph”, the little boxes you see at the bottom are days on which I write the posts. Suddenly the trend of comments is a lot more interesting as I can see the correlation.
There is one more powerful thing about this graph:
Not only can I hover the mouse to see the name of the post (nice) but it gives me a summary of how many comments I got on that post. Now here is the interesting part, the comments came over the next few days and the tool is smart enough to total them and show it to me.
Imagine a daily trend of Revenue on your website as the trend, the markers are each of your campaigns and if you hover your mouse it tells you aggregate of all revenue generated by each campaign (not just on that day). Would you agree that would kick butt?
Confidence. Discoverability. Context. Short-circuit Thinking.
There is no reason that every single web analytics tool out there does not ooze these principles and make life easier for every type of users. Get Newbies comfortable and understand data instantly with no training or empower the casual users to ask the right questions and do relevant deep dives or get the Numbers Gods the kinds of power they need to do what they want.
Having a great UI does not mean providing more, it means providing less and only what is relevant. Having a great UI means you no training for end users to get 100% of their initial answers from the tool. Having a great UI does not just mean sexy, it means useful. Vendors that figure this out will have unlimited number of profitable customers.
I disagree with the premise that Web Analytics is hard. It is not. We make it harder than it needs to be. It can be a positive experience and something that can be downright fun.
What do you all think? Care to share yummy features of the tool you use? Please share your feedback and critique via comments.
[Like this post? For more posts like this please click here.]