In a world where phones, tv, and even boards are smart… what happened to our websites?
Let’s take a look back.
At the beginning, the web used to be static. Then AJAX made it 2.0 and user content stuffed. That lead to social, and now… here we are, in the “big data” era, where everyone talks about it, but few really are ready to understand what’s going on.
So… Why not using all the data available to make our websites smart?
What is a Smart website?
The smart website uses data in user’s benefit.
The Smart web
Let me introduce this concept with an example. You just watched with your partner a TV traveling show. The chat goes on:
- When was the last time we made it to Paris?
- We should go back this year!
- When is the next holiday available?
- Let me check…
Seconds later this is the Google result for “uk holidays calendar“.
Cool, right. But… wait a second: This is not the calendar you asked for!. Nope. It’s true. This is an answer to your problem.
You Googled for the piece of info that you knew that could lead to your answer. The website tho, was smart enough to understand and solve your need and save you some time.
That’s smart. And that’s, in some way, what Google tries to accomplish with Hummingbird.
Smart with data
The secret of being “smart” is that website smartness comes from data. In the gov.uk case, probably data showing what users did on this page led to this design, highlighting with the “next bank holiday” and not with the calendar.
There are two types of data we can use to make our website smarter.
- User’s data. A smart website is a personalized website. We know a lot about our registered users: Previous shopping, previous behaviors, preferences, CRM activity, open tickets, demographic data… What else do they have to do so we can anticipate their needs? We are talking about registered users who come to our site logged, so we should be smart while complying with privacy laws.
- Cohort data. When we know nothing about the user…. we still know a lot about the user. We may not know about who that specific visitor is, but we know a lot about what happens with visits like it.
Cohort data requires imagination from the analyst to identify segmenting variables that can be correlated to an specific behavior onsite. And a skilled UX designer that uses that data in users benefit.
Let’s use another example. This time, something more… offline.
The smart ticket machine
The ticket machines at Paddington Station are very smart. They don’t know me at all. But when I’m coming back home from my commute, they show this “home” screen:
“Buy tickets for tomorrow morning”
Again, they don’t know me. But after 17:00 they assume that I might be needing tickets for tomorrow and not fro today. And they are assuming right. Why our websites don’t do the same?
The smart analyst / designer
In the train’s example, the data was there (tickets purchased after 17:00 are for the next day). Some analyst discovered it and a designer used it to change something, saving London commuters some taps on the screen, and increasing customers satisfaction getting them home earlier.
Let’s see some examples:
- Smart e-commerce: The second visit after purchase prompts shows you the shipping status of your order. The next visit after it arrives suggests you to write a review the product.
- Smart travel site: In my first visit the form appears filled: the most popular destination from my visit location, the nearest airport, dates filled to the next bank holiday / weekend and set for 2 people (of course, depending on the cohort data).
- Smart support page: Don’t you hate it when visiting a website where you are logged in they make you fill a form? Save time on user’s side.
- Smart blog: if you come from google “not provided” it means that you are logged in Google. Why not suggesting you to follow the blogged in G+? You could also inform the user of what’s the average reading time of the post, and what’s the most popular content on the blog.
- Smart download page: Doesn’t wait for you to click a download button!
The great thing about this approach is that all this simple changes will fit in many users needs and ease their experience… but it will also qualify many other users who browsed your website undecided. By anticipating their needs, you are suggesting a path to the funnel to them.
The Smart website
- Uses user and cohort data to anticipates user needs.
- Uses data to create hypothesis but experimentation confirms them.
- Thinks of personas, not visits.
- Uses the voice of customer to understand the behaviors occurring on the site.
- Understands that the user is smart and makes it easy for it to achieve his goals.
The dumb website, in the other hand…
- Makes it really difficult to us to reach our goals, even when it has all the data that it needs from me.
- Tries to make everything and makes nothing. Stuffed home pages… yes, we are talking about you with that link that less than 1% clicks but “needs” to be “just in case“.
- Thinks that by making things big and prominent and intrusive I’m going to change my goal. Big should mean “usable, intuitive and easy” not “do as I say“.
- Tries to hide the options I need to complete my goals, sending me to their goals pages.
- Assumes that every visitor has a clear goal and visit intent (it is: doesn’t qualify).
Any website that is following one of these rules and not using data to improve our experience, is a dumb website. Make it smarter!
There’s plenty of data. Too many, for sure. The challenge for the analysts is to find the actionable data, and deliver changes with it.
Remember our post on motivations: visiting your website is no ones goal. Solving a problem, most of times some times a hidden need, is.
Help user to solve this problem anticipating his actions, and you’ll deliver an awesome experience in your website that will pay off.
Commenting this post IS Smart
Last updated byat .