What you need to know about Eye Tracking Harry Brignull
Aug 17, 2014
What you need to know about Eye Tracking
Harry Brignull
Blog: 90 percent of everything Company: Madgex
Before we start, a story…
Anyone know what this is?
Pedoscope – used to x-ray feet
Invented by Dr Jacob Lowe for use in WW1 field hospitals
sHe patented the pedoscope and licensed it to shoe shops
In popular use from 1920-1960!
sHe patented the pedoscope and licensed it to shoe shops
In popular use from 1920-1960!
• Reveals the invisible• Exciting for clients• Visually compelling• Differentiated the owners as ‘experts’• Generates sales
…but completely unnecessary for fitting shoes!
Image credit: Tobii.com
Image credit: Tobii.com
The Tobii T120 Eye trackerCost: ≈ €28,000Anyone can buy oneMinutes to learn to operateYears to become an expert
Reveals the invisibleExciting for clientsVisually compellingDifferentiates owners as ‘experts’Generates sales
Image credit: Tobii.com
The Tobii T120 Eye trackerCost: ≈ €28,000Anyone can buy oneMinutes to learn to operateYears to become an expert
Reveals the invisibleExciting for clientsVisually compellingDifferentiates owners as ‘experts’Generates sales
Just like the pedoscope, Eye Tracking can be misused by novices for trivial things
Image credit: Tobii.com
The Tobii T120 Eye trackerCost: ≈ €28,000Anyone can buy oneMinutes to learn to operateYears to become an expert
4 common misconceptions
Misconception 1Eye tracking allows you to see
what people are thinking
Clooney or Crook – which do people prefer?
Clooney or Crook – which do people prefer?
Eye tracking gives you evidence of what people look at.
This data alone does not tell you whether they like it, understand it or want it!
For this reason, ET is usually paired with other observational data
Caveat: Eye tracking is more useful for some
tasks than others
When you have a simple goal e.g. “Do users notice branding within 5s?”
Eyes forward Eyes left
When you have a simple goal e.g. “Do users notice branding within 5s?”
Eyes forward Eyes left
If only web design were this simple!
Web pages serve many different functions - for many different people- doing different things- in
their own chosen ways.
Misconception 2If there’s no ‘heat’, users didn’t see it
Just because there’s no ‘heat’, doesn’t mean people didn’t see it.
Users can pick up information through peripheral vision!
Misconception 3Eye tracking is scientific, by definition
Let’s a an initial analysisThink of a hypothesis regarding this image:
Let’s a an initial analysisThink of a hypothesis regarding this image:
Congratulations, you’ve done your first piece of qualitative eye tracking
research!
You know what it reminds me of?
Hey, that cloud looks like a rabbit!
Hey, that cloud looks like a rabbit!
In other words: looking for patterns and attributing a rationale.
This is like any qualitative research - but ET is particularly prone because it is visually abstract and easy to misunderstand.
In an quantitative, empirical ET study you demarcate “Areas Of Interest” (AOIs) like this:
Then you use statistics to find out whether people fixated on one face more than the otherand whether the difference in “heat” is down to chance alone!
Misconception 4Heatmaps are generalisable
The user’s goal has a huge impact on eye tracking patterns!
Task: count the columns Task: count the people
Example
This heat map is based on aggregated data from 54 participants during the first 30 seconds.
The report states “all boxes both on the right and the left side of the page are practically ignored”
But what was the task given? Without knowing, this heatmap is meaningless!
http://bit.ly/tobii-realeyes
Conclusions
Perhaps you now understand why some don’t like ET!
“One of these days, I’m going to make a ‘Just Say No to Eye Trackers’ t-shirt.”
“How about a Ouija Board? They run about 1/3000 the price and produce just as good predictions of what works and what doesn’t.”
Jared Spool (2009)
ET advocates have been slow to respond to these criticismsThere’s a fissure growing between advocates and opponents
The killer question:Is ET any more effective at improving design than
conventional methods like think-aloud?
We just don’t know!
Where are the flagship case studies?ROI Examples? Findings that could not have been uncovered through other means?
Quite an inconclusive conclusion – but that’s the current state of the industry, folks!
Huge Thanks toAaron Young & Rebecca Gill of Bunnyfoot
James Breeze of Objective Digital
Harry Brignulltwitter.com/harrybr
90percentofeverything.com