The number of clicks
Are fewer interactions with a system better?
For a while now, an increasingly common soundbite has been taking root in product discussions: the fewer clicks something takes, the more easy and efficient it is to use. Let’s examine this further!
For simplicity’s sake let’s say by “clicks” we may mean taps, swipes, long-presses, really just about any singular interaction with a product’s interface.
By counting clicks we’re measuring physical effort, not necessarily cognitive effort. The relation between the two is not constant. Most people make somewhere between 5000–8000 clicks per day — some of them easy clicks, others tedious.
When more clicks can help
If it’s a completely new experience, people might need the extra time to build a mental model, discover what your product or service is all about and in this case, friction can be good. Not everyone processes information at the same pace or the same way. It’s not likely that all of our customers will be super tech-savvy with the relevant domain expertise.
Tasks that are big on complexity can sometimes benefit from more granularity — breaking them up into a few segments or sub-tasks would increase the number of clicks or interactions required, but would reduce cognitive load and be more beneficial in the long-run.
Control, accuracy and trust
Automation and interaction are a balancing act. Typing in your city’s name and the form auto-filling your postcode is magic. When we reduce input required, we reduce the cognitive effort required — even possibly reduce the chance of errors in the short-term. However, when the system does something on our behalf, not all of us will understand what has happened or why it has happened. This would leave some people without valuable pieces of system feedback, feeling of control and ultimately trust. Not everyone appreciates magic.
When more clicks could be detrimental
- Accessibility — not everyone clicks with ease — people with reduced motor functions or other physical impairments might struggle with a large number of interactions, especially ones involving a mouse.
- Power users — when people know a system extremely well they may benefit from fewer actions to achieve a task — in these instances automation, shortcuts and hotkeys are great.
- Redundant actions like unnecessary confirmations or standard time-saver features like quick selection or mass action features for repetitive tasks.
What could we do?
So if clicks aren’t a meaningful way to measure ease of use — what do we measure instead?
- Time on task — but don’t look at the average and take it down. Instead, it can be helpful to look at the minimum and maximum outliers and ask — why are some people quicker than others? It helps to segment new vs returning customers since there will obviously be a contrast there.
- Error rate — where do people stumble? Why do they stumble there? How can we prevent those stumbles? In some cases, removing a click might resolve the issue while in others adding a click might be of more help. Again, clicks are rarely the issue, but what lies underneath them.
- Success rate — how many people succeed to do the thing that they came to do? (This differs from conversion rate since it’s what the people using your product wanted to do, not what you wanted them to do).
Usually, each interaction has some purpose. People don’t hate clicking, they hate meaningless interactions. Examine the purpose of the interaction and how it can be achieved better, rather than quantifying and optimising the physical movements required to execute it.
So why do people use this metric? It’s easy.
It’s easy to sit on a video call, count clicks inside a flow and forge hypotheses for optimization — It’s much more difficult to observe how people use the product and involve them in the decision making process.
It’s easy to assume an interaction is annoying or unnecessary as a person that is very familiar with the system — it’s much more difficult to investigate if that statement holds true to everyone using the product.
Judging the experience that our digital product provides based on just clicks is like judging our experience at the bank by counting how many footsteps it takes to go from the entrance to the clerk’s desk.
Don’t be afraid to keep or even add some interaction. It might help people build a better mental model for your product, save some cognitive processing power and give them a sense of predictable outcomes and control.