Getting forms right: how better words lead to better results

A section from a survey in Arabic. There's an introduction with the words CAROLINE JARRETT and some text followed by an item with a 10-point response scale. It's from a hotel asking for feedback after a visitContent design is crucial for good forms, so I was thrilled when Content Club in London asked me to talk about forms for their September 2024 meeting.

They chose a rich variety of topics about forms. We started on the design of forms, and went to think about testing, complexity of forms, the difference between forms and surveys, and bilingual forms.

I’ll pick out a couple of them to mention here.

Form questions are used singly, survey questions are aggregated

Surveys ask questions, forms ask questions – there are many overlaps, and it’s not at all unusual for designers and developers to consider that something is a ‘survey’ when it’s built using a tool that describes itself as a survey builder, or a ‘form’ because it’s built using a form builder.

I think it’s more useful to consider the reason why we are asking the questions. For me,

  • It’s a form when we intend to use the answer specifically for a purpose related to one individual (for example: delivering a specific service)
  • It’s a survey when we intend to aggregate the answers to get some insight into the respondents as a group (for example: saying what proportion of our target audience has used our service).

I admit that it does get complicated because sometimes we can use one answer in two ways. For example, we might deliver a specific service to an individual but also report on how may people in total have used it  – which leads me, inevitably, to the need to understand how we plan to use an answer before asking a question, also known as creating a question protocol.

Thanks for the help on a matrix of complexity in forms

I particularly valued the group’s feedback on designing complex forms.

I’ve been developing a matrix to illustrate the many ways in which creating and using forms becomes more complicated, comparing

  • how many people are involved (an individual, several people, several organisations)
  • how often they deal with the form (daily, or less frequently).

I had a suspicously empty area in my matrix: forms that are accessed by multiple organisations, but very frequently.

A matrix comparing number of people involved in a form with the frequency of use. Example of an individual/daily form: congestion charge Infrequent/multi-person: Census Multi-organisation/one-off: Grant application
A possible complexity matrix for forms

It was great when an attendee immediately suggested an excellent, challenging set of forms for the empty area in my matrix: the forms that relate to managing complex medical conditions, which may range over different departments within a hospital, different hospitals, and other areas such as social services, privately-paid helpers, and unpaid carers inside or outside a family.

It was lovely to chat about forms

If you’d like to hear about the other topics we covered, you can watch the full recording on YouTube: Getting forms right

Many thanks to the attendees for their contributions, to Nia Campbell and Adrian Ortega for their help as organisers and selecting topics, and to Content Design London for their support.