Description
Asks many users to pick their few most important tasks from a long list, revealing the small set that matters most.
Most products have a "long neck": a handful of tasks account for most of what users care about, followed by a long tail of trivial ones. Top task analysis compiles an exhaustive list of 50 to 100 candidate tasks from analytics, search logs, support and stakeholders, then asks a large sample to vote for their top five. Because people must choose quickly from a long randomised list, votes reflect genuine priority. The top 5 to 10 tasks typically receive about a quarter to half of all votes. Sauro uses it to choose which tasks to benchmark; McGovern uses it to focus whole organisations.
Attributes
- Discipline
- Information Architecture & Content
- Loss type
- Active
- Stage
- Strategy, Definition
- Data
- Quantitative
- Effort
- Weeks
Steps
- Gather every candidate task from analytics, search, support tickets, competitors and stakeholders.
- Consolidate into 50 to 100 clearly worded, non-overlapping tasks in users' language.
- Survey a large representative sample (several hundred) asking each person to choose up to five tasks from the randomised list.
- Rank tasks by vote share and plot the long-neck curve.
- Use the top tasks to focus design, navigation and benchmarking, and the tail to decide what to de-emphasise.
Sources
- Benchmarking the User Experience (Sauro)
- Factor and Discriminant Analysis (Sauro & Lewis, MeasuringU training)
Also known as
- Top tasks
- Customer carewords top tasks
Categories
- Content
- Information Architecture & Content