We think about personalization at four levels:

Customization – the explicit statement of preferences, for example:

• I like my data in tables as well as charts
• I have deuteranomaly colour vision deficiency
• I have cerebral palsy.

Crowd – learning from collective behaviors, for example:

• People who ask that question prefer this presentation …
• People who do that expect to see advanced tools.

Segmentation – identifying similarities between individuals, for example:

• People like you prefer simple language summaries without charts
• People like you relish comparative / benchmarking information
• People like you love to run regression analyses.

Unique – implicit, interpreting the individual’s proclivities uniquely:

• Setting up a pension HI based on an existing banking HI
• Turning signposting off after three sessions, ‘as normal’
• Offering up an end-of-month fitness report, ‘as usual’.