Theme Management Committee Covid-May 6 2020 - Comité de gestion des thèmes Covid-6 mai 2020

Back to / Retour à > TMC meetings (COVID) - Réunions CGT (COVID)

Agenda

1. Updates

2. Page success widget - Penny/Isabelle/Patrick

3. Planning for recovery - Laura

Notes and action items

1. Updates

TBS

• Attended a meeting with Gerry McGovern who is doing an international top task survey and card sorting exercise for the WHO. Findings are being shared with participating governments.
• Finishing initial testing on Indigenous audience tasks. Will present findings Monday, May 11.
• Have done some initial testing on Student Benefit content with CRA. Will do more testing after the launch next week.

Health Canada

• Participating in Structured Data pilot, using FAQ schema and How to schema
• Moving forward with landing page updates shared last week

PCO

• Asked group: Which reports or data are people finding most useful in their day-to-day work for informing decisions? And what would they like to know but don’t have visibility on? Posting in the Analytics Slack Channel about this and would like your feedback.
• Accessibility - there are conversations happening at the senior level. Celeste would like to know what challenges around accessibility you have when publishing and if you have ideas for improving accessibility.

No updates from Principal Publisher

2. Page success widget - Penny Day/Isabelle Waltzing/Patrick Lajeunesse

Goals

1. At the page level, understand:

• what people are coming for
• are people finding what they need
• if not, why not

1. Use that data to make improvements or suggest areas for further study and testing
   • Where and what are the problems?
   • Frequency of problems (how many people reporting it)
   • Do we need to do some testing?

3. Assess the impact of changes

“Report a problem exists, but provides no way to report success or what the problem was.”

The page performance widget instead asks people if they found what they came for.

• Widget text reads “Did you find what you are looking for?” with buttons for yes or no.
• If the user selects no, another screen appears and asks “What was wrong?” Options and a text field allow users to provide feedback.

Data goes where it can be used

• Quantitative data goes to Adobe Analytics
  • Build segments based on task success/failure
  • Track what people did after not finding what they’re looking for
  • Track changes to pages and measure the impact
• Text-input goes to the responsible department (Health Canada for the pilot)
  • Email for now
  • Later into a separate datastore

To start - deploy on limited COVID pages

• Assess the effectiveness and administrative burden
• Make changes to text as needed
• Consider wider deployment

Anticipated volume of emails

Email volume = 1,500 emails per day (estimated)

Processing Emails

Short-term

• Use rules to filter and file emails into designated folders
• Remove any personal information from emails
• Export data from email to Google sheet

o  Email is formatted to enable simple copy/paste into spreadsheet

o  date/time;page title;page url;yes/no;what was wrong;details

• ·       Analyze data by web page, content owner, issues reported, frequency

Long-term

• Data will go directly from widget to datastore

Privacy

Since we are not asking for personal information in the feedback, there are no privacy implications.

However, in the event that we receive unsolicited personal information, we sought advice from:

• Privacy Management Division, Health Canada
• National Security Management Division, Health Canada
• IT Security, Health Canada

Assessing the pilot

• What % of visitors use the widget?

• Are we getting useful data from the text entry?
  • What % include a comment?
  • What % are useful?
  • Require intervention?
• Are the choices for why they couldn’t find what they wanted sufficient?
• Do they include anything unnecessary?
• Do changes in pages lead to changes in reported success/failure?
• What’s the management burden?

Next steps

• Proceed with pilot and assess the results
• Make changes as appropriate
• Complete development of the datastore for user-entered text
• Plan for wider deployment across Canada.ca
  • Should it coexist with Report a problem?
  • Who will be responsible for managing user-entered text?
  • Can we pre-process text to detect threats and personally identifiable information?
• Consider how it can augment/inform task success survey

If you have any questions, please feel free to reach out if you have any questions and/or comments in the Slack channel.

3. Guidance check in - Laura Piper (DTO)

Check-in to see if anyone has any questions/challenges around guidance that we can help with.

Note about alerts. Analytics are showing that people are more attracted to link text in doormats at the navigation level. Please make sure you’re moving to more concise alerts (as recommended in guidance) and developing the destination web content to support them.

• Alert guidance

If you have any questions/challenges, email [[1]]

Are there any missing pieces of guidance that would be helpful? You can

comment in the #themes-thèmes Slack channel.

Ordre du jour

1. Mises à jour

2. Widget du succès de la page - Penny/Isabelle/Patrick

3. L'enregistrement de l'orientation - Laura

Notes et actions à prendre