As usually, an insightful event! Though I have not worked on an intranet for a while, it was a pleasure to learn from the expert of the field. Unrelated curious observation: many people in the audience complained about issues with site search mechanism, and difficulty in correcting it. Our entire industry seem to be in the same boat 😉
Public sites sometimes used by employees to find internal information… because this is the only place they can find this information. As a result, this could lead to a resistance to changing public sites.
Allianz Intranet currently has 91% satisfaction rate and 74% of employees claim that they can find what they are looking for. The industry benchmark: 65% should be able to find what they are looking for.
Intranet redesign steps:
- Content inventory
- User testing
- Focus group – card sort (categorization)
- Focus group – naming conventions
- Intranet naming contest (get employees excited)
- User testing
- Contests (to introduce the new intranet)
Key – involve employees throughout the process.
Allianz has employee survey response rate at about 65%. Though the company surveys employees about 20 times a year and uses this information. The survey results define topics that executives (and managers) answer during general meetings. One of the questions is “What is the buzz?” Company’s management wants to be a part of the water cooler conversations rather than ignore them.
Communications department owns the intranet – departments can not decide where the information will be placed, though they receive advice from communications team. Communications team has a web writer, who places the content into needed location. The intranet exists for the users rather than catering to departments’ desires, that can be based on industry jargon.
All intranet data is shared with management twice a year.
Intranet search tool is not functioning well, and it is a known issue. Unfortunately, correction of the tool on the current platform would be prohibitively expensive. Communications team explained the situation to the employees directly – yes, replacing of the tool would cost the company too much, and we can find better use for this amount of money.
Two types of metrics are used:
- Engagement (how much users look at “my staff”)
- Business metrics (make an impact on the organization)
- “I feel well-informed” – Allianz at 94%
- “News are timely” – Allianz at 94%
- There is open and honest two-way communication at this company” – Allianz at 91%
Communications send employees only 6 emails a year. New (improved) intranet became more popular than email.
The site is updated Tuesday and Thursday at 1:00 pm, and, based on the survey results, it is considered by the majority of the employees just about right.
Leadership is involved: 20 members of leadership team writes stories – communications does not write stories for management, though they can edit and correct grammar. Ability to rate the story creates competition among writers.
Communications, who is viewed as consultants, recommends as short stories as possible. As it is an intranet… no worry about SEO 😉
Top intranet pages: benefits and careers. these sections were so important and popular, that communications had focus groups for each of the sections.
Top ten search terms are used to promote the searched information throughout the company in any available form (monitors in hallways, etc.).
Intranet-specific metrics (survey):
- I’m satisfied with the intranet
- I can find information
- I’m confident Intranet info is accurate
- Intranet has improved during last year
- Articles are easy to understand
Additional questions, which are the most beneficial for improvements:
- What are things you like?
- What would you improve?
Allianz intranet does use a slider with top articles, and everybody seem to like it.
Term confusion: “Archives” was not quite understood… was replaced with “What you missed” – questions related to “where can I find old articles?” disappeared.
Savvy approach: one executive was fixated on individual survey answers, and wanted to see changes based on these random requests. Solution: showing trends analysis before any samples from the survey.