07 September 2010

User Surveys for Intranets

Dorthe Raakjaer Jespersen (sorry for the English spelling) at J. Boye recently posted a blog entry “New Intranet: do we really need user surveys?”  It was a good read about how user-centric research is an excellent foundation for a new intranet. 

Early in the article, she points to somewhat typical exchanges between intranet managers and company management; the archetypical intranet manager management thinks surveys will help the intranet team better understand the needs of the users.  The equally archetypical manager thinks they need to start building immediately – surveys will take far to long to complete.  Dorthe goes on to provide an excellent set of talking points intranet managers can use to encourage management to agree to use surveys and highlights some of the potential risks in forging ahead without the user data.

However, once you’ve gotten approval, what’s next?  What questions do you ask?  What sort of feedback should you expect to get from your end users?  How many individuals should be included?  The answers to all of these questions do depend on what you need from your user community, the size of your organization and the scope of your intranet.  To help you get started, here are a set of basic questions we would recommend you include in your survey:

  1. In what department do you work?
    This question provides context to the answers.  Expect every department to have different needs and desires.  The question may also give you a picture of future departmental participation and support, as well as an early indicator of departmental adoption.
  2. What is your age range?
    This sounds like a question you’d want to avoid.  However, the answer will help you evaluate responses.  Many organizations we’ve worked with recently have somewhat older populations.  The age of an employee tends to affect what features of an intranet are compelling.  Your Intranet should reflect the population age average.  If you have a somewhat younger workforce, social features like status updates or blogs may be more compelling.  Older workforces will likely be drawn to a more “traditional” informational Intranet.  Keep in mind that you cannot and should not over generalize; age, in and of itself won’t tell you a lot.  However, the data will provide another useful dimension to the analysis.
  3. How often do you use the Internet to support your work?
    In some organizations, the existing Intranet is little used or there’s no Intranet at all.  In its place, many employees will turn to public sources of content and information.  This specific question will help you gauge how much external vs. internal sources of information help people get their work done.  In fact, the answers to this question should be a limited list of options like: no access, my work is paper-based, I know the right person to ask, not interested, etc.   A side benefit of this question may be the discovery  of what services to offer through a newly designed intranet (assuming you follow-up this question with one that asks about their specific usage).
  4. How often do you use the existing Intranet?
    It’s likely if you’re checking your server logs, you already know the answer to this question.  However, getting direct feedback is helpful is validating your objective data.   Where possible, include a list of selectable ranges of time like: once a day, once to twice a week, monthly, etc.
  5. What functionality, types of information or content is important to you?
    Provide your survey participants with a list of functionality (e.g. applications), content and information that you’re either considering or have surfaced through the Intranet.  Let them choose what’s most important to them.  Use the answers to help guide your development.  More than one of our clients has chosen features or content poorly without this data (leading to community dissatisfaction with the Intranet).  Further, the answers can help validate or invalidate the inevitable “nice to have” functionality that often is introduced during development; if it wasn’t important to your user community, it shouldn’t be important enough to get included in the Intranet.  You can also follow-up this question with a free-text question allow people to suggest items that weren’t included in the list.

The preceding five questions should really be the start of about a ten (10) question or less survey.  In all cases, try to make survey questions very specific to your organization.  For example, you could ask about the location of the employee (if you have multiple physical locations), ask a matrix question that allows the participant to agree or disagree with various statements regarding your intranet (or what they’d like to see) and ask about usage outside of work (e.g. at home or through a mobile device), if those access methods are important. 

The survey should be one of the first exercises you perform when gathering requirements and/or intelligence on a new or redesigned intranet.  They generally provide very valuable insight into user behavior and needs, plus has the added benefit of allowing you to validate features, functions and “requirements” against direct user feedback. 

In closing, here are some other tips:

  • Surveys should be only one of a few data sets used to evaluate your needs.  While it’s a good foundational insight source, it can also be misleading; if someone doesn’t have a good answer for a specific question, they may simply make something up.  As a result, the size of the survey population has to be significant enough to counteract errant responses.
  • When designing the survey, keep the overall survey length to no more than 4 to 5 “pages”
  • The survey should take between 10 to 15 minutes to complete; longer surveys may work, but you’ll get lower response rates.
  • As mentioned earlier, the surveyed population should be significant, but also representative.  If you 10,000 employees, interviewing 50 people is insufficient as it represents less than 1% of your population.  Further, if you have a mix of factory employees and corporate, desk-bound employees, be sure to get people from both communities to participate.
  • Ask relevant follow-up questions.  For example, if you ask about access methods (e.g. from home, mobile devices, etc), ask if they access the Intranet from home, because they have no access in the workplace (e.g. many factory workers have restrictive work rules that may prevent access to PCs during their shifts).
  • Offering incentives for participation will greatly improve participant involvement.  Some clients offer gift cards for all participants, while others use the “win an iPod” approach with a participant chosen at random.  What you choose should reflect your culture and drive the sort participatory behavior you desire.
  • Follow-up this survey with a similar version post launch.  A follow-up survey will give you very clear measures of improvement and help to establish basic metrics for judging on-going success.  For a discussion of other Intranet metrics, see a great article by Toby Ward on Intranet Metrics.