Help with intranet user research

18 Jan 2011 - 2:50am
1 year ago
6 replies
1390 reads
kevvstone
2011

Hi,


I was wondering if anyone had any advice or experience when it comes to intranet user research? Currently I am conducting a number of user interviews with colleagues across my company, but I am starting to question the size of my sample, and also how best to analyse the research.

Does anyone have any thoughts/ideas regarding my 2 current questions:


1.

My initial plan was to interview around 45 people, which absolutely is a lot of people – but the aim was to try and ensure that I had a cross section of all of the different departments (the questions are related to their work behaviours, information needs etc, and not confined to the Intranet itself). However, I am now mid-way through (about 24 people interviewed) and am finding that the general answers I am getting seem to be the same.

Does this mean I should stop my interviews? I had initially expected the different areas of the business to have different goals and behaviours around their work, but with the exception of 1-2 clearly defined groups, it appears that most areas of the business seem very similar.


2.

Does anyone have any advice as to how to analyse all of this data when it is completed? My initial thought was to do affinity mapping, however with such a large sample (and largely, only me doing the work) I am concerned this may be too big a task.

Another idea I had was to take the behavaiours/goals/characteristics of each group and then almagamate them to produce a single set of notes for the groups (i.e. As you would when making a persona). At present I am very open to any and all suggestions anyone may have regarding this analysis however.


Does anyone have any advice or experience relating to any of the following, as it would be greatly appreciated!

Many thanks,

Kevin
 

Comments

24 Jan 2011 - 6:34am
Ali Naqvi
2008

Hello kevsstone,

1:

according to Mike Kuniavsky, 7-9 people is sufficient.

I have done a similar research and only interviewed 7 people.

If you already have a general findings, no need to interview 45 people.

2:

Not sure if this can help you:

http://www.utsouthwestern.edu/vgn/images/portal/cit_56417/19/62/205397Guide_for_Focus_Group_Analysis.pdf

http://www.utsouthwestern.edu/vgn/images/portal/cit_56417/19/61/205402FG_Analysis_Workshop_-_FINAL.pdf

I could send you my report and how I analyzed the data for focus groups regarding an Intranet. BUT the stuff is in Danish so I assume it will be useless for you...

regards

Ali

 

 

24 Jan 2011 - 8:05am
Chauncey Wilson
2007

Magic numbers make me a bit nervous.  The number of participants in user research depends on several factors:   1.  The number of distinct user profiles or personas in the population of interest. If you have distinct user groups, then you need to consider whether you need to have a sample of from each group.  If you have 5 groups, then you may want to interview a sample from each group (so 5 groups with 5 interviews each might be 25 people).  With an intranet, you might have a smaller set of user groups and a generally more homogenous population but you still are likely to have several groups who use the intranet in different ways. 

2.  The importance of the design issue and the risk of failure. 3.  Whether your stakeholders will find a small sample credible. 4.  Whether you have existing data to supplement the interviews. 5.  The ability of interviews to get at the underlying issues.  You might need to consider logging data, diary studies, or observational studies depending on the issues around redesign. 6.  The skill and experience of the interviewer. 7. What percentage of user needs to you need to uncover?  This is often an unstated assumption. Edward McQuarrie discusses the "sampling frame" and how many people are needed in his book "Customer Visits". 

  Chauncey  

  On Mon, Jan 24, 2011 at 8:06 AM, Ali Naqvi <Ali@amroha.dk> wrote:

Hello kevsstone,

1:

according to Mike Kuniavsky, 7-9 people is sufficient.

I have done a similar research and only interviewed 7 people.

If you already have a general findings, no need to interview 45 people.

2:

Not sure if this can help you:

http://www.utsouthwestern.edu/vgn/images/portal/cit_56417/19/62/205397Guide_for_Focus_Group_Analysis.pdf [1]

http://www.utsouthwestern.edu/vgn/images/portal/cit_56417/19/61/205402FG_Analysis_Workshop_-_FINAL.pdf [2]

I could send you my report and how I analyzed the data for focus groups regarding an Intranet. BUT the stuff is in Danish so I assume it will be useless for you...

regards

Ali

 

 

(((Please leave all cont
24 Jan 2011 - 4:57pm
Yvonnia Martin
2009

Just as an aside from #3 of Chauncey's list: In my experience I would suggest sending out a survey as a supplement to combat any issues of stakeholders thinking that sample of participants is too small. Also, sending out the survey will help to cut down on  attempts at sabatoging your project  (I know this is PD 101, but still worth mentioning Laughing).

Just my 2 cents.

--Yvonnia

25 Jan 2011 - 9:41am
kevvstone
2011

Thanks for all the excellent comments.

After more thought about the interviews and these comments, I think I am going to reduce the amount of interviews overall, as clear user types seem to have formed. I will therefore use these to try and understand any patterns etc, and keep a note of the others who I have not interviewed - to be used in the future in case I need any refinement.

 

Yvonnia/Chauncey: A previous survey was undertaken some months ago in fact, and so I will hopefully be able to use this as additional material and also to ensure that a wide group of people have been consulted. I am currently very aware of the political aspects of such a project, and so I am trying to reduce any chance of this rearing its head where possible.

 

Thanks once again!

25 Jul 2013 - 9:33am
Antonio
2009

Hi Kev,

I am currently approaching a project that will require me to do some user research to initiate the implemenattion of a intranet. I was hoping I could ask you a couple of questions to understand you approach to this task.

Did you use any questionnaire before the interviews?
How long were you interviews? Were these 1-to-1?
Did you do a stakeholder interview before the user interview?

I would be fantastic if you could share any idea/experience on how to run user interviews for intranet design.

Best regards,

Antonio

26 Jul 2013 - 6:22am
Fredrik Matheson
2005

Hi Kev,

in my experience doing UX work on intranet and collaboration tools inside a company is quite different from customer-facing projects. You can know the users by name, sit with them as they do their work, see how the tool fits with all of the other things they're doing and how they negotiate with their tools to get their work done.

How many people do you need to interview? I've found that this is a factor of how big and diverse the organization is. You'll usually find that similar tasks are done in different ways across locations (geographic or in the building), specialties and departments. To succeed you'll need to look at the work itself, not just the workers.

Happily, Clay Spinuzzi has written a book on how to conduct this kind of research. It's called Topsight and is available on Kindle and in print. In it, you will find a guide on how do design your field study, how to conduct it and how to interpret the results. It's very hands-on and includes several tools for mapping the work and making the small, medium and large-scale issues visible. Topsight focuses on the tangible stuff – information, interfaces, tools, processes. You won't risk your intranet research project turning into a ethnographic study of corporate culture (which is in itself great, just not as valuable for our projects). 

I'd love to get your feedback on the book's design and content. Clay wrote it after teaching my team a half-year course in field studies and analysis methods from activity theory and actor-network theory. We've incorporated Topsight into several projects here at BEKK (where I work) and the results have been very positive. I'm interested in seeing how designers can use these methods to uncover how work is actually done and how we can help information flow more appropriately.

- Fredrik

|   ||  ||| | | ||| |   ||| | | |||  |  ||  | |  | || | ||   || || |||| || || | |||     ||    | ||| |   || |  | ||   |  || ||   ||  | |

Fredrik Matheson, IxDA Oslo
fredrik.matheson@ixda.no | ixda.no | @IxDAoslo + @movito

Syndicate content Get the feed