Part III: The Dream of SwissPeaks
It seems apt that this final part of the SwissPeaks trilogy is being written in Lucerne, the place where it all started. I haven’t been here in over 9 months but now, sitting in the KKL World Café looking out across the lake at the buildings of the Old Town, I ponder over what the bigger picture for SwissPeaks is.
My vision for SwissPeaks is two-fold. We want to use modern day software efficiently and correctly in order to:
1) Support interviewers and respondents
2) Support end users (researchers and decision makers) retrieve insightful data and information.
Although the industry has evolved since I first chanced into the world of market research as a Software Engineering undergraduate. At its core the industry still relies on some pretty old and clunky practices.
Firstly, we can’t stress enough that strong data management set up – surrounding questionnaire and interview script development – has a huge impact on the credibility of the end product. For example, we have worked on many projects where the task is multi-country and where each agency in each country script the questionnaire into CAPI unilaterally. The net result? A survey for each country that cannot thereafter be fused together with the other surveys into a “homogenised” project dataset. We believe that projects with multiple surveys should allow an end user to view the data and information as one project, rather than a set of discrete surveys that then have to be interrogated individually.
Additionally, CAPI, the modern face-to-face data collection platform, could be much more structured, efficient and effective. Getting credible, accurate data directly from the data collection stage is key and CAPI, as we have seen it, is seldom used in the way that we believe it should be.
The “script shell” as I’ve coined it, is the predefined script that contains the general survey actions: the interviewer control checks, interviewer help features, routines for masking, routines for respondent selection and so on. SwissPeaks have predominantly been working on projects based in Africa and what we and our partners afriQuest often find is what we like to call “Uhuru-Parking” (read about it here if you’re unsure what I’m getting at).
Getting credible, accurate data directly from the data collection stage is key and a predefined script will facilitate that. This way the interview doesn’t become arduous or mundane and interviewers take pride in their work, rather than finding ways to cut corners and seek an easier route to getting paid. Now that is a dream!
We also want to use modern day software efficiently and correctly in order to help end users retrieve insightful data quickly. Typically, the researcher will spend a considerable amount of time setting up a deck of PowerPoint charts and slides and have little time to interrogate and mine insight from the data.
As data visualisation, online analysis techniques and dashboards are becoming more abundant in the work place, we need to get researchers using the data and information they have effectively. We want more actual research and less “Microsoft Office-esque” building of charts, tables and slides.
We continue to invest into products like Tableau and our own bespoke online analysis solutions. We’re aiming to have a collection of applications available to our clients and in our existing online analysis solutions we pride ourselves in being able to:
Overall, our dream is to make the project experience more efficient, more modern, more insightful and most importantly, to do this in Africa – no more Uhuru-Parking! We want to see efficient data management techniques, data collection platforms that are used optimally, and researchers/end users tapping into their data readily and with more understanding. Over the next few months we will keep you updated on our data management journey, sharing our practices and quality charter with you and also bringing our showcase out into the public domain.
And with that, I return to sitting here and relaxing in Lucerne with my cup of tea. We have a lot to do over the next few months so I can’t see that I’ll get to do this for a while now…
