A Research Experiment — Help?

I know, I know… seems that all I write here lately is a request for contributing to surveys and research.  Too many, right?

Add the ones you get from every vendor, consultant, and even on Twitter and you feel abused.  Worse? the length of all these surveys and how few insights you get from that (infographics are the farthest thing from analysis and insights – they are The National Enquirer of the research world: all the sensationalistic data it’s not fit to print anywhere else).

I want to change that, and I would love your help.

With the help of my official customer service usage and adoption research sponsor ™ Salesforce we are aiming to change the model.

We were debriefing following the research we did last year (if you never got the data – its here; no report was done, we used the analysis in other ways… feel free to download and use, but you have to attribute thinkJar for the source of the data) and we came to three conclusions:

  1. Length.  Over the past five years I conducted this survey we added, molded, and improved left and right – that made the survey almost 30 questions long.  My rule of thumb for surveys (9 +/-2 questions) was totally violated.  Number of responses dropped dramatically (almost 70%) as the survey increased in length.  Worse, most questions became matrixes – it was almost impossible to understand and answer.  Too ambitious.
  2. Insights.  The more data the less insights we were able to get.  This is a correlation of the complexity of the survey (see above) that yield too few answers for proper cross tabs and the complexity of making sure that questions don’t contradict each other.  We tried, but there were still a few values that contradicted each other – happens often with long surveys with overlapping topics.
  3. Logistics.  Massaging data (cleaning incomplete records, finding the best records to follow up, conducting post-survey interviews) became a hassle.  Took too much time, again – see length above, and too hard to coordinate and aggregate.  Too many single cases as opposed to data points that aggregated into interesting trends and patterns.  Hard to write interesting insights when the work that went into it was more than the value that came out of it.

This is more than a mea culpa – it’s the catalyst to trying a new model – and we need your help.

  • We are running six mini-surveys (each no longer than 8-10 questions) that are more focused in specific topics:
    • Future Trends in Customer Service
    • Building Super Agents
    • Leveraging Mobile for customer service
    • Self-Service usage in customer service
    • Modernizing Customer Service
    • Social Customer Service
  • We want to deliver explicit insights for each survey, an evaluation of each topic in depth versus trying to compete with 6-8 different insights in a single report.
  • We want to expand the list of people we ask to participate and have them commit to either do all six, or at least two or more of the surveys instead of trying to find six different  groups of people to participate.

This is what I need from you:

  1. Give me your information, volunteer to participate in the surveys – become part of the community and we will provide you with first look at the data and insights for each survey.
  2. Take the first survey.  tell us what you think are future trends for customer service.
  3. Click-through the emails you will get in the future and take future surveys (don’t worry, will remind you).
  4. Come every four weeks to read the insights and data from the past survey (will also remind you)
  5. Enjoy
  6. Tell your friends, have them tell their friends, and their friends, and their — you get the idea…
  7. Let me know any comments or questions in the comments section right below

disclosure: yep, someone is sponsoring this research – come on, i told you above – Salesforce is my official customer service usage and adoption research sponsor ™.  they pay me, i pay my expenses, feed the kids, etc. and i give them a chance to participate in discussions on topics, questions, etc.  then i write the questions, conduct the surveys, conduct post-survey interviews, write all the content, publish it, and am fully responsible for everything.  nothing they say can change what i do and how it do it – nor can any other vendor. i know, incredible – right? someone paying to just be mentioned? they agree with me – we are trying to build bigger pies for more people to eat.  and yes, final veto for everything is mine.  period.

One thought on “A Research Experiment — Help?”

Comments are closed.