Monday, March 11, 2013
Inane CIO Surveys
We are conducting a survey among the top CIOs on IT trends, data centre efficiency, Cloud computing adoption, business analytics, mobility, leadership, top 5/10 priorities, technology priorities, top 5/10 IT challenges, business challenges, social media adoption, big data, security, … phew ! You name it and it is there; every day I get a couple of invites to participate in surveys with varying time indicated, from 5 minutes to an hour. They are run by all kinds of IT vendors, research companies and publications.
Surveys normally have MCQs (Multiple Choice Questions) or a matrix in which you select different options; some also want descriptive answers. Most of them end up taking twice as long to what they mention and what they would like you to believe. For example if a survey indicates that it will take no more than 10 minutes of your time, in all probability it will take 30 unless you are a speed reader. If you actually read all the text in the MCQ before you check/click your answers, then it may take longer !
Why do CIOs participate in these surveys ? They are busy professionals with paucity of time; I think it is to put across their views and get to know what others think (most surveys promise to send the results to participants, 50% do so too). But nowadays they come with incentives attached; take the survey and you stand a chance to win your favourite gadget/device or a shopping site voucher or …. The incentive increases participation rates, gravely said one such surveyor; from 5% to about 20%.
The first page normally gets genuine thoughtful answers; then you hit a block with questions where answers don’t reflect your context or reality; there is no way to skip the question, you end up selecting a random option. Questions are constructed in a way that support what the surveyor wants the outcome to be. Soon you hit a complex grid or matrix where you are expected to balance options in a way that you select one option per row whether it matters to you or not. The more interesting ones are when you keep track of totals in a column across multiple columns.
That is when most respondents start randomly marking answers or rush through the pages. Any online survey or questionnaire that requires more than 5 minutes or has more than 10 MCQs tends to get into indifference zone. What you get is random responses or a middle path or a horizontal row of answers if the questions use a scale of 1-5 or 1-10. Many are kind and desert the survey mid-way rather than input garbage. Why are most surveys so painfully irritating and difficult to respond to ?
You can now understand why most CIOs don’t agree with the results; because the input is rarely a reflection of reality of the participants. The responses appear to be from a different planet, the analysis bewildering, and the implications or actions disconnected from what makes sense. CIOs blame the outcome when they largely respond with indifference to surveys that seek their inputs on what matters to them. The starting point is the survey itself. So how to overcome this situation ?
To begin with reduce the complexity of the information required. If you want a split of percentages across routers, switches, wireless, bandwidth, network management, which is a subset of hardware, which is a part of the infrastructure, then you will only get what you deserve. CIOs do not measure these, neither do IT Managers. Ask questions that can be answered without a calculator or the IT budget spread sheet which has numbers, not percentages. If I don’t do social media, then I need an option “Not Applicable” or let me simply skip the question.
Now I rarely participate in surveys and if I do where possible put in textual answers with my views. Surprisingly no one has as yet written or called me back despite having my email id or cell number. Do people really read the answers ? Do they care about what is being said if their agenda or hypothesis is met ? Maybe the surveys are just to say that we conducted a survey and keep the data aside to publish what you wanted to anyway.