Podcast: Play in new window | Download (Duration: 12:00 — 22.0MB) | Embed
Subscribe: Google Podcasts | Email | RSS
Due to a technical glitch, the first version of this podcast failed to upload fully. The full version can be accessed using the player above or via the “Download” link.
We see numerous surveys, especially at this time of year when predictions and year-end recaps flourish. Many of them are just deleted unread because they aren’t newsworthy or have one of several flaws. So we put together this quick podcast that we hope can help you craft better surveys and hopefully get our attention for next year’s roundups.
Surveys often use too small sample sizes. Nowadays you should have at least 1,000 respondents that are culled from a global audience if you want any impact. Take a look at this survey from Bitdefender that interviewed 1,050 IT professionals in several countries to find out their cloud security purchase decisions. That is authoritative. Sometimes you can get away with fewer responses, if the respondents are highly qualified. An example is this piece that Paul wrote about where firms aren’t prepared for industrial IoT. It surveyed a few hundred people, but all were decision-makers.
Question design is also critical, and while we don’t delve into the specifics in this podcast, it’s important to not be leading the witness but trying to go after statements of fact. You want to ask about evidence that is directly observable, not just their subjective feelings about old chestnuts.
Stories that lend themselves towards “man bites dog” or unexpected results are always appreciated. Contrast this with this Barracuda survey, which found that 74 percent of respondents stated that security concerns restrict their organization’s migration to the public cloud and have prevented widespread cloud adoption. Really, that is news? Not!
Also good are surveys that come from unexpected sources or break new ground, such as this one that Paul wrote recently on how poor capacity planning can lead to wasting money on cloud computing.
Finally, take a look at places that we often cite in our own work, such as studies from Ponemon on the cost of data breaches or the Verizon Data Breach Investigations Reports, both of which have been fielded for years and are encyclopedic, comprehensive and analytical.
:49? 1.5 Mb? I think the file got clipped. Please re-upload!
Paul Gillin says
Thanks. The full version is now posted.