Jan 20, 2016 10:39:45 AM by John Weathington
Do you use surveys to conduct analysis for your requirements? My guess is no. But I hope that will change after reading this article. Surveys are a great way to collect data to support your analyses. They serve the invaluable function of validating what you suspect your end users are thinking. I remember working with a large utility company in California where the technical team was trying to deploy an upgrade to their Geographic Information System (GIS). They were sure their end users knew the upgrade was coming; however, when I ran a simple survey, I discovered that more than half of the user base had no idea what was going on. This is very typical in my experience, and it emphasizes the value of running surveys. Let me cover some of the basics.
If you're wondering how to technically conduct a survey, it's pretty easy. If you're in the IT or IS department, it's likely that you have no idea what technology is used to conduct surveys at your company. But, your marketing department does. So ask around to see what they use and what it would take to get a license. Hopefully there's no cost, but if there is a cost, it's usually nominal compared to the value it will bring to your analysis. If you hit a dead-end, try asking your manager if you can use something like SurveyMonkey. I've had an account for years and it's easy to learn and use.
Once you have the technology piece worked out, you need to understand some basic principles. The most important thing to understand is that surveys are used for validation, not exploration. Before running any kind of survey, conduct interviews and/or focus groups to understand your fundamental themes. For instance, if you're trying to determine what features to include in your next release, start by conducting interviews with experts to see what they think end users will want. Then, use a survey to validate which features end users care about, with multiple-choice questions. As such, avoid using free-form text boxes in your surveys to collect ideas--this is not the right use for a survey.
Next, keep your surveys simple with four- to seven-point multiple-choice questions that are balanced. Balanced means that the middle-choice is neutral and, as you go out from the middle, each side has an equal level of contextual intensity. For example: Strongly Dislike, Somewhat Dislike, Indifferent, Somewhat Like, and Strongly Like. You can use an even number of choices to force end users to choose a side (i.e., there is no neutral choice).
Finally, keep your surveys simple. Most novices over-engineer surveys because they're trying to be too clever. It is hard enough to convince your end users to take any kind of survey, let alone a huge, complicated one that makes them think too much. If you've done your thematic analysis with interviews and/or focus groups, you really shouldn't have that many questions. Keep it down to one page with only a handful of questions.
Surveys are great tools that are largely overlooked by business analysts. Take some time to understand what technology your company uses to conduct surveys and get some basic education on best practices. I've given you some tips in this article and it's not hard to find an easy-to-read reference to cover the rest of the basics. Then, put your newfound knowledge into action! It's a lot of fun and you'll be pleasantly surprised at the results.
Written by John Weathington
John Weathington is President and CEO of Excellent Management Systems, Inc., a management consultancy that helps guide organizations to achieve strategic goals, improve critical processes, and leverage the power of information. For over 20 years, John has helped clients of all sizes including an impressive list of Fortune 100 firms to include Visa, PayPal (eBay), Hewlett Packard, Sun Microsystems, Hitachi Data Systems, Cisco, and Silicon Graphics. His unique blend of leadership, management, and technical talent and skills are a rare find in the consulting arena.