Cvent Survey

This morning driving into work, the morning show I listen to was discussing a recent story about a woman who quit her job to go to the Royal Wedding. She'd requested the time off, booked her tickets, and at the last minute was told they couldn't get coverage so she couldn't go. Instead of staying home and going to work, she quit. In an interview with KMBC, she said she wanted to experience the modern version of a fairy tale.

The story spurred the question, is there an event you would quit your job for: a ride in a space ship, go to the Olympics, see your significant other who is home on leave?

Job satisfaction was the dividing line between those who said they would and those who said they would not. The radio personalities all gushed about how much they love their jobs and couldn't imagine quitting for anything other than an extreme family emergency where the station wouldn't work with them. I don't think it's any secret that employees who are satisfied in their jobs are more motivated, loyal, dedicated and productive. However, in the last few years, companies have gotten lazy. They've begun to feel secure that the perception of job security alone would keep their workers from leaving. This isn't the first story we've shared to the contrary. In fact, CM Arnold recently wrote about how workers quit just because they're dissatisfied. Callers shared stories about how they quit just because they wanted to watch the Superbowl. One caller shared a story about how her husband was coming home from Afghanistan on emergency leave for bi-pass surgery, and her office wouldn't let her take off because her boss was going on a cruise. I was mortified! How can an organization expect employees to go to bat for them, to put in the extra hours, to really go the extra mile when the organization isn't behind the employees?

At the end of the day, no organization is perfect and not every employee is going to be happy and love their job. However, it's management and HR's responsibility to try to maximize job satisfaction and reduce churn, which ultimately will impact revenue growth. It's tough to meet aggressive growth goals with unmotivated employees or high turn over. An entire Voice of the Employee program starts and ends with taking a good hard look at the workforce and what they need to be successful. Feedback is the linchpin of any good Voice of the Employee program. Without employee feedback, it's tough to make the right changes and measure employee satisfaction and engagement. Very similar to the framework of a Voice of the Customer program, a Voice of the Employee program should look something like this:


If your Voice of the Employee program is just in its infancy, make sure you have management buy in and set clear goals. Start small and prove the value, if you try to boil the ocean you'll likely fail and the program will stall. Most importantly, for any employee feedback management program, make sure you listen to your workforce and don't get to a place where they're unhappy and the Royal Wedding leads to resignations.

Online survey tools are excellent for capturing consumer and B2B marketing data, but the fun begins when we analyze the data. We can review questions individually or in combination to isolate patterns.

Univariate frequencies look at responses to one question at a time. The tables below show two questions, one for work satisfaction and the other for perceptions regarding compensation. Over 9,600 respondents answered the job satisfaction question with 383 not responding (‘missing’). Looking at the valid percent column approximately 44% reported being ‘satisfied’ with less than 15% reporting they are not satisfied with their current position. The second table shows that 58% of respondents feel they are not fairly paid.

Satisfied at work

		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Satisfied	4228	42.0	43.7	43.7
	Somewhat satisfied	4036	40.1	41.7	85.3
	Not Satisfied	1419	14.1	14.7	100.0
	Total	9683	96.2	100.0
Missing	System	383	3.8
Total		10066	100.0

Feelings on Compensation

		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Fairly paid	4206	41.8	41.8	41.8
	Unfairly paid	5860	58.2	58.2	100.0
	Total	10066	100.0	100.0

Individually these questions are interesting, but we should be asking if there is a relationship between the questions. One of the simplest ways to assess a relationship between nominal or ordinal scaled questions is to perform a cross tabulation. This provides a joint distribution. We assess the relationship’s strength by employing a common statistic known as the chi-square. This metric compares observed frequencies with what would be expected if there were no relationship between the questions.

Only respondents who completed both questions will be included in the analysis. The tables are often referred to by the number of columns and rows they use (a 2 x 3 table has two columns and three rows). The more columns and rows used, the more difficult it will be to extract meaningful information. It is often helpful to collapse the number of rows or columns to facilitate analysis. The 2 x 3 table below shows the joint distribution of work satisfaction and feelings on compensation.

Satisfied at work * Feelings on compensation cross-tabulation

			Feelings on compensation
			Fairly Paid	Unfairly Paid	Total
Satisfied at work	Satisfied	Count	2442	1786	4228
		% within satisfied at work	57.8	42.2	100.0
		% withing compensation	59.6	32.0	43.7
	Somewhat Satisfied	Count	1306	2730	4036
		% within satisfied at work	32.4	67.6	100.0
		% withing compensation	31.9	48.8	41.7
	Not Satisfied	Count	346	073	1419
		% within satisfied at work	24.4	75.6	100.0
		% withing compensation	8.5	19.2	14.7
Total		Count	4094	5589	9683
		% within satisfied at work	42.3	57.7	100.0
		% withing compensation	100.0	100.0	100.0

Chi-Square Tests

	Value	df	Asymp. Sig (2-sided)
Pearson Chi-Square	764.064a	2	.000
Likelihood Ratio	774.612	2	.000
Linear-by-Linear Association	699.614	1	.000
N of Valid Cases	9683

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 599.96.

The first place to look is at the marginal percentages in the ‘total’ columns and rows. Note, the tabled percentages may be different than those from individual questions because of missing data. From the table we see that 57.7% feel they are unfairly paid and that this varies between 42.2% for those who are satisfied and 75.6% for those who are not satisfied with their job. Similar variation can be seen in the ‘fairly paid’ column. This level of variation is reflected in the large chi-square value (764.1). Note, however, that larger sample sizes will inflate chi-square values. If we were to randomly create a subset of the data, the chi-square value would not be as large, but it should still be significant (p <= .000).

Since the chi-square value is large and significant, we can say that job satisfaction and feelings about compensation are related, in this case in a negative fashion. However, we cannot say that one variable ‘causes’ another. That would be outside the scope of this analysis.

Whether using a form to create a survey or using a form for lead generation and customer acquisition, form building is more of an art than a science.  There are general things you should know for form building, especially if you are using forms for customer relationship marketing.  Here are a few tips:

• Use radio buttons as much as possible.  While open fields give you greater insights, radio buttons are user-friendly, simple, and fast.
• Include open text fields first on your form.  Making customers put in some effort early on in the form is OK as long as the rest of your form is simple.  This way, you tend to avoid customer drop-off. Some open fields are expected, but try to keep them to a minimum.
• Keep the number of questions to a minimum.  Most companies want as much information as possible from their customers.  Unfortunately, customers (unless incented) are not as willing to put in time for something that is not mutually beneficial.  Keep your relationship marketing or survey forms to a minimal amount of questions so as to keep your customers happy.

Having a teenage daughter involved in high school chorus has certainly opened my eyes to new ways of thinking. This was further reiterated after reading a post by Laura Patterson on growing the bottom line through voice of the customer (VoC) research. In her post, she indicated that firms with an established VoC program saw improvements in customer loyalty, willingness to recommend and increased incremental purchases. Gains in these areas will add to the bottom line.

Just as a chorus has many voices which need to act in sync, so to does a functional VoC program. As researchers, we need to continuously think about our customers, both internal and external. Each of these groups has a ‘voice’ and it may indeed be as different as a tenor is from a soprano. The success of any marketing research initiative, be it consumer or B2B focused, will depend in part on the degree in which it meets the needs of its intended segment. Remember each segment has a distinct voice!

When we think customers, we should not limit our perspective by thinking all customers are the same. If we can leverage transactional and marketing data in our CRM systems then we can clearly see not all customers are alike. In fact, there may be several distinct segments within our customer base alone. The question then becomes: can one survey be used to capture the diversity of opinions that can be expected to arise from different customer groups?

As we continue to expand our thinking to include prospects and internal clients (e.g. sales force, distributors, customer service, product management etc.), it becomes easier to see that we will need targeted surveys, perhaps coupled with some form of qualitative research if we are to hear the different voices in the chorus. Online survey tools when integrated with CRM systems such as salesforce.com can be used to distribute surveys tailored to specific groups.

Like a vocal instructor who sets standards differentiated for the skill levels within his or her chorus, we should think about different metrics for the groups we target in our VoC program. More valuable customers perhaps should be given additional weight in the reporting of program statistics. Conversely, if a company is in growth mode then the weight can be shifted toward measuring the concerns of new customers and prospects. It is important to keep the initiative’s reporting and measurement goals aligned with current business strategy.

When conducting market research (and in some cases when conducting a survey), you want to ensure that you have a representative sample that meets your brand and project objectives.  If you are not sure how to define a representative sample, here are a few tips to use for your market research or your survey:

• Determine an "n." "n" is most often defined as a sample set and can be statistically calculated to ensure you have enough customer feedback to get significant results.  Make sure you have defined an appropriate "n" if you need to, and that you recruit enough customers for your project.
• Determine defining criteria.  Recruiting criteria can differ depending on a range of factors, but ensuring your representative sample has the basic set of criteria needed for your market research or survey is critical to obtaining valid results.  If you need women, aged 30-45, who have one or more children then those parameters become your defining criteria for your sample set.
• Track other details from your representative sample. Sometimes, when recruiting on one set of criteria, you end up uncovering other trends relevant for your research or survey.  For example, in the criteria defined above, perhaps all the mother's end up being stay-at-home moms.  In the latter case, moms who raise their kids might offer a new insight into your brand that you may not have otherwise considered or realized.  Track these details - they can become critical in later stages of your research or survey.

As someone who regularly creates business cases for my team and writes persuasively, finding data and research to validate your idea or hypothesis is key to ensuring that people buy into your thinking.  This is especially important, regardless of whether you're focus is B2B or B2C, where others are competing for the same dollars as you.  Where do you find research and data that helps support your business case though?  Here are a few places that I like to look.

Start with what you have.

In many cases, I have data provided to me from the client about (for example) types of traffic that their site is getting, or competitive data.  These are forms of market intelligence because they give you some baseline to understand what is on your client's radar and how they see/track their business.

Try well-known sources.

Sites and papers like McKinsey, TNS Research, and Forester, are all great resources to search for information and statistics for whatever business case that you are trying to make.  Additionally, these data warehouses tend to have various kinds of market intelligence that spans many business verticals so they probably have something for you.

When in doubt, just try an internet search.

Today, people blog, post, broadcast, and tweet the information they want peers to see.  The internet can be a great way to find bits and pieces of data that you would not otherwise have known or come across in your search for market intelligence.

One of the primary advantages of an online survey tool such as Cvent is the ability of the researcher to incorporate audio/visual content in the context of their survey. This might take the form of sample audio formats testing preference for radio commercials, ranking page design or logo alternatives, or sample television ads.

Most platforms now incorporate a slate of interactive questions. Building off the scenarios above a researcher may want to have survey respondents view a series of visuals and provide a rank order of preference. This allows the survey designer to put visual elements before the respondent allowing them to make their choice based on actual visual experience. In the graphic below Zyzx, Inc. is interested in changing their corporate mascot and wants to know what their current customers think about three options.


The order in which a series of stimuli are presented can impact the rankings of those stimuli. To accommodate for this, it is important to randomize the presentation order. This is similar to the concept of randomizing the order of questions in a multi-item battery.

This rank order data should be analyzed against other variables in the survey including customer segments, demographics, prior purchase history, etc. in order to isolate potential differences between these key groups.

Visuals can be uploaded and stored in a graphics library for use on multiple projects, both on the actual survey and in subsequent reporting. Incorporating visual (or audio elements) with interactive questions allows the researcher to create a more engaging survey, while maintaining the structure of a controlled experiment.

Last month Facebook launched a new online feature called Facebook Questions which allows you to easily poll your friends and fans. It's similar services like Yahoo! Answers, Quora, or Mahalo, but it's nowhere near the online survey software that's available on the web.

The greatest advantage of Facebook Questions is that it’s already integrated into the Facebook ecosystem. You simply write up a question, create some multiple choice options, and post it to your wall for your friends or fans to answer. This makes it easy to compile quantitative data, as opposed to asking a question in plain text and then sifting through a list of comments.

Facebook Questions can be used on your personal profile as well as Pages (but it's not yet available for Groups) so as a business with lots of 'likes' you’re able to quickly gather quanitifiable market research from your existing community and on a fairly "quick and dirty" scale.

For example, say that you operate a bicycle shop with a cycling club that meets regularly, and you need to decide between three different group ride locations for next week. Why not crowd source the decision to your Facebook fans (likely your best customers) and ask them what route they prefer? With a large enough sample size, a perfectly split result is quite rare, so you should be able to make decisions more confidently.

The one tricky part about Facebook Questions is that you can only ask one question at a time, so if you need more probing answers (like what do they mean by "other") you'll need to stick with online surveys for anything substantive. It's great fun to watch Facebook evolve, but traditional online research is at no risk of disappearing anytime soon.

As we head into conference season across business verticals, the focus for many companies will be getting people into their booth and detailing them about their brand or company.  But once people come into your booth, surveying them can be just as critical as getting your brand message across.  Here are some sample conference survey questions to get valuable customer feedback:

What caused the customer to walk through the booth?

Understanding what attracted your customers to come through your booth can help your team amplify foot traffic into your booth at your next conference.  Maybe customers know the brand name; maybe your booth had an outstanding visual display; or maybe your conference booth team was very friendly and inviting.  Understanding this aspect of conferences bu surveying your audience can be key throughout conference season.

What aspect of the booth stood out most to your customers once in the booth?

Knowing whether the main booth display is doing its job is another key survey question to ask your customers. If your main booth display is not "working" for the purpose for which it was designed, reconsidering how to make your main booth display more customer-friendly can help you optimize your conference presence.

What were the key messages that customers took away from the booth?

Every company or brand has an objective and/or goal that it wants its customers to gain from stepping into their booth.  Understanding whether your message was conveyed to your customers and stuck with your customers is a good gauge of how well your booth achieved its desired company/brand results.

Businesses conduct informal surveys on Twitter all the time. They want to know what their customers think about the products or services they provide. Sometimes they want to let their current and pontential customers know about new products and services that will be available in the future. As noted in a previous post, that's great when done in extreme moderation.

Several marketing and business experts participated in an email interview that led to the creation of the following list of Twitter marketing Dos and Don'ts:

DO - Address customer problems in real time. Be there whenever the customer or prospect may need you. If possible, even before they need you, said Lindy Moses, marketing associate for FabulousSavings.com. It's all about being available.

DO - Use Twitter as a source of information for customers and prospects. Don't just gather information from your followers for your market research projects. Share useful information, tips and links [that will help] your followers to see you as a credible source, said Holly Fisher, executive director for Charleston Step Ahead, Inc.

DO - Start by engaging people you already know. Find current clients, prospects and friends and have conversations with them, recommended Jeff Gibbard of True Voice Media, LLC. Acknowlege their content and participate in the community that emerges.

DON'T - Bombard your followers with your company's brand messages. A big no-no is sending automated direct messages to everyone who follows the company, cautioned the director of marketing for Reach Local, Mike Merrill.

DON'T - Retweet yourself to the point of distraction. Over-targeting individuals through repeated use of the same messages or targeting and retargeting the individuals is as annoying as spam email or political flyers in your door, said Matt Van Hoven, the communications director for Skinny NYC.

Whether you use Twitter as a way to conduct informal surveys and market research, to engage with customers and prospects, or both, heeding the warnings and following the advice of those who have done it successfully could help you increase business and promote your company's brand naturally.