Voice of customer This year’s buzz word in the world of web analytics is "Voice of Customer" or VOC for short. Essentially this boils down to presenting a survey to your web visitors asking them to respond to questions that can be used to ascertain how they feel about the web experience they have had.

Why voice of customer surveys are so useful

As you are no doubt aware, web analytics tools and methodologies are great for telling you the "what" and the "when" of your web site visitors. That is, what happened (a goal conversion event, a transaction, a specific pageview or a combination of pageviews etc. or any kind of engagement on your site) and when it happened (time/date, do they repeat the same thing over again and at what frequency etc.). This is quantitative data that is invaluable for identifying poor performing pages and poorly targeted marketing campaigns.

However, the missing link from web analytics has always been the qualitative data - the "why". For example, why did 30% of visitors leave your web site on that particular page, only view one page (bounce), not convert, not buy, not contact you, stay for less than 10 seconds etc. The only way to obtain such information is to ask them why, and a survey does just that.

In fact, without a survey, ascertaining qualitative information is often guess work. For example, following a site redesign, why has your time on site metric increased? Is it because users are more engaged with your content, or is it because they are lost in your new navigation layout? Taking a guess can lead to a very invalid assumption about your visitor satisfaction.

Surveys bridge the gap between anonymous, aggregate web analytics data (traffic) and the views of your visitors (personalised responses from individuals). As an aside, I strongly recommend that personalised responses also remain anonymous for best practice privacy reasons, unless of course the visitor expressly wishes to give their personal information.

When voice of customer surveys can damage your brand

So with so much to benefit from deploying a VOC survey, what can be bad…?

What I find quite astounding is the current trend to use pop-up windows as the method of survey deployment. That is, an unsolicited window pops up in front of the visitor (usually selecting visitors at random) requesting them to participate in a survey. By unsolicited, I mean the opening of an additional window without any action or knowledge by the visitor. This is not the same as a visitor’s action opening a new window - for example, the clicking a link or button.

Pop-up windows are one of the oldest and most annoying forms of interruption marketing on the web - a phrase I borrow from Seth Godin (whose books I highly recommend for viewing the bigger picture of digital marketing).

They are so annoying that they sparked the creation of a whole industry of anti-popup and ad blocking software in the late 90s and early 2000s. In fact, the pop-up blocking capabilities of the Google Toolbar was a main reason for its success – one of the most popular software releases ever, with an estimated 100+ million downloads to date.

Unsolicited pop-ups are simply to email spam. That is, hated by Internet users. Their use to display a survey, not only puts you in the category of a spammer – annoying your visitors, it is also likely to skew results. For example, consider the following scenario:

A happy visitor on your site enjoying their user experience becomes irritated with an unsolicited pop-up requesting their feedback. They either leave your web site, decreasing your survey participation rate, or provide negative feedback because of the annoyance.

Of course, it is possible they welcome the option to provide feedback, but how likely is that in reality? Probably just as likely that they would want to receive a spam email from you. It happens, but it’s so rare that that damage to your brand caused by the interruption, far out weighs any feedback gain.

For this reason, survey participation rates using pop-ups are very low - rarely rising above 1%.

How to increase survey participation rates and get better responses from the correct target group without hurting your brand

Here are some alternative ways at soliciting feedback from your visitors without the use of an interuppting pop-up. The key is to get as many engaged visitors to take part in your survey as possible – a higher participation increases the likelihood of obtaining a statistically relevant outcome.

As a guide, if you have 10,000 visitors, you need survey results from 370 visitors to be 95% confident in your results to a level of + or - 5%. That’s a participation rate of 3.7%

Travel example Retail example How Google do it
BBC SME website

The rationale of these examples is to solicit feedback from visitors at both ends of the user-experience spectrum. Why? Because those visitors that are unhappy with their experience are always more likely to provide their views. This leads to the “squeaky wheels” syndrome. That is, you have to be careful not to assume that everyone feels the same way just because you have not received as much positive feedback.

Check list for a best practise deployment of a survey

  1. Avoid pop-ups – allow users to provide their feedback using a standard text link or button
  2. Keep it simple - use transparent language for requesting participation
  3. Add your participation request close to the call to action or user event you wish to gain feedback on
  4. Keep participation anonymous – or keep personal information optional
  5. Avoid random participation – provide all your visitors the choice to send their feedback. If people want to talk with you, they will!
  6. Deploy continuously – as with web analytics data, a continuous and ongoing survey system provides you with much greater insight than running a survey for a single campaign or set time period.
  7. Integrate survey results with web analytics click stream data (the subject of a future post )

Try my survey

I am obviously keen to understand my readership’s details and hear your voice - what countries are you from (my publisher first suggested 80% of book purchasers will be American, are you?), is there a need for me to publish in other languages, what type of organisations do you work for etc. etc.

Rather than interrupt you with pop-ups, I simply post a link to the here using the excellent SurveyMonkey tool.

The goal is to reach a 3% participation rate over a 3 month period, calculated as total respondents / total number of visits. I’ll post back the anonymous results as to how successful I am at getting respondents in a future post.

1 Star2 Stars3 Stars4 Stars5 Stars (7 votes, average: 4.14 out of 5)
Loading ... Loading ...
Email This Post Email This Post     Print This Post Print This Post