Implementing a “popup” survey is a (mostly) trivial task. Our solution at Surveylab requires 1 file to be uploaded to the website and then a couple of lines of code are inserted onto the page(s) of the website where the survey popup is to appear. Other survey software suppliers have a similar installation process.
Visitors to the website will see the popup, and perhaps 2% click through to take the survey, depending on a whole host of factors such as where the popup is sited, the message, why the visitor is on the site in the first place, and so on. We once managed to get a 27% response rate (in 2002, for visitors who completed a mortgage quote enquiry), but most surveys have seen less than 10% and some just 1-2%. Which is fine if you have enough traffic, and depending on your research objectives…
Making a popup survey into a genuine exit survey so that when the visitor leaves your website a message displays asking them to take part in a very quick exit survey is technically possible… but it’s not pretty!
Read the text in the screenshot below (read it carefully).
There is no way to edit the first and last sentences of the popup message, and clicking the [OK] button will continue the visitor’s journey away from the website. If the user clicks [Cancel], we can then display the exit survey. A big IF, because this is counter-intuitive – “OK, I’ll take the survey,” visitor says clicking OK, “hey! Where’s the survey?“
I have no data to back up this view, but I believe that if a visitor has clicked away from the website or is closing the browser they’re very unlikely to pause to take an exit survey. You would be better off looking at your site analytics.
And actually, that’s not a bad idea. Site Analytics (analysing your websites traffic logs) can’t tell you everything (e.g. age and gender of visitor) but it can often tell you what they were searching for, what website they clicked through from, etc. as well as their path through the site and which pages have higher “bounce” rates. Tip – install a popup here to engage these users about a survey.
Dan Wardle
We used to get asked this question a lot – an enquiry would start out asking for help with an online survey, but ultimately what the enquirer really needed help with was acquiring a suitable email list. If this is you – then you need a specialist list provider such as Sample Answers, Ciao or Mo’web Research (and there are lots of others).
A post on Word to the Wise – another list purchase horror story – reminded me of a very similar experience a few years ago when a client supplied us their email list (we knew it was a purchased list, from a reputable, well known company) and on inspection it contained huge volumes of info@, sales@ addresses. We sent it back because there was no way we were going to email the list.
Never purchase email lists. If your survey needs the general population (or a niche segment) to take part, find a specialist list provider that will send emails to their database (you won’t ever see the actual email addresses). Usually, they charge on a per response basis which can vary enormously depending on the demographics and volumes required.
Dan Wardle
As reported in The Times today, in response to a shareholder’s question at BT’s AGM, Chief Executive Ian Livingston announced that BT is moving 2,000 call centre jobs back from India to the UK. This apparently resulted in a huge round of applause perhaps reflecting the UK consumers’ hatred of the offshore call centre.
The move is probably more about the UK economy and minimising redundancies than improving BT’s customer service but it is a great step forward.
Readers may recall that Surveylab conducts the UK Customer Care study and last year reported specifically on the total dissatisfaction UK consumers have towards offshore call centres. At that time we reported that 66% of our respondents reckoned their call to a UK company ended up overseas but they were far from happy with the experience:
The vast majority of respondents experienced one or more problems talking to the overseas call centre:
- 88% found it difficult to understand the agent
- 57% felt that the agent had difficulty understanding the caller
- 42% felt the agent had insufficient knowledge of the product or service involved
- 41% felt that the call centre agent relied too much on scripts
- 38% felt that the agent had no or limited authority to resolve the caller’s issue
- 35% felt there was poor call quality or noise on the line
- 21% said that promises made were not kept
One of the reasons given for “offshoring” was to make it easier for customers to contact but only 5% of respondents thought it was. Just over a quarter felt it made no difference and 68% worse.
At that time BT was the most reported organisation for offshoring customer service calls, receiving twice as many mentions as any other organisation. Lets hope that in time the remaining 3,500 jobs (BT apparently had 5,500 agents working in India) are repatriated.
John Kemp
One of the blog posts that generated some chatter in our office this week were some survey findings from Neilsen – Consumers Trust Real Friends and Virtual Strangers the Most.
The opening statistic reports:
Ninety percent of consumers surveyed noted that they trust recommendations from people they know, while 70 percent trusted consumer opinions posted online.
The online “word of mouth” effect is a lot more prevalent than I expected, although thinking about the findings and my personal observation of the number of customer reviews on Amazon these days perhaps I shouldn’t be so surprised. Providing customer reviews of products is central to many online retailers’ sales strategy today – it isn’t just a “web 2.0″ feature – as customer reviews can help to generate trust and credibility (perhaps the feeling of a busy, vibrant shop full of customers, instead of tumbleweed blowing through the aisles of a warehouse out in the sticks… ).
That said, I don’t expect the range of answers to the question “Approximately how many people have you told about this problem” to increase by huge numbers. It’s long been known that the proportion of customers who post reviews versus those who don’t is about 90:9:1 (lurkers : one-off contributers : repeat-contributors) although perhaps the growth of twitter will change this?
BTW – if you can access BBC iplayer, comedian McMichael McIntyre considers who are these people who write a review about toasters? Settings 2 to 6 are an embarassment :-) (watch the Comedy Roadshow Brighton – 2 minutes in – you’ll see what I mean!)
From today’s Observer: The makings of a brand leader
Advertising, he says, has become even more hollow because of the way people now use the internet. “I Googled ‘best razor’, found out which one got the most stars in reviews, bought that one, and noticed that the one I had currently only has two stars. And it clicked for me; this is the way people are going to shop. For brands to rely on advertising and marketing is a huge mistake because people find you out, which is why we put emphasis on product and service innovation.”
As people become more internet savvy, I couldn’t agree more.
Dan Wardle
From a questionnaire I chanced upon the other day:
What is your attitude towards DIY?
A range of answers were offered from I enjoy DIY to I loathe it and pay someone to do every job no matter how small (or something like that). I wanted to add another answer option
- Ask me when I’ve finished my project
Ok, my answer is tongue in cheek but it highlights an important consideration when fielding a customer survey – when is the best time to ask the customer to complete a feedback survey?
Soon after the event is better. It doesn’t have to be the moment the support ticket is closed but give the customer a chance to recall the event as well as give yourself a chance to perform service recovery on any “customers at risk” that raise alarm bells in your results.
If you haven’t used an online survey for customer feedback in the last year avoid contacting customers that you last had contact with 12 months ago (even 6 months is a long time). Findings from 300 responses in the last quarter will be fresher, more reliable and more actionable than any findings from 1,000 responses in the last 12 months.
Dan Wardle
Might also be of interest: Tips for better customer surveys #3: Contact recent customers and (on blog.vovici.com) Follow-up/Transaction Survey
