If you are sending emails to your customers to ask them to take part in a survey, then you are performing a form of email marketing. This means that all the rules (whether law or best practice) still need to be adhered to.
When you prepare a list of customers’ emails to use for your next customer survey, take a moment to consider who you are inviting, especially if this survey is an annual survey. By that I don’t mean cherry pick those customers you think will give you good answers! Of course, you want your sample to be representative of your customer-base, but find out how long ago did you last communicate with these customers? Or when did their last transaction or customer contact actually take place?
You don’t need 20,000 responses to a survey
300 survey responses gives a high level of confidence in the data for populations that range from a few hundred to tens of thousands.
By population I mean the size of the customer base – the number of people that could be asked to take part in the survey.
If you want to analyse results for two segments of customers, for example repeat customers and new customers, then you want to have confidence in data for both segments (so 600 responses). You might also plan to drill-down into the numbers for other demographics (e.g. customers who contacted the call centre), so include this in your decision on who to contact too, but you don’t need to collect 20,000 responses to get robust data from your customers.
We also find that the response rate for customer surveys improves considerably where the customer contacts or transactions are recent (within the last 3 months), which helps to ensure the results are current and relevant too. Very important if you plan to act on the findings!
If customers aren’t opted in to a regular newsletter or email from your organisation (in other words the last email you sent was about their last transaction, perhaps 12 months ago) then 10-20% of older emails could well be “bad addresses” (bounces).
The frequency of customer interaction should be considered when selecting the sample to email. Car sales typically involve contacts once a year (returning to the garage for a service), membership of a DVD club potentially involve many more contacts.
Test a mailing on a small sample first
When your survey is ready to launch, send email invites to a small sample of your selected customers first. Four or five hundred emails are usually enough to collect at least a dozen responses within an hour of hitting ‘Send’ (50 – 100 responses overnight) and review how many emails bounce.
Make sure all bounced emails are removed from your email database, but analyse which emails bounced and which are clicking through to get a better idea of what should lead to a better response rate. Then refine your email list.
The worst that can happen?
One of the issues we face when sending survey emails on a Client’s behalf is that inevitably for some customers something went wrong and they had a bad experience – so the last thing they want to do is complete a survey. Are those customers more inclined to hit ‘Report Spam’? Our spam complaint ratios have been 4-12 times higher than a typical complaint ratio (1 per 3000-10000 emails sent). It really is in your best interest to minimise complaints to help ensure future emails are delivered to the inbox.
In an article entitled Email marketing: the worst that can happen the author Mark Brownlow points out that spam complaints can have a major impact on your ability to get emails delivered. Insisting on emailing all 200,000 contacts from the last year to measure customer satisfaction could have repurcussions elsewhere in the business… If overnight the survey mailing causes a spike in the email volumes seen by the ISPs with an increase in spam complaints and bounced emails, then ISPs may start holding up delivery of your regular customer newsletter and/or other communications.
Clearly, that’s not desired, but there are a lot of organisations who collect customer feedback and don’t realise the connection between an ad-hoc survey email and email marketing. It’s almost a case of actively conserving resources (your customers’ emails).
Read more
The always excellent No man is an iland newsletter by Mark Brownlow gives lots of practical advice and information about email marketing, pointing to real world examples and experiences. Another recent article Targeted opt-out email: busting some myths is also worth reading.
An older (still relevant) help article about sending emails for customer surveys.
Nice blog, very useful tips and experiences.
Thanks for the feedback Arantza.