There are many ‘types’ of donor surveys, but this particular survey type, the supporter survey is the biggest source of targeted bequest leads. They also help boost future direct mail income dramatically and they are a fantastic major donor lead generator.
Communications and fundraising author Jeff Brooks said ‘If you are not doing a supporter survey like this, you cannot be donor focused, and are missing out on millions of future potential”.
Sean Triner from Moceanic, has been a pioneer in this area and has revolutionised the supporter survey in Australia. If you would like more on this subject then Click Here to check out his work
Once your survey is ready for designing there are a few things you can do to increase the accuracy of the data captured. In the coming issues we will provide some useful tips, which will improve the efficiency and accuracy of your data capture process.
In this issue, we are going to cover the effects of colour in your survey design.
A thing or two about colour
When designing your supporter survey, it’s important to consider when and where to use colour. Colour can be your worst enemy or it can be your best friend.
Colour and the type of colour in the correct area can help provide assistance to, and help the flow for the person filling out the survey.
There are 4 pillars which are required to execute a successful campaign. Fundraising team’s strategy, Analytics, survey design, data capture
Only 3 are required to get the survey out of the door, but the fourth, data capture, is just as important, but quite often is the last to be considered. Because there is a misconception that data capture is easy.
“Data capture is simple, but not easy” David Packenas – Crystal Clear Data
What’s all this got to do with colour?
Scanning is an excellent way of automating data capture, but a brilliant graphic designer may produce a survey which is a disaster when comes to scanning,
Once surveys are scanned, an image of the survey is created in black and white. Black and white images are easier to read electronically, and the extracted data is far more accurate.
During the scanning processes, certain colours can be dropped out from the image, pastel colours will drop out automatically when using a dedicated document scanner. The problems can begin when colours do not drop out, or sometimes worse, partially drop out.
OMR – A crash course
OMR stands for Optical Mark Recognition. OMR fields are used to read tick boxes. Typically, a scale of 0% to 100% is used to determine if a box is ticked. Generally speaking, if a box is 80% or more filled, it would be safe to say the box was ticked. There are lots of caveats here, which will be covered another time.
However, an example of one of these caveats could be if it’s more than 95% filled then send to the data verifier for a decision, if it’s 95% or more filled, it might not be ticked. It might be crossed out, or it may have a background colour which is interfering with the reading.
So you now have a basic understanding of how an OMR field works, anything that interferes with this calculation can have a dramatic impact on the accuracy.
Scanners – A crash course
It’s important that your surveys are scanned on a dedicated document scanner, these scanners come with intelligence, double feed detection, colour drop out functions, which are missing on scanners that are built in with your printers, otherwise known as Multi Function Printers (MFP)
If your document has colour which is designed to drop out, your MFP becomes a lot more unpredictable, colour can be scanned in as a grey scale, which can then interfere with document recognition and percentage filled on your tick boxes.
Colour borders should be avoided, especially greens and yellows as they tend to drop out, which in some cases, this is not a problem, it’s where the colour is in that mid range, where it doesn’t quite drop out or doesn’t completely fill in.
When calculating the percentage filled of a tick box, the border is taken into consideration, if it’s a nice solid border, then the percentage of the area can be subtracted from the overall read.
The Green borders below are in that mid-range where the borders partially drop out, which will cause a big problem and can just about make it impossible to get an accurate reading of percentage filled.
This next image shows, how after scanning the box becomes problematic as it breaks up
There is no mistaking this next example, the borders are clean and sharp. The base level of percentage filled will be accurate within fractions of a %
We are starting to hear more and more, that these colours cannot be changed or subdued due to branding. I’m not sure how to respond to these sorts of objections, it’s a survey, a data collection tool, where data accuracy has more impact than the subtle change in design we are suggesting here and will have no impact on branding at all.
Below is a good use of shading, it helps to separate the questions, and the pastel shading will drop out leading to more accurate results.
The shading in this next example is too dark, and the shading does not easily drop out, the resultant image in is a dirty image, additional settings may help improve this situation by removing image noise, but this problem could have been avoided at design time.
This is just one of many little aspects, that turns your surveys into raw measurable data, there will be many other tips coming in 2018 so keep an eye on our blog. Or if you have any question, don’t hesitate to contact us here.
PS – Moceanic’s step-by-step course on surveys is designed to help fundraisers experienced with, or new to, this kind of survey.
I’ve negotiated a special discount for CCD blog subscribers – use the exclusive code “CCDLOVE” and follow this link to get US$100 off the course, bringing it down to US$297 (about AU$370 today).
There is nothing to lose – they guarantee you will learn something new on their course AND if you enroll by 18th February I will give you your enrollment fee back when you choose CCD to help you get that valuable data entered!