When Privacy Policies Go Bad and How to Fix Them

Hot on the heels of yesterday’s post about the importance of carefully selecting your positive and negative keywords, I’ve some remarkable online test insights to share from ContentVerve.com’s Michael Aargaard.

Michael wanted to find out what impact adding a privacy statement alongside a registration form would have on sign-up rates. I think you’ll agree, the results are far from obvious!

Using a registration form on a high traffic online betting site, he started by adding a simple statement “100% privacy – we will never spam you!” just above the Sign Up link. This simple change resulted in an 18.7% drop in sign ups compared to a control cell:

That’s a huge drop in sign ups through the addition of statement that is regularly used to increase conversion!

Suspecting the ‘spam’ word to be a possible cause of this drop, Michael ran another test, this time with a privacy statement that read “100% privacy. We keep all your personal information secret”:

This time there was no significant difference between the two cells. The privacy statement was neither helping nor impairing conversion to sign up.

Time for another test, this time with a subtly more reassuring statement “We guarantee 100% privacy. Your information will not be shared.” The results were remarkable:

That’s almost 20% more sign-ups from a simple, reassuring promise. Michael’s hypothesis is that the word ‘guarantee’ serves as a much more credible and authoritative commitment to fair play than the previous statements.

There was just time for one more test, this time reintroducing the ‘spam’ word to see if it would have the same negative impact as seen previously:

Incredibly, all of the good provided by the ‘guarantee’ word appears to have been undone by the reintroduction of that toxic ’spam’ word.

I’m indebted to Michael for sharing his test findings with us. But what can we learn from this exercise?

Firstly, as I’ve reported before, the little changes sometimes make all the difference. Without careful testing, the impact of inadvertently giving visitors reasons to be fearful through a mention of ‘spam’ might never have been spotted. And neither would we have been able to quantify the powerful reassurance felt by offering a ‘guarantee’.

As a result of this testing, we now know that for this particular website the word ‘spam’ (and all related synonyms) should be treated as strict negatives in future. As will be any other terms that could fuel alarm amongst potential customers. And more work can now be done to verify the reassuring power of a ‘guarantee’ and further testing carried out to see where else this keyword can be used to good effect across the site and on other marketing materials.

Remember, your mileage may vary. You should never treat the findings from one website as certain proof for your own site. But you can take heart from these results and embark on your own testing programme to establish if your site visitors will react the same way.

Sometimes the most obvious things in life are far from self-evident. The inquisitive marketer can learn a great deal from Michael’s methodical testing process. Perhaps your website will be the next proving ground for some remarkable marketing insights.

Further reading:
ContentVerve.com’s Testing Summary: http://contentverve.com/sign-up-privacy-policy-tests/