weeks of hard work had been put in this project - finding content, writing content, handcoding the emails...all in a effort to customize messaging to twenty-two segments, and two brands. yes you read that right, a total of forty-four emails.
the emails look good, they work on principle and are loaded in the email tool, so what's the problem? the technology is too good.
here's what i mean by that. we use a great tool, and we integrate that with salesforce like many other companies, to create a great nurture campaign series. we lead score on the back end, and we also assign different internal contacts to "speak" to the prospect. but all of that happens too quickly!
we started to test - the team went to the web, filled out the web form, and got them customized email and messaging. the first round of testing was great! then we had some members of the team fill out the form again - some members deleted themselves out of the database before filling the form again, others didn't. those that were "fresh" prospects worked as expected. those who were repeat prospects had a different experience. the subject line was correct, the tokenized content was correct, but the dynamic customized content was the same as what was originally sent to them.
so what happened? when the first email was sent the processes looked like this:
- the web form is submitted
- the tool receives the submitted information
- the database is updated with the new information
- the tool builds and sends the email
- the web form is submitted
- the tool receives the submitted information
- the tool builds and sends the email
- the tool updates the database
so, how do you deal with this new opportunity? there are a few different options that you can try. first off, you need to know that to discover an issue like this you need to do multiple tests and multiple q-a sessions. try to break the system. and if you discover an issue like we have, turn it into a marketing opportunity. possible solutions we are looking at:
- changing the timing of the email send - if we put a wait step in the email logic, then the tools, scoring, and back-end processes will catch up to each other...this of course will have an effect on the user experience since receipt of the email will no longer be instantaneous.
- look at all the possibilities - is tokenized content or dynamic content that is causing the problem, and can it be corrected by swapping out either
- think outside the box and look for ways to accomplish the additional requests for information in a unique way
the overall moral of this posting though is to test, test, and test again. look for ways to not just accomplish what you are looking for - in this case delivery of a custom email based on an attribute - but for ways to break the system. think like your users... if they don't get the email in a few seconds after submitting they'll resubmit, especially if they want the information; this could break the logic. they'll submit a different attribute to get additional information, shortly after the first submission, this could break the logic. so in other words, don't just test - break the system.
0 comments:
Post a Comment