More on the Google AdWords Experiment

In a comment to this post, Frank Johnson wonders how we’re going to measure the effectiveness of our Google AdWords campaign. So Frank, here’s your answer …

Yes, Church of the Resurrection is a large church with over 14,000 members including adults and children. And you're quite right that this makes it difficult to measure ad campaign effectiveness.

But I'm in a quite unique position. While I'm the IT Director at Resurrection, I'm also the volunteer tech guy for my wife's brand new church plant, Living Water Christian Church. Like your friend Scott's church, it too is small enough that we can find out easily what's working and what isn't. This turns out to be a great advantage from a technical perspective because I can prototype things on a very small scale at Living Water, before moving them up to beta in the big church, and then finally rolling them out in the big church. So that is our plan in this case.

For this experiment, I’m working with Chuck Russell who is also a contributor to this blog. He has experience running Google AdWords campaigns at the national, denominational level for the United Methodist Church.

We’re presently wrestling with two questions: 1) What is the objective of the campaign? 2) How will we measure the results? Now clearly these questions relate strongly to each other and to the design of the campaign.

For example, if we define success as someone hearing about us first through AdWords and subsequently attending worship, that creates some significant measurement obstacles because the Communications folks at Resurrection are reluctant to put “How did you hear about us?” on the worship sign-in sheet. (Jokingly, our Communication Director has wondered out loud about some kind of coupon strategy, a marketer’s standard technique in a situation like this, but that’s good only for a laugh.) Another difficulty with worship attendance as an objective is that it is often difficult to separate multiple influences on why someone came to worship for the first time. Someone may well have heard about us first through AdWords, but then heard about us in a number of other places before deciding to attend. By then, they’ve forgotten that AdWords was the first exposure.

So we’re considering other objectives, with other measurement methods, which will give us a way to determine effectiveness with greater confidence. For example, what if the objective is to get people to subscribe to something such as an e-mail newsletter or a podcast or a blog? Or, maybe the objective is to get them to download a PDF or watch a sermon online. Perhaps the objective will be to get them to send us a prayer request. The key here is to get people to take an action online that we can directly measure. It must be something beyond simply clicking on the AdWord itself because we’ll never know if that click was relevant or effective. We’re also planning on having a separate landing page on our website for each AdWord so we’ll be able to measure any relative differences among the AdWords.

As you can see, right now we have more questions than answers. My previous post was just to let you know we were working on this and would have results to report by sometime in November.