I sat across from a client the other day discussing how they measure digital marketing efforts. They had just committed and entrusted millions of dollars in online ad spend to our agency only weeks before. The SEM and Display Media teams had already come in to discuss ideas and strategy. Now it was my team, web analytics, to come in to measure and prove that we could get the most conversion out of these ad dollars. On the line – display dollars and social media.
As marketers, we know that display is the hidden lift behind search and conversion. We even realize that social media, beyond the hype, has as much if not more value than the a creative placement. It is intuitive. And, to be honest, executives get it as well – they just don’t know why they have to pay so much for it. Yet, I still get the questions I got the other day, “How do we defend our display budget? Is there really ROI in social media?”
Here’s the thing, if as online marketers we are in doubt, a definitive ‘NO’ is going to come down from above. Research conducted by such reputable firms as ComScore Networks and eMarketer isn’t enough to change minds. It wasn’t enough to convince the marketers across the table from me. They wanted to know if display and social media spend was working to the advantage of our other clients. If so, how did we know?
The current approach has been pick a few tactics, launch, and measure. Results come in and they are lack luster; we chalk this up to not working and move on. Why the test didn’t work is not always assessed. The results tell it all. Not really. The results only tell us the outcome of the test. It does not tell us the validity of the test. This is the beginning of defining attribution to our display and social media tactics.
The Reality:
Not all campaigns and applications are made equal. The trick with display and social media tactics is that you know what you want them to do for you. What is the goal, awareness, drive to site, engagement, or conversion, or all four? Once that is determined, what is the role of display and social media in attaining these goals – direct or supportive?
The Application:
It pays at this point to think out of the box. Social media as commonly thought of – Facebook, Twitter, LinkedIn, etc. – is only an aspect of social media. What makes social media work is the communication, proliferation, engagement, and connection it drives. The venue of a network or blog is only a placement, the components that facilitate the experience of the venue are the engines. Display as well is more than a billboard. Approaching display like you would an email campaign with a targeting strategy, crisp copy and creative, and a strong call to action is the key to making display work.
Next, taking into account how these tactics support the goals will define measures and metrics. Saying that click through rate for display and social media are indicators of conversion is only a small portion of the value. As seen, click through rates are dropping dramatically for display and marketers have yet to directly connect social media interactions with conversion. It is not to say that conversions can’t or won’t happen. It is just that the likelihood is much lower. So, you need to measure how awareness and perception contribute to conversion. And this is the crux of the matter.
Test, Test, Test Again:
The test at this point should be more clear. You know what you will launch and why. You have a perspective of how to measure performance and its link to goals. It is time to develop the test plan. Here are a list of things to keep in mind when developing the test plan:
1) Know your baseline. This is more than having a control. A control segment assumes you have tested a baseline. Don’t assume that several weeks prior or even a few months of data is representative. You will need to measure a baseline of performance over a period that allows for seasonality, marketing cycles, and market forces.
2) Consider how long your display and social media campaigns need to be in market to be able to measure impact. You need to attain a threshold of measurable sample and you need to take into account the length of time required for exposure. You may reach a sample that is significant but if it was attained in a week and a display or social media standard is 3-4 weeks in market for impact to be felt, you don’t have a viable test.
3) Start simple and progress to advanced. Applying a simple in market – out of market approach can get you headed in the right direction and give you what you need. As you begin to exhaust simple testing methods, this is when attribution analysis can kick-in and allow you to apply complex strategies.
4) Got results? Test again. This is an iterative process.
5) Those results you got, leverage them elsewhere. Insights you gain from one set of tactics or a campaign can seed an expansion of activities. Then, test assumptions in the new application.
