Starting a new ad campaign takes a lot of work. There are creatives to build, and audiences to target. There are meetings to have and objectives to meet.
And after weeks or months of work by sometimes dozens of different professionals on your team, you launch your campaign and eagerly await the sales to start rolling in.
Perhaps at this moment, you remember to tell your analytics team about your new campaign. Or perhaps, you get a message from your analytics head, “Hey, boss, did we launch a new campaign or something?”
Does this sound familiar to you?
Or how about this quote from a recent AdExchanger news update:
One startup manufacturer decided to reduce Facebook advertising “for the sake of their business and their own sanity,” instead testing a 15-second TV spot. Returns were harder to prove without the data feedback, says one executive, but at least “[y]ou can push the button and get on with your life.”
Measurement is often a task thought of after campaign creation, if at all.
First, a story about hiring: In 2005, a group of researchers at Stanford conducted a research study in which they asked participants to read the resumes of two qualified candidates for the job of police chief.
One applicant was street-wise and got along with fellow officers but had only a basic education and lacked administrative skills. The other applicant was highly educated and well-experienced in administrative issues, but lacked the street experience and rapport with other officers.
Participants were then asked to rank which were the most important qualities for the job. Oh and there was one more factor. The streetwise candidate was named Michael. The educated candidate, Michelle.
Participants overwhelmingly rated the important factors for the job to be “tough, risk taking, physically fit, and getting along well with fellow officers” while “well educated, has administrative skills, has political connections, able to communicate with media, family oriented” were all listed at less important.
However, when the researchers reversed the names on the resumes, making Michelle the streetwise officer and Michael the educated administrator, the factors that were considered important ALSO switched. Without consciously realizing it, the participants used the factors to justify their internal beliefs.
So why am I telling you about this study on gender bias in this article about measurement? Because it’s not really about gender, it’s about honest data.
This is how many people treat their analytics! Instead of deciding beforehand which factors are the most important to track (or better yet, looking at the careers of successful police chiefs to see what skills they may have had in common), many marketers impose their own internal interpretations onto the data.
Which test is better? Which channel provides you the most ROAS? Which is your most popular product? Which online interaction is the most critical for closing an offline sale? Which candidate would be a better police chief?
Without building your measurement into your media campaigns right from the beginning, you are missing out on vital performance data. The factors that you hone in on will be influenced by your own bias.
CMO and marketing leaders must hold their data to the highest standard. Start by insisting on honest data, gathered and analyzed for what it is, not what you want it to be. Crush your inner sense that you already know the answer and make the data prove itself. From there, you’ll know that you are working from the information that reflects your real needs and strength.