Imagine that this client's email arrives in your inbox. "I read that it's a best practice to include your company's brand name in text ads. Do I need to include the brand name in every ad copy?" Now that we already have performance data on text ad variations, we can start calculating numbers. CTR chart with balloon "Wonderful victory" Answer "yes". "Ads that mention a brand are nearly three times more likely to be clicked. We'll replace them with new branded text ads by the weekend." advertisement Continue reading below Five minutes have passed when the client replies: "Interesting.
And do you think you controlled whether the ghost mannequin effect service brand was a search keyword?" "good…" In a hurry, this time consider whether the search query is branded and calculate the number again. Your new conclusions are inconsistent with what you have just provided. Clickthrough rate graphs for branded and non-branded ad copies "For search queries that don't include a brand, ads without a brand name are 50% more likely to actually be clicked." advertisement Continue reading below What happened here? You are proud to be a data-driven, quantitative marketer. But now you have to explain to the client why your analysis was wrong until you reaffirmed your process. So what was the cause of the error? Marketers wear many hats and can make mistakes if they are in a hurry. But the real problem is deeper. Most digital marketers are not trained statisticians, analysts, or data scientists. Collectively, we're working to improve account performance, so we're hacking the path to data literacy across the industry.
In an ideal world, every marketer would be a statistician. But if you don't have the bandwidth to win a second carrier, here's how to avoid the overwhelming pain of your data and find valuable insights in your PPC data. You've probably built a paid search career based on finding wins and opportunities, rather than disagreeing with your own conclusions. Yes, it's more fun to have a party for the growth we think we're driving. However, this actually leads to lazy thinking that can reduce performance. Real statisticians are trained to be skeptical of simple conclusions and observations. You may have trouble drilling holes in your analysis (especially at first). But that's very bad: After committing your ideas to a PDF report,