This week we’ll look at how terms that were trending in late June are faring today. By June, most U.S. states had reopened. Black Lives Matters protests were growing worldwide, and retail sales grew more than expected.
“Local marketing” intent among brands has been steadily increasing since mid-September. In June, both “demographics” and “local marketing” saw increases. Marketers may be seeing location as more efficient targeting criteria than other audience characteristics, especially as states vary drastically in their COVID response and number of infections. Zeroing in on areas where business has been sustained or is on the upswing may allow marketers to tap into audiences that are eager (or less reluctant) to buy.
Similarly, “content marketing” intent has dipped in the past month among the MediaPost 500 brands. Declining intent around content terms correlates with a recent increase in intent around search terms such as “pay-per-click,” signaling a focus on marketing tactics that need fewer resources to execute.
Intent among brands for “content personalization” saw a steep climb in June, as measured by Bombora. While the term saw a few spikes in intent since, it’s been trending downward overall. Getting content personalization right requires investment in audience segmentation, mapping, and analysis, as well as content optimization tools. In June, we were in the earlier stages of brands attempting to redefine the demographics, sentiments, and behaviors of their audiences, which may explain this particular intent trend.
Michael, I don't quite get it.
The Y-Axis is labelled "% of MediaPost domains with interest spikes", and the article refers to the MediaPost 500.
The 'spike' is now just over 0.4. Does that mean that the basis is the nett sum of all the domains of the brands that constitute the MediaPost 500 brands (i.e. each brand can have more than one domain).
Achieving a 0.4% (up from virtually zero on Aug 02) could be seen as a 'spike' when looking at it in a relative way, or if you looked at at the big picture (that more than 99.5% saw no increased activity in local marketing) then it would be a 'blip'. I'd love to see the trend on the traditional 0-100% Y-Axis scale and see if I could spot the spike.
Thanks for the feedback. We are looking at a relatively small sample scale from week to week and accordingly, our commentary is simply meant to provide insights into interests and,from this, indicate potential intent to take an action. It is a hyper focused view into the challenges brands and agencies are being faced with every day – with the goal of bringing awareness to those areas where we might be able to anticipate needs. We focus on those topics with the with the highest % of interest spikes each week. Hope this clarifies. Michael
Thank you for the explanation Michael.
I take your point that it is a small sample. However, with small sample sizes it is way more prudent to NOT read 'spikes' into the data as the chance of a 'false positive' or a 'false negative' increases exponentially. In essence a single respondent could be the reason for the 'spike' ('blip') and one person does not make a trend. If there is replication of that 'spike' in numerous other respondents then statistical confidence increases. Awareness is important, but so is accuracy and replication,.