Issue link: https://read.uberflip.com/i/1174618
what content marketers do and what results their activities get. As usual, most of our results show correlations in the data. The notable exception to this is the first discovery: content placement. For all other results, the sophisticated reader can feel free to adhere to the old maxim: correlation does not necessarily imply causation. Just because we see a correlation in the data, this does not mean that we have identified the cause of this cor- relation. However, as a first at- tempt, it is important to identify these correlations, because they may point at what the best prac- tices can be. Observing correla- tions is still a step above simply looking at what other content marketers are doing. Content creation and place- ment stands apart in this man- ner. Instead of simply looking at a correlation, we observe the dynamic response of items to being placed in a "stream." This is a causal impact analysis. We employed a fixed-effects regression analysis to see if the views of content pieces increased after placing them in additional streams. The re- sults show us that even in the same domain, by carefully organizing content, and mar- keting to relevant audiences, it is possible to increase the viewership of existing content. This is in contrast to the well- known SEO practice of hav- ing content in one copy only. We present evidence that it is good to have multiple copies of the same item, even on the same domain, to create differ- ent experiences for different audiences. Thankfully, instead of boring you with the details of this analysis, we are able to express the results with a helpful visual. We hope that you have found it interesting, relevant, and convincing. 81 APPENDIX LETTER FROM THE DATA SCIENCE LEAD