Working Diary 04 - Looking back, what should we do next?

Every week during our internship, we met to review the past week's marketing effectiveness and make adjustments to the pace of future placements. As such, recording and analysing data in the back office is an ongoing daily task for me.

(To ensure the confidentiality of the company's data, I have cropped some items, as well as kept the original Chinese text.)

We look at a total of two platforms, one for the marketing platform and one for the online shopping platform. The data that we want to analyse and use as guidance for future marketing is a combination of both platforms, which is very much a test of multi-dimensional data analysis skills. For example, when we analyse whether the structure of a marketing blog post makes sense and whether it stimulates the consumer desire of the viewer, we do so through two dimensions: the first is the click rate of plug-ins on the marketing platform, through which we can know the extent to which our products are exposed as plug-ins with shopping links in different blog posts and attract viewers; the second dimension is the click count of product detail page links on the online shopping platform, which is particularly useful for most blog posts do cover more than one product, so by comparing the link clicks with the plug-in clicks we can see how much a particular product attracts readers to click on it in that post to see the details. By doing a case study of the more optimistic blogs and analysing elements such as storytelling, article structure, and pitching language, we can see what kind of articles are more appropriate in a marketing context.

I also learnt that a key point, a potential dimension, is 'time', which is also an important factor in data analysis. Data is limited by time and marketing planning needs to change for different periods: for example, everyday marketing needs to emphasise products' helpfulness in everyday life, shopping festivals need to emphasise 'short time limits and great deals', holidays like Valentine's Day or Mother's Day, this like, need to emphasise the fit in scenario, and so on. Also, multidimensional analysis cannot be carried out for only a few days, but must be based on macro analysis of data over a long period to find regularities.