Q: how do we do this with competitive data analysis, high page arrival, good billing, high price conversion? In dealing with such issues, i hear the quadrant method is the most common. What's good? What about the four quadrants
A: the four-perimeter analysis method, as shown in the figure below. Extension data are analysed from four quadrants (levels) to learn which areas are deficient and perform well, and then targeted or drugged。
As to how it should be read:
In the quadrant method, cross-references represent our consumption of advertising costs, and even if the coordinates represent the amount of conversion generated by advertising, one by one we see:
First quadrant: there are three attribute labels, i. E. High conversion, high consumption and high cost. There are many factors that need to be optimized for this type of data, with a focus on bids, hits/transformation rates (needs to analyse conversion rates under the o and cpa models) and the precision of textual creativity。
Second quadrant: this part is the best advertising data, which translates into high-cost, continuous observation and, of course, appropriate scaling up of rich masters。
The third is: it's obvious that this part of the ad failed to go out and that the exposure level is too small, so we need to give priority to presentation when we optimize. I don't think you have to tell me how to optimize the performance
Fourth quadrant: in the final part of the data, the amount of conversion is low and the cost of conversion is too high, and our preference is to optimize the delivery options (including creative/landing pages/direction, etc.) and the accuracy of the flow。
(chu added: the quadrant method is a basic data analysis method, but it is not always appropriate for your promotion, mainly because the amount of conversion generated by advertising is not well defined or not well defined. Practice is therefore of paramount importance to learners. Do not be too committed to this theoretical approach in order to avoid going to extremes, without analysing the right results, leading to errors in the direction in which the competition spreads, or to a circle that does not emerge
