Imagine brand tracking this way: spending an hour on your balcony counting passing pedestrians in order to predict how many people pass by during the entire day. It is fair to say that your guess would probably be pretty inaccurate because one hour of observation does not provide a lot of information. However, if you repeated this exercise every day for a month, your prediction would be a lot better. Why? Because as humans, we learn over time. We start to notice patterns, for example, how traffic increases during rush hour and decreases when it is raining. See, the more information we collect, the more we are able to build a general sense of how different factors contribute to what we are trying to measure.
This thinking process comes naturally to humans. However, in the world of brand tracking, it is almost entirely missing from traditional survey-research methodologies. This means that if we used a traditional brand tracking tool to measure a KPI every day for a month, the methodology would treat each day as a completely new day, making a new prediction from scratch every day. So by the end of the month, the prediction would be no more accurate than the prediction at the beginning of the month. No prior information was taken into account so nothing has been learned over time to improve the prediction.
Starting from scratch every day means that the prediction is vulnerable to daily fluctuations. Consider the balcony example again. On the thirtieth day the one hour of observation happened during a rainstorm and no one walked by. Would you predict that no one walked down the street for the rest of the day, even during the sunny hours? No, because by the thirtieth day we would have a general sense that rain means fewer pedestrians. As a result, our prediction wouldn’t depend entirely on what we happened to see in our limited, rainy hour of observation. This is not good brand tracking.
Learn how you can use Latana Brand Analytics to improve your brand marketing and grow faster.BOOK AN INTRO →