Why 5 years of coping behavior data to establish a trend?
Question: In last Friday’s MRx column, you mentioned that you need five years to be able to present the consumer behavior coping data as a time series so you can analyze for trends. We’re corporate strategy planning officers and in our annual business planning, we’ve used just two years to talk about likely trends for the following third year.
We’ve attended your consumer coping survey presentations since it became nation-wide in 2008. Every year that you present, your data always were just for two years. Then when you spoke of how things are likely to be on following year, you just referred to how the two years’ data showed an upward or downward movement. Isn’t that trending on just two years?
That actually makes your claim for “five years” to establish a trend inconsistent with your previous years’ presentations. Please clarify. Why do we need five years of consumer coping behavior data to be able to establish a trend?
Answer: It actually depends on the direction, and change in direction, of the data movements. These two factors define four possible trends. If the movement is year-to-year, the four possible trends would be:
— A year-to-year upward trend;
— A year-to-year downward trend;
— A year-to-year flat unchanging trend; and
— An upward, then downward, then upward, and then downward year-to-year trend.
As you can imagine and see, it’s the fourth series of movements that requires a 5-year data set. The first upward movement has 2 years. The next downward data connects the second year data with the third year. Then the next upward movement of data links the third year data with the fourth year. Finally, the second downward data connects the fourth year data with the fifth year data. And so, the 5 years.
Now, consider the case of changes every 2 years in the fourth category of movements.
If you work out the up-and-down sequence, you’ll come to a total of 9 years of data series. As most of us who deal with time series data know, the reality of year-to-year data movements are not as neat as in our 4 categories. They are a hard-to-predict mix of 2 or 3 of the 4, and can vary as well in terms of time length, like year-to-half-year-to-every-2-years, etc.
But the longer the time series, the more likely it will reveal the real underlying trend.
As I mentioned last Friday, this year’s presentation marks the fourth year that the coping survey is nationwide. But even just for the 4 years, the survey data already show interesting trends for certain product categories.
Consider, for example, the case of pandesal in Metro Manila. From 2008 to 2012, housewives considered pandesal a nice-to-have product category. To them, these were not a necessary recurring expenditure item but they were just nice to have around.
But from 2012 to 2013, Metro Manila housewives began seeing pandesal as an expenditure product category they generally felt they could not do without. Pandesal has become a necessity, although not absolutely.
Then from 2013 to 2014, Metro Manila housewives started feeling that they definitely cannot do without pandesal. It has become absolutely necessary. How did that happen?
The 4-year trended coping survey data for Metro Manila housewives showed that from 2012 to 2013, and then from 2013 to 2014, the “maintainer segment” significantly increased. In other words, housewives in Metro Manila who had pandesal in their budget rose in number for two consecutive periods.
At the same time, those housewives who were dropping pandesal in their budget moved in the opposite direction during those 2 time periods. Those giving up on pandesal dropped just as steeply as those maintaining pandesal increased over the same 2 periods.
As my 2010 Market Segmenting book said: “The ultimate source of growing your business are (behavioral) market segments. Products are only secondary.”
How much exactly were those increases and decreases will be specified once the coping survey results are presented industry-wide on May 8.
I’ve been asked in previous years’ presentations how could one manage, for example, a pandesal-like category to move from near-staple to a staple. I’ve been honest in saying that the quantitative data of this survey won’t be able to give the answer. But doing qualitative research on the housewives’ behavior change will provide the answer. What I’ve done for this is to sample and interview via an FGD (focus group discussion) or an IDI (in-depth interview) those survey respondents who, for example, raised their maintainer or retention behavior and/or lowered their lapsing behavior. The FGD or the IDI probed into the reason or reasons for this behavior change.
These insights uncovered the motivation behind the behavior change.
You did not indicate the product category or categories where your brand belongs, but if your categories are recurring expenditure items, the 151 product and service categories of the survey should likely have covered them.
My May 8th presentation won’t cover all 151 categories, but only the top 3 of each of the 5 category classifications of:
— Staple or categories consumers definitely cannot do without;
— Near-staple or those that consumers generally cannot do without;
— Nice-to-have or those that consumers regard as no longer necessary but still nice to have;
— Generally dispensable or something the housewife regards as something she most likely can do without; and
— Definitely dispensable or something the housewife feels she certainly can do without.
For those categories below the top 3 in classification that I won’t be able to cover in the presentation, the supporting survey data will be provided. If your categories belong here, you can analyze those data by copying the analysis I will do for the top 3 categories per classification.
There are product categories whose movements differ from one study area to the next (NCR, Balance Luzon, Visayas and Mindanao) and by SEC (Class AB, Class C, Class D and Class E) within each study area. These tell you how business growing opportunities differ from one area to another, and from one SEC segment to the next.
This is how rich the survey data are and can be made even richer with the correct and appropriate analysis.
Keep your questions coming. Send them to me at [email protected].
