Promoting Trial in Packaged Goods

One of the best methods of using database marketing in the field of package goods lies in promoting trial of your product to heavy users of competitive brands. This is hard to do through FSIs because the vast majority are redeemed by existing brand users. It is possible to obtain through outside sources, lists of high category users of completing brands. These names can be added to a consumer database, properly coded.

To do this type of analysis you use of a decision tree similar to the following in the creation of promotions to these names.

To “aggressively steal share” the marketer should offer an extremely valuable coupon, amounting in some cases to a free large size product. You will notice that loyal users receive no offer at all. The frequent coupon users and the light category users receive lower value coupons. This clever targeting method makes optimum use of the marketer’s dollar.

To deliver the coupons, there must be a selective insertion process which does not treat every household in the same way. If done correctly, it is possible to put these coupons in the mail to targeted households at a cost effective rate, and to measure response with some accuracy. Targeting households in general, or category users, may not be cost effective for the marketer. Targeting high competitive brand users, on the other hand, may be extremely cost effective, and should be explored.

How do the economics of such targeting work?

In the first place, the goal of the program is conversion of incremental purchases — getting heavy users of competitive brands to switch some or all of their purchases to your brand, and getting the less-loyal occasional users of your brand to increase the percentage of times they buy from you.

To determine what value to assign to the offer, the marketer can use a formula which calculates the conversion rate they will need to achieve if the promotion is going to payout in 12 months.If the conversion rated needed is absurdly high or low in their experience, then the decisions concerning “if” and “how” to target the groups in question are relatively clear. If the conversion rates are in the grey zone (maybe we can achieve them, maybe not) then the client can adjust the design by altering the definition of the target group or changing the value of the offer to see if a more achievable conversion rate can be produced.

Once the buying and coupon using habits of each household are known, the value of each coupon necessary to assure redemption by a desired percentage of households can be determined with scientific accuracy. These rates can run from 5% to more than 50%, depending on coupon value, and targeting of households. As an example of how this works, picture four possible target groups which may be targeted using a decision tree:

Looking at these four groups, the marketer could decide on the value of the coupons to send to each group. The percent redeemed is simply a function of the type of household targeted and the value of the coupon. This can be tested and adjusted, and therefore, predicted with some accuracy.

The variable costs derived from this chart are used as input to another chart which computes the conversion rate required to pay back the variable promotion costs in one year.

It is also possible that the marketer can explore use of household scanner data. Household scanner data (data from households that show their shopping card at checkout time) suffers from several serious limitations: not everyone why buys has a card, and not everyone who has a card uses it every time they shop. People shop at more than one store, which means you cannot be sure that the data you have is all the important data on the household. Finally, there is the volume problem.

The amount of data generated at the cash registers in a supermarket is so gigantic that even the largest chains find it almost impossible to keep track of household purchases by SKU (product and brand). The volume chokes even the largest computers. They tend to accumulate information by department (meat, grocery) by month, rather than keeping track of every tube of toothpaste bought by each household. The problem is that the tube of toothpaste data is precisely what the manufacturers want to know.

Use of household scanner data is becoming more common as the public becomes more aware of the value of such shopping cards, and as retailers and manufacturers provide more benefits to those who use them. As household scanner data increases in accuracy (if it does), however, it may offer real promise. Knowing what each household actually buys every week certainly has the potential, one day, to beat asking them what they remember buying.

Post Footer automatically generated by Add Post Footer Plugin for wordpress.