When Sampling Works

Gone are the massive national sampling programs where twenty percent of households were targeted and tens of millions of samples were handed out. Many brand managers have either eliminated sampling as a marketing tool or greatly reduced its scope as a driver of trial. The prevailing opinion in the marketplace is that sampling rarely pays out: it is an expensive marketing tool and there seem to be more efficient marketing levers that will generate trial. Sampling activity also seems to have been reduced in part because it has been difficult to measure the incremental effectiveness of these programs.

But is there a set of optimal conditions where sampling is likely to be not only effective but also efficient in driving trial and volume? Is sampling a good strategy only for those situations where the product is high quality, but price is a barrier to trial? Or are there other conditions under which sampling programs should be used? According to recent research, there is good evidence that sampling works.

Ipsos-Novaction has conducted over 8,000 brand tests and evaluated over 50,000 brands. Analysis of this database has yielded some interesting and not necessarily intuitive findings on the impact of sampling. Our analysis allowed us to examine the incremental effectiveness of sampling because we have analyzed the trial potential of an item that might be due to all other factors (advertising, price, distribution, etc.)

Price & Quality

Sampling effectiveness is very sensitive to product quality. In Table 1 , we see that sampling effectiveness increases by 50% above the average effectiveness of these programs when the product quality is perceived to be high. And similarly, when product quality is low, as measured against our database, sampling is only 60% as effective as the average sampling program.

However, sampling effectiveness has an inverse relationship to price. We found that sampling programs were roughly 11% more effective for relatively low priced items than they were for high priced items.

Table 1

We also considered the impact of price and quality on sampling effectiveness ( Table 2 ). Sampling is very effective for high quality products, regardless of price. And sampling does not work well for products where quality is low, also regardless of price. The conclusion is that sampling effectiveness is much more sensitive to quality perceptions than to price.

Table 2

Ad Driven Trial Potential

Initially, we found that there was almost no relationship between ad driven trial potential and sampling effectiveness. Sampling programs were of average effectiveness whether ad driven trial potential is low or high. However, when we analyzed the relationship between quality after use (a key driver of repeat purchases) and ad driven trial potential ( Table 2 ), we found that there were conditions where sampling programs could be very effective.

Table 3

In situations where quality after use is high, sampling programs can be very effective at all levels of ad driven trial potential. However, when perceived quality is low, sampling programs are clearly not effective. And sampling effectiveness declines as ad driven trial potential increases for low quality products. This is possibly due to the relatively high trial that would be generated in the absence of sampling, so the incremental volume attributed to sampling would be lower.

Conclusions

1. When quality after use is high, effective sampling (i.e., right size and delivery) must be considered. In addition sampling is especially effective at high price and when ad driven trial potential is inhibited (copy, pack visibility, price).
2. As long as there is a proper value balance, sampling can generate significant incremental volume for even moderate to low quality products.
3. By contrast, sampling should not be considered if ad driven trial is high and quality or value are low.

The study results do not support a return to large national sampling programs, but indicate that sampling as a marketing tool should not be ruled out. There are conditions when sampling can be an efficient driver of volume. Clearly, the more efficiently potential users of the product can be identified, the more cost effective these programs can be. With the security of a payout analysis, you can ensure your sampling program will drive trial and volume effectively and efficiently.