Eclectic Thought Samples From Larry Burns

Continued Proof that Sampling Works

Share Button


In the era of “big data,” the notion of executing programs in a scan-based targeting and analytics frame has become available to the sampling market. What that means is that by utilizing behavioral data (purchases a consumer makes in a particular set of stores), we can understand how those purchases are tied to the individual. Then, we can aggregate that information and look at a how the sampling tactic impacted in the marketplace.

The way we marketers do this is by executing a sampling program in the real world. We have discovered a very interesting way to “look back” at naturally occurring experiments in data about programs we have been running. It boils down to simply that one group of people receives a sample, while another group does not. Then, we look at the difference between theses matched test and control groups to see whether or not the sampling program was impactful.

For example, we recently received some really remarkable results from a company that operates within the grocery channel – results that demonstrate the power of sampling to influence consumer-purchasing behavior. We did a beverage category study, and quite literally saw a product trial rate four times higher within the test group. In other words, four times as many people who received the sample actually went into the store and made a purchase than the number of people who didn’t receive the sample. We also saw a very high increase in actual unit sales within that sampled group in fact an almost 700% increase in measured purchasing by people who received the sample was present.





So, Sampling Works! It’s something that we’ve been talking about on this blog for quite some time. It’s getting ever more difficult to question these scan-based results or somehow dismiss samplings power because we’re reading actual purchasing behavior.

Interestingly, when we do this, each of the products sampled can of course be purchased in other retail channels and stores beyond the ones we are able to measure, so in effect we’re always looking at a conservative measurement of response and impact.

Yet, even using these conservative measurements, we see very exciting and measurable impacts, which certainly show that sampling programs, done well, can really pay out. Another test program that proved successful involved an over-the-counter remedy in the supplement arena. We conducted this program measurement in the grocery channel. As an old researcher I will admit to being somewhat concerned that we would see as a measurable result, since so much of the purchasing of this particular product occurs in drug or convenience store channels. So, we ran the program, and lo and behold, we saw an increase of more than 200% in unit sales from people who received the sample and those who did not – and slightly more than double the trial rate in just the grocery channel!





As I’ve been discussing for a very long time on this blog, sampling is a method of reaching consumers and giving them the option to try a product they’re interested in. By doing this, you give the person a chance to experience that product. What’s very interesting about that is that when they then go into the stores to make a purchase, they are not really making what we refer to as a “trial purchase.” Instead, they’re coming into a store knowing that they’re interested and knowing that they want to buy the product. Consider the flip side: if you tried a product and didn’t like it for some reason, why would you ever buy it?

The folks we drive into stores because of their sampling experiences are a much more committed audience than folks who have not sampled the product. People driven to the store by other forms of awareness building and promotion marketing tactics need to make that ‘trial purchase’ where that consumer runs the risk that her family will reject the product.

With sampling, because these consumers are pre-sold, if you will, when they do go in and buy, they are much more likely to continue to make purchases. This makes them a more valuable shopper for the retailer.

We’re seeing more and more evidence that people are really beginning to utilize big data (#bigdatause). That’s what this is about. People are using the data in intelligent ways for marketing and, most importantly, as evidence.

The whole industry, especially the packaged goods world, is requiring, demanding and seeking proof – show me clear results of my marketing spend. In our particular part of the world, the product-sampling universe, we have the means, methodology and tools to show direct, purchase data-based results. We encourage all of the brands working in the sampling universe to expect clear, cogent and unambiguous results.

**Methodology Note: Data reported above comes from studies conducted by Spire (A Datalogix Company) for StartSampling, Inc. In the graphs above households that received our mailed sample were identified by matching to known shoppers. A proper control group (similar buying pattern at same stores in pre period) was created and then the analysis was performed to see if the dynamics of brand and category purchasing differed between the two groups. The data was handled in an anonymized fashion so all companies involved received only aggregated data reports without individual identification being possible.

Categorised as: Marketing, Sampling, Sampling Works!