Product Sampling - Data/Measurement/CRM/Predictive Analytics

Project Description

We are an agency holding company consisting of six subsidiaries, each with their own areas of expertise. For this project, we will be focusing on our Lifestyle Sampling & Fulfillment subsidiary.

This subsidiary distributes a high volume of Consumer Packaged Goods (CPG) Product Samples to a variety of venues (hotels, spas, colleges, events, trade shows, etc.), and also handles warehousing, storage, and overwrapping of the CPG Products prior to distribution. 

Background/Problem: There is currently no good measurement tool in existence that measures Product Sampling Effectiveness.

The marketing mix models (MMM’s) were built for marketers to input the specific details of their marketing spend to determine Return On Investment (ROI) at the market and tactic level. These models do not include Product Sampling in the mix, due to lack of volume (have to reach 1% of the population with any marketing tactic in order for it to be measurable within the models). These models were built many years ago to account for traditional media, including TV advertising, radio, print, out of home media, direct mail, in-store media, social/digital media, etc., commissioned by advertising agencies that do not traditionally control the product sampling spend, and therefore they self-servingly omitted measuring tactics, like product sampling, that do not drive revenue into their agency. Therefore, marketers that had historically spent billions of dollars on Product Sampling every year have begun to greatly reduce or all together abandon sampling as a tactic, because they are unable to quantify sampling’s short or long-term impact.

Our previous strategy in using market research have utilized Attitudinal research methodology, whereby we would interview consumers after they received a sample in order to determine their previous purchase habits, and their future purchase intent, using a test/control methodology to determine incrementality. The limitations of this methodology (and other similar methodologies) are many, including the inability to correlate actual purchase with claimed purchase, and the lack of any direct link to in-store sales purchase verification, causing many marketers to believe that it is impossible to measure out-of-store sampling effectiveness. We would like to solve this problem, as we believe doing so will not only help us demonstrate the power of this tactic and ultimately result in more manufacturers embracing product sampling and specifically OUR product sampling solutions, but we can also offer this measurement solution to the manufacturers as a SAAS platform to evaluate their proprietary sampling programs, and license to other sampling vendors/suppliers to empower them to demonstrate the effectiveness of their sampling vehicles. 

As some added information, below are a few of the other current (rudimentary) methods being used to measure sampling effectiveness: 

1)    Provide a sampling campaign specific coupon along with the samples, which has a unique barcode, and can therefore be tracked and measured (the main problems with this tactic are that: (a) it does not speak to retention; and (b) many times people will not clip and save and bring the coupon with them to the store to make the purchase).  The goal of a sampling program is to induce trial that leads to purchase conversion.  Evaluating sampling effectiveness by adding the additional hurdle of sampling to people who are both inclined to use coupons and who follow through and redeem the coupon at purchase eliminates the inclusion of all purchase that occurred without a coupon and negatively impacts perceived ROI. 

2)    Determine a baseline of sales prior to sampling and then measure sales a few weeks/months post sampling to determine the increase. The main problems with this tactic are that there are so many other variables that could be effecting the baseline, it is hard to attribute it directly to sampling.  Sample distribution can often occur over a wide geographic area, and sometimes over a long period of time (weeks or months).  In addition, purchase cycles impact the timing of purchases (if you just bought a bottle of shampoo, it could be 2 months before you buy another one) and purchases can happen across hundreds of retailers, making it impossible to see any meaningful lift at retail. 

Project: Create a model that will allow manufacturers, retailers and marketing services suppliers to accurately measure, post-program and/or predict pre-program, the effectiveness of product samples distributed via a variety of different methods, including through venues, via direct mail, via on-line request, and in-store. 

This will allow specific campaigns to be measured, but also ultimately power a predictive model companies will be able to use in order to plan their sampling campaigns based on projected ROI. Cross-tab segmentation might evaluate effectiveness of subgroups based on things like demographics, geographics and psychographics.  We need to be sure to attempt to predict both: (a) increase in sales in the short term; and (b) the long-term impact of those sales. The probability of retention (multiple purchases and lifetime value of the conversion) is an important factor.

In addition to gathering quantities of market research findings and developing a measurement tool, we desire the ability to re-engage participating consumers via opt-in marketing campaigns for ongoing CRM. The database will be populated in two ways: 1) product sampling recipients who participate (both those who do and don’t participate in the research) might be incentivized to participate and/or opt-in by offering them access to future sampling campaigns and/or sweepstakes opportunities; and 2) call to action messages prompting consumers to sign-up will be delivered with the +75 million product samples BC distributes on behalf of our clients annually. The long-term vision would be to have a database whereby we can: 1) conduct research studies and gather insights; 2) execute highly targeted campaigns for marketers; and 3) own a consumer-facing website/app where people register to receive samples and special offers. 

Expertise Required: Market Research best practices/Data Capture/CRM/Predictive Modeling

Data Sources At Our Disposal: We will be able to provide what product we are sampling, what venues the sampling is being distributed in and the quantity of the samples. Other than that, we will need to gather any other information we deem necessary to complete out goals as the technology is rolled out. 

Deliverable: We want to create and own a platform that has Data Capture Technology capabilities, with the end goal of capturing enough data to develop an algorithm that we can plug: (a) product; (b) spend; and (c) network into in order to predict the short term sales increase as well as long term retention of such consumers. Lastly, the technology will have to have the capabilities to function as a CRM solution as well.

Please provide the amount of hours required to complete project in your proposal.