Keeping your brand fresh and your portfolio strong can make the difference between winning the customer journey or falling behind the competition. Yet it’s common that portfolio decisions are either made without the benefit of consumer insights or with approaches that don’t truly optimize your learnings. Either way, the result can be devastating if the right tool isn’t applied or the right information isn’t captured.
Whether you are considering introducing a second product to your lineup or are managing a portfolio of dozens of SKUs, there are a host of advanced analytic techniques that can bring clarity to your decisions. Here are the 5 major approaches you should consider and an assessment of when each one should be applied:
1. TURF via Max/Diff
Total Unduplicated Reach and Frequency, while not the most statistically complex technique, can prove to be powerful for identifying a subset of pre-configured products. It works by first prioritizing your products via a simple trade-off exercise and following this by identifying the degree to which the appeal of one product overlaps with the appeal of other products in your portfolio. The result is an understanding of which products can be dropped while keeping a relevant substitute product.
When to Use: Most appropriate with a relatively small number of pre-configured products.
2. Purchaser Profiling
The idea behind purchaser profiling, like TURF, is to identify products that overlap in terms of who is purchasing. By understanding who is purchasing, we can both determine groups that are already served by existing products and those groups where a white space exists for a potential new entrant.
When to Use: Most often used when portfolio management was not a primary objective of the research and we want to reverse engineer portfolio optimization into an existing study.
3. Market Basket Analysis
Like the above two options, Market Basket Analysis is not a particularly computationally intense technique. However, by determining product co-purchase, it helps you understand the degree to which some products are complementary. By identifying complementary products, we can identify bundling opportunities. The actual purchasing behavior data assists in determining the degree to which each product generates significant share and revenue.
When to Use: Can be used either with actual POS/behavioral data or from self-reported purchasing data. Because this approach can be applied on actual sales data, additional data collection may not be required.
4. Line Optimization Tests (LOT)
With an LOT we start to get more advanced in our approach. Here we’ll use a discrete choice or allocation-based conjoint to determine the optimal mix of pre-defined SKUs. The decision as to which exercise is best will depend on the category. We want to make sure we align our technique with how consumers interact with products in your category. For fast moving categories or categories in which consumers may have heterogeneous preferences (for example, a consumer sometimes wants Product A and other times prefers Product B — for example, yogurt flavors, movie genres or, among physicians, therapy choices), an allocation-based is more appropriate as it allows the user to specify these differentiated preferences. For occasional purchases (for example, a refrigerator or car tires), a Discrete Choice is oftentimes more appropriate.
When to Use: LOT is most appropriate when we have a set of pre-configured products. Those products may vary based on price, but all the other features are fixed.
5. Portfolio Optimization
The distinguishing element between a LOT and Portfolio optimization is that in a LOT, the products are all pre-configured SKUs. In a Portfolio Optimization, we are simultaneously optimizing the product(s) and the portfolio. As with LOT, we have options with respect to the specific type of choice exercise we’ll use. Discrete Choice and Allocation-based Conjoint are two available techniques. But, because we are manipulating not just product availability but also product configuration, our range of techniques is opened up to other solutions such as Adaptive Choice-Based Conjoint (ACBC) or Menu-based Conjoint (MBC).
When to Use: Portfolio Optimization differs from LOT in that the products are not pre-configured. We don’t yet know what the products we should offer look like yet in terms of features or price.
If you’re in the process of scrutinizing your portfolio for its overall impact and to keep it fresh and relevant, it’s important to take a hard look at each of these options to determine which one is the best fit for your situation. We’re happy to help you navigate this decision.
Contact us to learn more.