Growth hacking to validate and optimize the value proposition for Instant Noodles with added protein.
Reduce risk by validating your assumptions early. Use structured testing before making significant investments in order to make smarter, more confident decisions.
Find what truly resonates. Test tailored value propositions with specific customer segments instead of aiming for broad appeal.
Use data to guide your next move. Run targeted experiments - like ads and landing pages - to gather real insights, fine-tune your offer, and build a solid case for your idea.
One of the top 3 global food and beverage companies had identified an important trend in nutritional preference amongst younger people. They found out that protein is an increasingly important ingredient in the products they’re choosing. With an upsurge in products like the high protein yoghurt variant Skyr, protein bars, and even protein pancakes, our client wanted to find out if they could sail into their own blue ocean by offering instant noodles with added protein.
Current high-protein products target health-focused consumers. Our client aimed to reach those prioritizing taste and indulgence but still desiring more protein, suggesting tasty, easy instant noodles with added protein. The client had interviewed people and identified personas, so we could move ahead with step one of our Explore Track.
We first need to identify what are the variables that need to be true for the product to be successful, our assumptions. We try to be as specific as possible when identifying these assumptions, so that we can design the perfect experiment to validate them. Based on the project goal, we divided them into 3 categories. Some of the assumptions fell neatly into one of the categories, some were a combination of categories.
After writing all the assumptions down, you will probably end up with quite a lot of them. Not all of them will be equally relevant to be validated. To select the most important ones, we use an Importance – Certainty – Matrix. Assumptions that are very important for the success of the product will be plotted high on the Matrix. For example: ‘People want more protein in their food’ is very important if we want this product to be successful. Due to the prior research however, we already knew the answer to this question. Therefore the level of certainty was fairly high. However, the assumption that people wanted 'a product that is tasty with a high amount of protein', was very important and far less certain: It needed to be validated.
Assumptions that have both a high importance for the success of the product, as well as a high level of uncertainty will end up in the top right quadrant. These assumptions are the assumptions that we were to validate in our experiment.
Pro tip: the best way to plot the assumptions is relative to each other. Start with one assumption. Place the next one more to the right if it’s more uncertain than the current assumption, and higher if it’s more important. This allows you to anchor your decisions.
Now that you know what to test, you need to group the assumptions based on what can be tested in the same experiment. There are a lot of experiments you can choose from, and the right choice will depend on the goal of the project and the assumptions. To best answer the questions in our 3 categories, we decided that we needed large-scale data on which value proposition works best for what persona. We designed a 3-step experiment.
Test your value proposition against very large target audiences to receive indisputable results.
Measure slightly different concepts head to head to identify the purchase intent for adding different features.
Collect valueable insights from the most interested section of your population and learn their key drivers.
Setting up a multi-layered experiment allowed us to test a large amount of assumptions in a matter of weeks instead of months.
We have created an off-brand product that represents the same values as the product that the client wants to place in the market. We created four different advertisements and ran them for two sprints of two weeks on social media. We split the experiment in two sprints to iterate on the value proposition to provide an even more precise recommendation.
Pro tip: Off-brand testing brings the advantage that you can create speed and test the value propositions on its merits without external influence, but it will be less predictive as on-brand testing, because brand associations are not taken into account.
We tested three different ways of talking about the protein in our product to see which one resulted in the highest purchase intent. We compared website pages that said 'Protein', 'High in Protein', and showed the exact amount of protein.
To test the type of noodles, we doubled the entire experiment and tested both the ads as the landing pages with two different types of noodles, thin noodles and bigger noodles. We launched both experiments in two countries and for two different target audiences, which implies that a total of 32 different advertisements were running simultaneously. (4 value propositions x 2 noodle types x 2 target audiences x 2 countries), leading to twelve different landing pages (3 claim tests x 2 noodle types x 2 languages). Each click on the landing page that showed purchase intent, leads to the 'smoke page’ where we tell the prospect that the product is not yet live, and invite them to share additional input through a qualitative discovery survey.
People that demonstrated clear purchase intent landed on a short, qualitative survey where we asked additional information regarding current noodle-eating habits and brand associations. This allowed us to gain deep insights into the potential customers of protein-enriched instant noodles.
With the results from the discovery survey we were able to have insights in the current noodle-eating habits of the potential customers and predict the expected cannibalization-effect towards the current offer.
With an large dataset flowing in, we have sat together to translate this raw data into concrete answers and recommendations. To do this, we went back to our assumptions and found for each one a data point that could answer the question. For the trigger test we measured the amount of people that interacted with the advertisements, and analyzing significant differences between them. We compared it with our base measure for showing significant interest and ended up with a ranking of most desirable to least desirable value proposition.
On the landing page we measured the purchase interest by measuring how many people clicked on the ‘find a store’ button. Triangulating the data from the trigger test with the data from the smoke test allowed us to create a business case for each value proposition and noodle type, indicating a clear winner.
Experimentation is key to product success. The large amount of data we gathered indicated that the market was hungry for protein enriched instant noodles. Thanks to the combination of experiments, we tested different layers of customer interests and behaviors. Triangulating the data points allowed us to provide a comprehensive answer to the research question. Getting these insights early proves that validation through experimenting before launch is vital for putting the development efforts in the right direction.
Having hard data that predicts the success of your product or service will make the difference when investing in new ideas. We welcome you to reach out to our amazing crew to talk about your next idea and how we might navigate your challenge together.
![[object Object]](https://static-files.umso.co/lib_SgVrTchGksgQdBiV/5to2tmehoshuavn2.png?w=306)