Measuring the moment: How product consumption data delivers true consumer behavior
Editor's note: This article is an automated speech-to-text transcription, edited lightly for clarity. To view the full session recording click here.
In Adrich’s session, Adhithi Aji, CEO and co-founder, focused on understanding consumer behavior through smart products. Measuring product usage offers researchers unique insights which help explain why consumers are behaving in certain ways. Understanding consumer behaviors helps brands make the changes needed to continue to be leading companies.
Session transcript
Emily Koenig
Hello and welcome to the session “Measuring the moment: How product consumption data delivers true consumer behavior.” I'm Emily, an editor at Quirk’s. Before we get started, let's quickly go over the ways you can participate in today's discussion. You can use the chat tab to interact with other attendees during the session and you can use the Q&A tab to submit questions for presenters. Our session is presented to you by Adrich. Enjoy the presentation.
Adhithi Aji
Good morning everyone. A pleasure to be here. In today's day and age, there are a ton of smart products, call it smart thermostats, smart shoes. So all of the products are getting intelligent and we at Adrich wanted to specifically understand how smart connected products could help supercharge research specifically in understanding real consumer behavior and I’m excited to showcase some of the work and case studies we've done and really understand where in the realm of AI and data do smart products fit in and how that really helps you get one step closer to your consumer and create smarter ways to engage your consumers as well.
So, one of the biggest challenges we have is how are consumers really interacting with our products, right? And I highlight the word really because a lot of times consumers report saying, for example, peanut butter, yes, I used it to make pb&j’s for my kids in the morning, but also you are possibly consuming it middle of the night when you got up and got hungry.
Similarly, let's take cleaning products. Consumers say, Saturday mornings are my deep clean routine. But let's face it, you've had a hard week and you just want to slack. And so really, you might not be following a routine. And so, there are certain blind spots, right? There's a say-do gap, and then there is also a recall gap because if you send a survey three weeks after someone uses the product, they don't remember to answer that question or three weeks ago what they were doing with the product.
So Adrich wants to help here by closing both of these gaps by capturing real time product interactions so we can get accurate insights and also trigger in the moment surveys, add point of views. So really we can eliminate the memory bias. And how we do that is through smart labels. We put this on any product, it's a peel and stick. You apply it and any product, it becomes smart, it gets connected and captures in real time, data of how consumers are interacting with the product. So, we get time of day of use, how often it's used, duration of use, storage temperature, dosage in terms of how much is being pour per use that can get really granular hours between use and also location of use in-home vs. out-of-home usage.
All of this data streams autonomously so we make sure consumers are interacting with the product, are using the products as they normally would, and the smart label works in the background and autonomously collects this data into our cloud where we present to you a dashboard which at an aggregated level gives you insights and it also enables surveys at different points of use. So, you want to understand packaging feedback so you set up a survey at first use and we have a bunch of triggers that really helps you get to the why.
So not only you are getting what is happening and how consumers are using the product, but why. And so, our mission is to use the smart products to give you a 360-degree view of how consumers are interacting with the products, which includes the what, the how and the why. So, let's dive into some case studies which really brings out some of the aha moments that we've created for the brands and learned and grown the brand together.
So, the first one is in a new product launch, understanding consumer behavior. This was launched in market and within a few months we rolled out the study and the brand really wanted to understand jobs to be done, all of our favorites, right? We have various usage occasions that come about when new brands are launched. And so really wanted to understand what the real use cases are and specifically because it was a disinfectant product, seasonality and understanding heavy users’ patterns. The hypothesis was this is going to be used primarily in the flu season and probably in most likely in the living rooms. And this we did as a longitudinal study, so 12-month study where we wanted to overlay weather data, understand seasonality and decided to roll it out as a single scent variant to keep all the variables constant in terms of the sense so that we get a baseline over the 12 months.
Interestingly enough, we captured patterns of usage between the days and the times. And it was interesting to see that evenings and early mornings were dominating and specifically certain weekdays. Now with the surveys we were able to also understand the most common rooms that were used in. So, after every spray or survey was sent to understand, okay, which room was sprayed, and not just that, but what exactly is driving that occasion. What was the usage occasion, what was the reason that they were using the spray?
And the very interesting insights came about. So as mentioned, the initial hypothesis that it's going to be used mainly in the flu season and then living room was understood to be the most likely use case, but we actually found that it was used as a routine and the bathroom was dominating the usage rather than the living room. And even though it was positioned as a disinfectant product, it was primarily used for odor removal. So almost a flip 180 in terms of the initial hypothesis, but interestingly enough uncovered many more use cases.
And so understanding the now what, what could the brands do with this? Certainly reposition the product for more routine use cases, target peak usage times with contextual campaigns, especially knowing the time of day of use and when exactly people are maximizing the usage, timing the advertisements and campaigns around that would increase the marketing ROI. Develop variance for different rooms, because it was a longitudinal, the smart label was able to detect when consumers were running out of the aerosol and we reordered on time. And so, they got the replacement on time and with that they were able to leverage not just demand patterns and what the reorder rates are, but also enable shipment at the right time for having continued usage of the product without any drop off.
So really the P&L impact as you'd have guessed, increased consumption over the years. So, it's not necessarily seasonal uncovered or routine use case, which could drive up consumption higher repeat purchase rates. We actually saw some of the heavy users finish the product within two months of use, which is interesting because the brand thought it'll take at least six months for people to even finish one unit. So, we did see a significant number of heavy users who finished the product in two months. So that again was a new segment and revenue uplift from the positioning. So, the fact that it goes beyond just disinfecting the air was certainly a big use case that the brand could leverage.
Another use case. This is common where we want to understand competitive analysis and with the smart products, so with the smart labels, we were able to get really granular consumption insights and specifically amount per use, the dosage. And in this case, our smart label is designed not to introduce any bias. So, we do use AI and neural networks to understand how much is being pulled without using weighted scales. And that removes any bias in terms of consumers having to weigh the product after every use. And so, it streams the dosage information as well.
So, here's an example of how the dosage was used. We grouped each occasion to a four-hour occasion, each use to a four-hour occasion. We were able to benchmark the competitor brand's dosage to the main brands dosage and got really granular in terms of which time of day, how much more was used, where the brand was having higher consumption compared to the competitors.
And really now the so what? First and foremost, the brand thought that mega size boxes would get completed within two weeks and a significant portion of people had not finished the product. Now, that applied for both brands. So that itself was a learning. However, there was a higher percentage of people who finished the brand compared to the competitor brand. So that was a win there.
There was also a higher total consumption of brands. So, the total amount consumed over the two weeks was higher for the brand. And by actually mapping out each usage occasions over the day, the brand uncovered more usage occasions in late afternoons, early evenings and late-night snacks. So very interesting learnings.
And so, applying this to, now what? How can the brands use this? So really doubling down on the usage occasions there were, beyond just breakfast, there were snacking occasions that definitely could be leveraged upon. We also understood that it's more family driven occasions through the surveys that were triggered to understand. They were more family driven compared to solo consumption, but also in some cases the competitor brand had higher dosage per use. And so, there was a potential to introduce smaller single-use snacking packaging that could leverage those new usage occasions.
So, with the P&L impact, of course, increased frequency of use, there were so many more times that consumers use the product, larger family packs which sort of would increase the overall usage per family and then higher volume sales, right? That's, at the end of the day, that was what impacted the P&L.
A third use case is sales does not equal consumption, right? Typical misnomer is, OK, we can make our decisions on sales and sales more or less reflects consumption. We specifically did this project to understand whether that is true and there were a lot of things that came about in this project that really showed that no, sales does not equal consumption.
So, this was a beverage brand and it was more of a portfolio-level study rather than each product and six beverage brands were rolled out of the same company within their categories. Six different flavors were rolled out. It was another 12-month longitudinal study, of which consumers out of the six consumers were allowed to choose three. So really understood which ones they're choosing.
And after they got the products, we also observed baseline consumption patterns in a natural setting, sort of the three products, which one they would pick to consume, how much was consumed, what was the order of consumption between the three brands. We also wanted to understand what the impact of weather and events, events like the Super Bowl, Thanksgiving, all these fluctuating or how these important milestone events over the year could impact consumption. And we had an interesting recipe engagement. So, the same triggers at point of use was also used to try out different recipes to see if a particular recipe engaged the consumers more.
And as mentioned, the initial hypothesis was out of the six brands there is a single sales leader and that likely is also the consumption leader. And when we dove into the data, this one example of what we saw, for one brand four was the sales leader, but actually when we saw the consumption brand one was the leader because it had significantly higher rate of consumption. But one other thing that came about which was non-intuitive and super interesting in that infrequent users showed a more loyal behavior.
In a natural course, if it's a three-month study or a month-long study, typically we would think, OK, the frequent users are using more of the product so they're more loyal and they are our target consumers. But because of the longitudinal nature of this project, we were seeing that even though frequent consumers use more of the product, they were not consistent over the course of the year. There was drop off. Either they would use some other brand in the summer or the consistency was not there. However, infrequent users, even though their consumption per usage occasion was lower, they had a more consistent usage over the course of the year. And we also triggered a purchase intent question when they were down to 20%. And we validated this kind of behavior with that purchase intent question where frequent users would respond, we don't care about the brand, we would probably buy some other brand in the same category. However, infrequent users preferred the same brand in the same flavor. And so clearly this sort of uncovered a whole new segment for the brands to focus on in terms of infrequent users, but consistent and more engaged users over the course of a year.
And so that was one big aha moment, strong Millennial brand advocacy was another segment that was outside of what the brand's hypothesis was. And we also were able to see which one along the weak, which brand was showing a consistent usage, but also this was used as a benchmarking study so we took averages across all six brands and understood how each brand is benchmarking against the average.
So, this was again a portfolio-level insights where it was almost like a connected home where three connected products were given to each users and really understanding how, at a portfolio level, the brands could work with each other and potentially bundle potentially understand the benchmarking and not only insights we gained at each individual brand level, but also at a portfolio level.
So, coming to the so what, of course. there is a difference between consumption and sales and specifically around the repurchase, when we talk about repeat sales, it is better to make decisions on consumption data, identified demographics, which more loyal users. Recipe engagement was a big one where it was triggered at end days of no use. So, when we saw drop off, a recipe was sent and we were observing whether there was reengagement, reinteraction with the product and there was significant comeback and reengagement with the product. And then purchase intent was more with infrequent users, even though they finished slowly, but they were consistent over the year.
Coming to the now what. So, with all these shifts and the learnings certainly basing the demand planning on consumption rather than sales improved the prediction, better, increasing repeat purchase and consumer lifetime value because we were to identify a segment that was more loyal. That was also interesting to note, that the repeat purchases were more likely happening from that demographic. Triggering recipe engagements really reduce brand switching. So, timing that recipe engagement when there was drop off so consumers don't switch brands was an interesting way and a new approach to potentially marketing and hyper-personalized marketing.
And then again, capturing incremental revenue from new segments that was not in the hypothesis of the brands originally. So really what we are coming down to now on the left, where does all this fit in the broad scheme of AI? Right on the left we are seeing the data that we are basing all our models and AI on is existing data but some analysis, but then there is no element of behavior. Behavior is also modeled and AI is as good as the data you feed it. And so, when you incorporate real, accurate consumption behavior or consumer behavior, now you have a foundation data set that really is very accurate. It improves your foundation on which you are building your AI algorithms. And so, you are also getting enough validation and that of course improves decision.
So that's where connected smart products fit in. That's where understanding, real interaction, real behavior and the level of depth you can get with sensor-based technologies and specifically with Adrich because we've done over 25,000 units in the field across 25 product categories and both at a portfolio-level as well as individual product level. So really bringing in the next generation of connected home and smart products and leveraging that into the realm of AI to give you more precise consumer insights, more accurate and more depth in terms of understanding your consumers so you can build better products suited to their needs.
Thank you so much. I'm Adhithi Aji, co-founder and CEO of Adrich. We'd love to talk more. Please feel free to connect and certainly I would love to dive deeper or answer any questions. Thank you for your time, I appreciate it.