Dogs and Data and Assumptions
Bonus Week and Off Topic
We’re wrapping up summer, and it’s hard to believe how much of the year is gone:
With five weeks in August instead of four, this is one of the bonus weeks of the year where we’ll explore some areas outside our normal areas of this newsletter. So we’re going slightly off topic, though not too far askew this time (I promise I’ll go really off topic in subsequent newsletters).
Product Thinking is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
We got a new puppy over the weekend. He’s a cute golden doodle, and he’s quickly winning over everyone he comes in contact with, as you can imagine. I mean, just look at that face.
The only one who still has any lingering questions is probably our 11-year-old golden retriever, who is still on the receiving end of most of the puppy bites.
So I was intrigued by an article that hit my inbox:
Am I more of a risk-taker? More of an adventurer like the picture from the article above portrays? Because I own dogs? And now even more so because we’ve got two?
I certainly like to think so. It’s an intriguing image. But I’m almost certain that I am not more of a risk taker, at least not significantly outside of an expected range given my age, socio-economics, family, and other circumstances. I don’t have the data on this for myself, but assuming I’m somewhere in the meaty part of the bell curve is probably a safe bet.
Of course, with me, N=1, so that’s not too convincing. But it gives me a starting point. A frame of reference.
Influence or Reflection?
As I pondered this idea, I can’t help but wonder how much we’re influenced by our pets (probably some amount) and how much our pets are a reflection or representation to the world of our existing traits.
I know many people classify themselves as dog people or cat people (or some have both). Not to reduce our pets to objects, but many people also have similar preferences around the types of vehicles they drive, the places they live, the clothes they wear, etc. These decisions also influence the decisions we make, but are also reflections of who we are. How much of either are they?
If I’m completely neutral on a risk continuum (whatever that looks like), will I move to a riskier side by adopting a dog? Or a less-risky side by adopting a cat? If I have no predisposition to either, could I be influenced significantly by my choice of pet?
Could a risk-intolerant person be made to take on more risk by adopting a dog? Could my 70-year-old parents be encouraged to take on more risk by adopting a dog? Even though their risk-tolerance may be quite low right now? Or would it be relative to their existing risk tolerance?
Some interesting initial thoughts, but let’s move into the actual article as well.
From the Article
The article starts with this idea:
Dog owners tend to take bigger risks and respond more to reward-oriented advertisements. Cat owners, on the other hand, are more cautious and more likely to react to ads emphasizing risk aversion. Those are the two main findings from new peer-reviewed research I co-authored.
Now I’m not formal researcher and certainly no statistician. So I won’t review the work that the researchers for this article did, though I read through the paper that they wrote in addition to the article. And while the research methods were likely sound and the statistics are probably good, it’s some of the bigger questions that linger with me.
The headline obviously was meant to grab anyone who owns a dog or a cat, while the research is really focused on how to market to dog owners vs. cat owners. So that is the first flag that stands out. But we’ll set it aside since all headlines are misleading these days.
The main studies in the article have to do with three areas of focus:
Bidding for ads
Our first pair of studies looked at pet ownership data in U.S. states and compared that with several crude measures of risk-taking. For example, we found that people in states with a higher share of dog owners, such as North Dakota, had a greater prevalence of COVID-19 infections in 2020 than states with more cat owners, such as Vermont. Although we controlled for political orientation and other variables, our results show only a correlation. The reason dog ownership seems associated with more COVID-19 cases, for example, could be that dog owners take more risks – or they simply have to take their pets out for walks more often, which means greater exposure.
Fortunately they acknowledge many of the limitations of this first study in the article itself. For me, I wonder if it doesn’t completely rule itself out. Does dog ownership vs. cat ownership become completely meaningless in this context when we’re talking about different states and infection rates?
I’ve worked for several companies throughout the pandemic, all with people located across the United States and across the world. We also did a little traveling during the pandemic. The differences in how different locations were reacting to the pandemic at different times were stark. While some states or countries were still very strict about mandates (Hawaii or California for example) other states were completely the opposite (Texas or Florida for example). This goes for countries as well, such as Australia and Italy.
Taking politics out of the discussion, the culture in the area will determine how others are reacting. And then demographics and other factors also play a part. While things like pet ownership are interesting, I wonder how much of a role they play in infection rates rather than are just traits of a given area. If people tend to own dogs then more people will tend to own dogs.
In one study:
We gave participants an imaginary US$2,000 and asked them to invest any portion of it in either a risky stock fund or a more conservative mutual fund. Dog owners, who made up 53% of participants, were significantly more likely to invest in stocks and also put more money at risk than cat owners.
And in another:
we asked 225 people to view four print ads featuring either a cat or a dog and then decide how to allocate a $2,000 investment, as in the previous study. We found that exposure to dogs led participants to be more likely to invest more money in stocks.
I’m always wary of hypotheticals. Sometimes they are the best you can do. But we should try to gauge actual behavior rather than “how would you invest” or “how would you do XYZ”.
I wrote about this in my review of The Mom Test:
People will often tell us one thing and then do another when it comes to actual money. It’s easy to say they will buy something “hypothetically” but when they have to put their actual money to it, they might balk.
So I wonder how do actual dog owners invest their money? How do actual cat owners invest their money? That seems like something we could find out by looking at actual portfolios and controlling for age, income, and demographics. And that sounds fascinating. I’d love to see someone do that because it would probably tell us a lot about actual risk appetite (at least in financial decisions) for dog owners vs cat owners.
Bidding for Ads
Students who recalled an interaction with a dog offered bids significantly higher when they were exposed to the reward-oriented rather than risk-aversion ads. In contrast, those who recalled a cat offered much higher bids when they saw ads focused on risk aversion.
Finally, the article discusses ads based on reward focused vs risk aversion. And once again, this confirmed the authors’ hypothesis of the stereotypes of dogs and cats. Experiences involving dogs primed the undergrads involved for rewards and experiences involving cats primed the undergrads involved for risk aversion.
Some of my same questions apply here. Since this was primarily with undergrads, how broadly applicable is it? Since it is largely hypothetical or dealing with very small amounts of money, is there enough risk involved to truly assess risk preferences? Do risk preferences change over time?
It’s interesting to think about how much our pets influence us vs. how much they are a reflection of our existing preferences. I suspect it is both, though how much is an open question. We certainly can’t escape how our decisions shape our environment, which in turn shapes our decisions.
It seems like a stretch to say that dog owners take more risks and conform to their pets’ stereotypical traits without adjusting for many of the other factors at play. Which is why it is so important to look at all of these types of articles and research studies with a critical eye. And ask questions and examine sources. When all the participants are around the same age and come from roughly the same pool, we may get some small amount of signal, but can’t really confirm too much.
Which also goes for our own research for our products and our companies. It’s easy to confirm our own biases and assumptions, but we have to dig deeper and question our hypotheses rather than confirm our stereotypes.