3 ms·

Found this link to the paper, since OP's SciHub link has rotted: https://dspace.mit.edu/bitstream/handle/1721.1/130136/The%20Surprising%20Breadth%20of%20Harbing

Found this link to the paper, since OP's SciHub link has rotted: https://dspace.mit.edu/bitstream/handle/1721.1/130136/The%20Surprising%20Breadth%20of%20Harbingers%20of%20Failure%20June%202019.pdf https://dspace.mit.edu/bitstream/handle/1721.1/130136/The%20... This is an interesting finding no matter what, but I think part of the splash comes from the idea that some people might be harbingers _independent of the rest of their identity_. Since it's not obvious that liking crystal pepsi and watermelon oreos are related, we can speculate that the people inherently like doomed products. On my read, there's a more boring hypothesis that is equally supported: in any given period, some demographics (age, income, cultural -- all the stuff that we know gets clustered geographically) are more profitable than others. The products of the less profitable demographics get discontinued. The main difference between this and the "harbinger" hypothesis is whether you can make surprising predictions. The paper doesn't really prove its predictive case because the calibration and holdout data come from different products in the same time period -- meaning that information about the time period leaks into the model that wouldn't be available if you were truly running a prediction. If you feel like you yourself are a harbinger of product doom, maybe you're just a part of a group that isn't making enough money for sellers! Your choice whether to call that unfair or a victory :)

lordgrenville 4y ago
> Found this link to the paper Hurts my eyes. SciHub URL + DOI (10.1177/0022243719867935) works for me.
avereveard 4y ago
Alternative: people that like to experiment are the first jumping on new products, and people that like routines aren't - that would be enough to bias the experimenters into the harbinger group, and it would be useful to check if there's an overlap when measuring for household that were early adopter of successful products
- evrydayhustling 4y ago
  Good thought - people with "high variance" tastes might buy an unusual number of duds, but they might also be more likely to pick breakout products. An analysis for that would need to include degrees of product success, not just survival.
- mrandish 4y ago
  This was my reaction as well. There are two different ways to be an outlier in prefering products more likely to be discontinued. Very early adopters who have uniquely discerning selection criteria or tastes may seek out new products which meet their preferences but if the product is "too early" to capture much market share, it is likely to fail. When their 'outside the norm' preferences end up catching on with the broader market they look like trendsetters or taste makers. When their atypical preferences don't catch on, they look like the "harbingers" discussed in the article. I suspect the reality is that both types exist. The article says harbingers tend to clip coupons more. This implies they may be price-sensitive, lower income or perhaps trying to stretch their lifestyle beyond their means. To me, harbingers sound like people who are open to trying novel things but also tend to make quicker, perhaps ill-considered decisions. Over time this would lead to a trend of making poorer choices on average, such as remaining in zip codes where the negatives might outweigh the positives but not obviously so at first glance - explaining another finding of the study. I've occasionally had strong preferences for product features clearly not valued as highly by the majority of the market. That means sometimes I have to seek out or pay more for the items I prefer. However, I'm not at all price sensitive when it comes to products that differentiate themselves in ways which are important to me. Also, I almost never clip coupons. Conversely, when I judge a cheaper product to be equivalent in the ways that matter to me, I'll go for the cheaper one - all other things equal (ie I wouldn't drive to another store to save 75 cents). Other times, products with features I find uniquely valuable fail in the market because there aren't enough others who value those features. But I also have a large number of examples where I was very early to a product niche which was not a 'thing' but then grew into a sub-category popular enough to be sustainable and competitively profitable. An easy example is mechanical keyboards. I was one of those people seeking out old IBM keyboards and adapting the old big DIN connectors to USB. A decade later I was participating in group buys to get good mechanical keybs. Now, another decade on, I can walk into Best Buy and find three or four mechanical keyboards. I'm unusual in that I'll pay >$500 for the exact keyboard I want yet seek out a very particular brand & model of shirt or pants I prefer from Costco or Wal-Mart despite being the cheapest, apparently undifferentiated offering. But in those cases, it's always because of specific features that item uniquely possesses over others. When friends ask me about some of my less conventional product choices I can explain in detail exactly why this one is objectively better to me than the others. In many cases, following such conversations my friends end up adopting either the product or at least the previously under-valued differentiating feature. A personal example of the negative outcome (failed product) would be the Zazzle, a cool toy which ultimately failed in the market despite my buying one when I saw it at retail. My Zazzle is sitting on the bar in my game room and once discovered by friends becomes a favorite. One friend told me they tried to get one on eBay only to find them selling for hundreds of dollars. I have a higher than usual rate of having discontinued products I bought when new which are now selling for multiples of original MSRP. I didn't buy them because I thought they'd go up in value, I just tend to buy stuff I want if I see unique value in it.
smcnally 4y ago
A corollary to this is when a single brand and formulation become multiple brands and formulations, but the original disappears. e.g. Triscuits, Cheerios, Cranberry Juice. I can find * Tzatziki Quinoa Basil and Garlic Triscuits, but not Original or Low-Sodium. * Apple Honey Nut Ancho Kobe Beef Cheerios, but not Cheerios. * Mango Chutney Cranberry with Turmeric & Essence of Nightshade juice, but not Cranberry see also: potato chips, hummus, frozen waffles
- AlbertCory 4y ago
  Funny because I have both Original Triscuits and regular Cheerios in the house right now. Panic! Maybe I shop at a Harbinger Store [tm].

lordgrenville 4y ago

> Found this link to the paper Hurts my eyes. SciHub URL + DOI (10.1177/0022243719867935) works for me.

avereveard 4y ago

Alternative: people that like to experiment are the first jumping on new products, and people that like routines aren't - that would be enough to bias the experimenters into the harbinger group, and it would be useful to check if there's an overlap when measuring for household that were early adopter of successful products

evrydayhustling 4y ago
Good thought - people with "high variance" tastes might buy an unusual number of duds, but they might also be more likely to pick breakout products. An analysis for that would need to include degrees of product success, not just survival.
mrandish 4y ago
This was my reaction as well. There are two different ways to be an outlier in prefering products more likely to be discontinued. Very early adopters who have uniquely discerning selection criteria or tastes may seek out new products which meet their preferences but if the product is "too early" to capture much market share, it is likely to fail. When their 'outside the norm' preferences end up catching on with the broader market they look like trendsetters or taste makers. When their atypical preferences don't catch on, they look like the "harbingers" discussed in the article. I suspect the reality is that both types exist. The article says harbingers tend to clip coupons more. This implies they may be price-sensitive, lower income or perhaps trying to stretch their lifestyle beyond their means. To me, harbingers sound like people who are open to trying novel things but also tend to make quicker, perhaps ill-considered decisions. Over time this would lead to a trend of making poorer choices on average, such as remaining in zip codes where the negatives might outweigh the positives but not obviously so at first glance - explaining another finding of the study. I've occasionally had strong preferences for product features clearly not valued as highly by the majority of the market. That means sometimes I have to seek out or pay more for the items I prefer. However, I'm not at all price sensitive when it comes to products that differentiate themselves in ways which are important to me. Also, I almost never clip coupons. Conversely, when I judge a cheaper product to be equivalent in the ways that matter to me, I'll go for the cheaper one - all other things equal (ie I wouldn't drive to another store to save 75 cents). Other times, products with features I find uniquely valuable fail in the market because there aren't enough others who value those features. But I also have a large number of examples where I was very early to a product niche which was not a 'thing' but then grew into a sub-category popular enough to be sustainable and competitively profitable. An easy example is mechanical keyboards. I was one of those people seeking out old IBM keyboards and adapting the old big DIN connectors to USB. A decade later I was participating in group buys to get good mechanical keybs. Now, another decade on, I can walk into Best Buy and find three or four mechanical keyboards. I'm unusual in that I'll pay >$500 for the exact keyboard I want yet seek out a very particular brand & model of shirt or pants I prefer from Costco or Wal-Mart despite being the cheapest, apparently undifferentiated offering. But in those cases, it's always because of specific features that item uniquely possesses over others. When friends ask me about some of my less conventional product choices I can explain in detail exactly why this one is objectively better to me than the others. In many cases, following such conversations my friends end up adopting either the product or at least the previously under-valued differentiating feature. A personal example of the negative outcome (failed product) would be the Zazzle, a cool toy which ultimately failed in the market despite my buying one when I saw it at retail. My Zazzle is sitting on the bar in my game room and once discovered by friends becomes a favorite. One friend told me they tried to get one on eBay only to find them selling for hundreds of dollars. I have a higher than usual rate of having discontinued products I bought when new which are now selling for multiples of original MSRP. I didn't buy them because I thought they'd go up in value, I just tend to buy stuff I want if I see unique value in it.

smcnally 4y ago

A corollary to this is when a single brand and formulation become multiple brands and formulations, but the original disappears. e.g. Triscuits, Cheerios, Cranberry Juice. I can find * Tzatziki Quinoa Basil and Garlic Triscuits, but not Original or Low-Sodium. * Apple Honey Nut Ancho Kobe Beef Cheerios, but not Cheerios. * Mango Chutney Cranberry with Turmeric & Essence of Nightshade juice, but not Cranberry see also: potato chips, hummus, frozen waffles

AlbertCory 4y ago
Funny because I have both Original Triscuits and regular Cheerios in the house right now. Panic! Maybe I shop at a Harbinger Store [tm].