Harbinger households: neighborhoods that reliably buy products that fail
December 12, 2019 8:50 AM   Subscribe

In The Surprising Breadth of Harbingers of Failure (Sci-Hub mirror), a trio of economists and business-school profs build on a 2015 Journal of Marketing Research paper that claimed that some households' purchasing preferences are a reliable indicator of which products will fail -- that is, if households in a certain ZIP code like a product, it will probably not succeed. The original paper calls these "harbinger households."
posted by Etrigan (60 comments total) 27 users marked this as a favorite
 
Finally, scientific research that shows why products I like are always quickly discontinued after I discover them.
posted by Ghidorah at 8:56 AM on December 12, 2019 [35 favorites]


I admit, this has the kind of tidiness that makes me think it will turn out in a few years to have been massaged pseudoscience. Based on nothing specific about the evidence, it just has that... smell of bad social psychology.
posted by tavella at 8:57 AM on December 12, 2019 [38 favorites]


Finally, science to study how my dad still talks about how much he loved OS Warp.
posted by Bulgaroktonos at 9:01 AM on December 12, 2019 [24 favorites]


My parents have always claimed that the more they like a show, the more likely it is to be canceled before the first season is over. I always thought it was confirmation bias, but ... huh.
posted by solotoro at 9:02 AM on December 12, 2019 [7 favorites]


it just has that... smell of bad social psychology

Not to mention perpetuating economic dependence on perpetual waste, ie, products that are made to last fail and products that are disposable succeed because people shop based on price.

In the short term though, some consultants are going to rake in a lot of moolah.
posted by CynicalKnight at 9:11 AM on December 12, 2019


The paper says that "harbinger households" tend to be headed by older, white, less-educated suburban people. A lot of these quickly-discontinued products are likely heavily marketed and then heavily discounted. So: people who buy cheap stuff that is heavily advertised and maybe don't put a lot of thought into what they buy. And, being in the suburbs which are dominated by big-box stores, the variety of choice is limited. We know that suburban white people cluster, so this is not a surprise.

I think it is maybe a distinctly different phenomenon than being a huge fan of The O.A..
posted by grumpybear69 at 9:13 AM on December 12, 2019 [23 favorites]


If you have enough data, you're going to find a 'pattern' in there that's there purely by chance.
posted by entropone at 9:16 AM on December 12, 2019 [2 favorites]


My parents have always claimed that the more they like a show, the more likely it is to be canceled before the first season is over.

Were your parents of an age to have watched You're In The Picture when it aired? That seems like a superhero origin story to explain their ability.
posted by ricochet biscuit at 9:16 AM on December 12, 2019 [3 favorites]


I was curious about these "harbinger Zip codes" and was imagining them as being full of rich old eccentrics or something. But that is apparently not the case:
Households in harbinger zip codes are relatively less educated (they are less likely to have graduated with bachelor’s degrees). They also tend to have proportionately larger white populations, with fewer African Americans, Asians, or Hispanics. We also see evidence that they are more deal sensitive, purchasing a higher proportion of items with coupons, using coupons that tend to offer larger discounts, and buying items that have a higher average unit price.
I wonder why products these people seem to like tend to fail, and if it's because they're not considered a really desirable demographic? I.e. if a company finds that a particular product is only popular among non-college-educated whites, in what I suspect are fairly rural or not economically booming areas*, does that mean they think there's no further growth potential?

* Based on the statement that "during periods in which house prices were generally increasing, harbinger zip codes experienced smaller house price increases than other zip codes"
posted by Kadin2048 at 9:19 AM on December 12, 2019 [7 favorites]


I wonder why products these people seem to like tend to fail, and if it's because they're not considered a really desirable demographic?

I believe it's more that the companies know a product isn't doing well, discount it, and these folks buy it because 1) the advertising makes it look good, and 2) it's cheaper. But the reason it's cheaper is because it's on the way out, and the folks are less clued in to what currently the successful products.

Thus, I think this is more using the aggregate sales patterns of a group to indicate product failure because those folks are motivated to purchase by the sort of marketing that gets used on failing products.

It's more a fascinating strong correlation than causality.
posted by c0nsumer at 9:25 AM on December 12, 2019 [28 favorites]


I also suspect that new items that are more out there are also likely to be heavily couponed to push trial - so if these do line up with heavy-couponners, then I think that's your correlation.

Meanwhile the boingboing writeup is overflowing with the typo "habringers" which is giving me a lot of joy for some reason.
posted by Mchelly at 9:30 AM on December 12, 2019 [5 favorites]


I do wonder how the “tends to vote more often for losing candidates” fits in.

I’m sure it’s just because I’ve been watching it a lot lately, but I have the theme song from Parks and Recreation running through my head.
posted by hurdy gurdy girl at 9:36 AM on December 12, 2019 [1 favorite]


Meanwhile the boingboing writeup is overflowing with the typo "habringers"

It's not like hamdinger is hard to spell
posted by GCU Sweet and Full of Grace at 9:37 AM on December 12, 2019 [21 favorites]


I do wonder how the “tends to vote more often for losing candidates” fits in.

They buy items that don't have broad general appeal only because they've been incentivized to do so (i.e. with coupons).

They vote for candidates that don't have broad general appeal only because they've been incentivized to do so (i.e. disproportionate attention is given to this demographic because they're "Real Americans.")
posted by explosion at 9:39 AM on December 12, 2019 [6 favorites]


Reading the research now and this jumped out:
To identify harbinger zip codes, we use data provided by a mass merchandise store. For confidentiality reasons we cannot identify which store, but for ease of exposition we refer to it as “MassStore.” The retailer sells a broad range of products including perishables, sundries, and durables. When customers purchase at this retailer they must identify themselves with their membership card, which allows us to link each transaction to each customer.
Emphasis mine. So -- Sam's Club, Costco and BJ's immediately come to mind. I know these stores send out coupon books. I realize the rest of the research (via "ApparelCo") is meant to disprove the argument "Maybe people who are keeping an eye on expenses through club store purchases are naturally more sensitive to coupon offers and are more motivated by price than by market popularity." And now I'm sitting here with the realization that the rest of the paper is dedicated to the argument that no, some shoppers are just going to consistently back low- to no-growth-potential products and that approach also affects their political engagement and their ability to use their house as a source of wealth, equity or generational.
posted by sobell at 9:44 AM on December 12, 2019 [4 favorites]


I wonder if there is a strong presence of dollar stores in harbinger zip codes changing the perception of what's an acceptable level of product quality vs value? Is there an Overton window of crappiness?

As for pop culture, I've joked for a long time that I'm an anti-bellwether for what's popular, but I think that's a totally separate axis from buying or judging product quality.
posted by BrotherCaine at 9:55 AM on December 12, 2019


Were your parents of an age to have watched You're In The Picture when it aired?

10 points to ricochet biscuit for this reference.
posted by briank at 9:56 AM on December 12, 2019 [1 favorite]


I know that as my household income has increased (and we've become less sensitive to discounts) my tendency to buy brands that are high quality and will be around for a while has gone way up.

Instead of buying whatever Android was most heavily discounted when my contract was up, I look for the phone (e.g. a higher end iPhone) that I want, and that I know I'll use for a long time, not just the phone I can afford.

As my leisure time has also increased (more or less in stride with income and career success) the appeal of niche items (in my case niche items were often for the kitchen) has gone way down. I now prefer high-quality widely-useful items. They take a bit more time to use, but baking, etc. is something I can take the time to enjoy even when it's labor intensive. (Because my one job more than covers my needs and I don't spend most of my waking time working or commuting.)

I suspect this study is less about the decision making perspective of people in the households and more about the sorts of people in situations to be affected by opportunistic marketing.
posted by oddman at 9:59 AM on December 12, 2019 [12 favorites]


So -- Sam's Club, Costco and BJ's immediately come to mind

This seems like a major flaw in this study, don't most of these stores usually just have one type of product (toothpaste, detergent, etc.) and so there's often no choice in the matter?
posted by jeremias at 10:06 AM on December 12, 2019 [1 favorite]


Costco stopped having coupons years ago, although they do advertise discounted products. I don't think the demographic fully fits with Costco given it's the higher end of those stores and the merchandise is better quality. But then again, my prez candidate of choice did lose last time.
posted by waving at 10:14 AM on December 12, 2019 [1 favorite]


The article describes data as coming from a single retailer, so dollar stores have nothing to do with this.
posted by PMdixon at 10:15 AM on December 12, 2019


I believe in this phenomenon, if only because I'm so good at picking soft drinks that go off the market.

I suspect that actually I just crave novelty, so when I see a tst product, if it's not absurdly priced, I'll try it. And then it turns out I really like Crystal Pepsi, and 1893, or whatever. But they were always intended to be niche, or temporary.

I would swear that I've done this same thing with snack food, but the particulars elude me at the moment. So it might just be colas that I win at picking failures.
posted by Zudz at 10:21 AM on December 12, 2019 [1 favorite]


It may not be science but I know a few people who are reliable anti-canaries for technology/devices : if they love something it’s doomed to be a niche micro interest, if they hate something its going to be the biggest seller on earth. Real anti-Cayce Pollards. I’ve gotten pretty good bets on where the future is not going by tracking what they’re really, really into.

Also there are a few pundits with whom you could get a better then 50% bet on how politics are going to shake out by just predicting the opposite of what they do,
posted by The Whelk at 10:32 AM on December 12, 2019 [4 favorites]


Finally, science to study how my dad still talks about how much he loved OS Warp.


Finally, science that can prove whether or not my mom is right to still be saying "But you wanted to buy a Betamax" to my dad 30+ years later.
posted by MCMikeNamara at 10:35 AM on December 12, 2019 [4 favorites]


Is there somewhere that they show their work on the actual products that they analyzed? They list off Diet Crystal Pepsi (1992) and Colgate frozen meals (1983) as examples, but those can't be part of the real dataset.
posted by AgentRocket at 10:40 AM on December 12, 2019 [2 favorites]


Neil Steinberg’s book has many delightful examples. Reddi-Bacon?
posted by Ideefixe at 10:43 AM on December 12, 2019 [1 favorite]


I can only read the word harbinger so many times before it loses all meaning. The third paragraph of the second link does it.
posted by Splunge at 10:49 AM on December 12, 2019


I admit, this has the kind of tidiness that makes me think it will turn out in a few years to have been massaged pseudoscience. Based on nothing specific about the evidence, it just has that... smell of bad social psychology.

Uh, was that supposed to be ironic?

I haven't read the 2015 paper, but I'm a bit curious about the off-the-cuff dismissal of a peer-reviewed 2019 paper that appears to be an attempt at large-scale real data conceptual replication of a 2015 idea? From the paper:

"Collectively, the results reveal that harbinger tendencies are correlated across a wide range of decisions, including purchasing decisions, political decisions, and housing decisions. We do not know what characteristics of customers’ decision making
contribute to these tendencies."

Also
"In this article, we do not aim to evaluate why the harbinger effect occurs. Instead, our focus is empirical."

So if I read correctly, they're open to unobserved underlying factors that are creating the emergent correlations, and specifically not putting forward a causal argument. I haven't delved into the modeling, but this appears to be a reasonable approach to empirical analysis of behavior patterns related to an interesting set of questions with authors showing an appropriate amount of caution in representing their results. Am I missing something?
posted by BlueBlueElectricBlue at 10:57 AM on December 12, 2019 [7 favorites]


Some years I’ve harbunged more steadily than others. I can imagine a couple of causes - people so unfashionable that they damage corporate support for a product, people who dislike liking what everyone else likes, people easy to market junk to the way c0nsumer says.

I’m getting a narrative picture of a difficult uncle who is really easy to scam because he’s so convinced he has to scam everyone else.
posted by clew at 10:58 AM on December 12, 2019 [4 favorites]


I can only read the word harbinger so many times before it loses all meaning. The third paragraph of the second link does it.

If you've played through the various Mass Effectses it goes from zero to shouting silly quotes in one repetition. THESE ZIP CODES ARE THE HARBINGER OF YOUR PERFECTION!
posted by GCU Sweet and Full of Grace at 11:02 AM on December 12, 2019 [5 favorites]


It's interesting that they're suburban and white. I wonder if part of it has to do with marketers and product developers who are also suburban and white creating products that they think will have broad market appeal but in fact only really appeal to their own demographic.
posted by L.P. Hatecraft at 11:02 AM on December 12, 2019 [4 favorites]


It might have been easier to just plot out where Zunes were sold.
posted by sjswitzer at 11:08 AM on December 12, 2019 [12 favorites]


So, people with limited disposable income, who are likely to watch ads on TV and other media that is targeted to them, and other people around them who are similar socioeconomically, purchase products that are ostensibly designed to have a higher failure rate due to quality of construction and/or novelty? Seems like you could almost make a self-fulfilling prophecy out of that correlation...
posted by Chuffy at 11:22 AM on December 12, 2019


I believe these are products that fail in the market, not in use. Lots of the latter are market-successful.

What’s the PRIZM name for harbinger zips? I skimmed the paper for PRIZM but didn’t see a mention.
posted by clew at 11:27 AM on December 12, 2019 [4 favorites]


Does anybody else wish you could type in your zipcode to see if your in a suspected harbinger zone?
posted by oddman at 11:30 AM on December 12, 2019 [16 favorites]


My Windows Phone is blushing.
posted by storybored at 12:06 PM on December 12, 2019 [7 favorites]


The poor, welfare reliant household that I grew up in was surrounded by family members and neighbors seemed to always have fad related products around and living rooms centered around the television. My mother was a contractor and we were always in the "nice" neighborhoods. Aka, "rich people houses". They had and fresh fruit in a bowl and television sets that were tastefully hidden or rarely used. I can tell you exactly what contributes to this if ya'll let me. Stupid television. My mother used to tell us that it made people exhibit bovine behavior.
posted by lextex at 12:08 PM on December 12, 2019 [1 favorite]


What’s the PRIZM name for harbinger zips? I skimmed the paper for PRIZM but didn’t see a mention.

Checking the Claritas listings according to Wikipedia now and there are so many more of them than I remember. They seem to be working to slice and dice the U.S. middle and working classes even more finely than before. (Their press release center reads like a primer on where marketers will be turning their attention; wait to see these percolate out to press coverage not long after they're released.)

If I had to guess, I'd bet some overlap in the harbinger counties and:

"Bedrock America: This segment is made up of families in small, isolated towns located throughout the nation's heartland. They have modest educations, sprawling families, and service jobs."

"Crossroad Villagers: With a population of retired seniors, residents are high-school educated, with modest housing, and live small-town lifestyles. They also enjoy the occasional dinner out."

"Red, White & Blue: Middle-aged, with high school educations, many of these folks are transitioning from blue-collar jobs to the service industry. In their spare time, they are active members of their local community organizations."

"Simple Pleasures: Many of its residents are over the age of 65 years old and mostly retired. Many are high school-educated seniors who held blue-collar jobs before their retirement."

"Heartlanders: This widespread segment consists of mostly retired older couples living in communities of small families. These empty-nesting couples pursue a rustic lifestyle and have many leisure activities."

" Domestic Duos: Made up of over-65 singles and married couples living in older suburban and second city homes. They tend to maintain an easy-going, predictable lifestyle often socializing with friends."

"Kid Country, USA: Widely scattered throughout the nation's heartland this segment is dominated by hard working families living in small towns."

I'm sure there are others I'm not immediately grasping. Esri's another fun one to explore and they let you see how your zip code breaks down by marketing segment.
posted by sobell at 12:34 PM on December 12, 2019 [10 favorites]


They don't really put Sams Clubs (and especially not Costcos) in middle-income middle-education suburban neighborhoods. So harbingers must be travellin' to different zip codes to get their discount wares.
posted by The_Vegetables at 12:38 PM on December 12, 2019 [1 favorite]


Thanks, sobell.

Are the Claritas euphemisms clear to everybody else? What do they mean by ‘technology use’ ? Streaming? Video games? New cellphones?

The papers cited by the original paper are another Snow Queen’s mirror if you’re unfamiliar with market segmentation.
posted by clew at 12:59 PM on December 12, 2019


Bear in mind when I floated the idea that the retailer they were looking at was Costco, BJ's or Sam's Club, it was based solely on the mention that shoppers pay an annual fee to shop there and are required to show a card. There may be other retailers who also require that with whom I'm not familiar!

In the study itself in re: the retailer activity, the authors mention they analyzed the following data sets they got from the mass retailer:

1. retail transaction data that identifies the customer, store, date and time of purchase, number of units purchased and price paid.

2. a complete record of coupons used in the customer transactions included in the transaction data. (" However, we cannot match coupons with specific items. As a result, we use the coupons to measure customers’ deal sensitivity, but we cannot use them to measure how often a product was purchased at a discount.")

3. customer containing demographic data for each customer, together with the five-digit zip code from each customer’s residential address; this is augmented by census data that describes the population density, ethnic composition, and average education attainment levels by zip code.

The data analysis of product sales was, if I'm reading correctly, done by sorting product sales by the product's "success" (i.e. longevity), then "we classify the zip codes into groups by focusing on purchases of the 2,324 new items in the classification category in the first 90 days after each new item is introduced. We calculate the proportion of new items purchased by customers in each zip code that succeeded (had sales after 18 months). To construct this zip code–level average, we weight by the number of transactions for each new item. We then rank the zip codes based on this average and organize them into groups based on quartiles of the average success rate."

So if I'm reading this correctly, they're looking at resident zips and not store zips. There is, as of right now, no geographic component mentioned ("harbinger residents drive an average of five miles to MassRetailer ...") but I wouldn't be surprised if one is factored into whatever private model is being constructed to figure out how to maximize sales/reduce losses based on harbinger behavior.
posted by sobell at 1:03 PM on December 12, 2019 [3 favorites]


Sorry to double-post --

clew, you asked, "Are the Claritas euphemisms clear to everybody else? What do they mean by ‘technology use’ ? Streaming? Video games? New cellphones?" ... I just downloaded a copy of the 2016 categories (interesting because it sifts on the axes of urban density and income) and how technology use is described seems extremely income-based.

For example, "Young and Rustic" is described thusly:
Young & Rustic is composed of restless singles and young families in the
nation's rural areas. They enjoy the outdoors on their ATVs but are also
big video gamers and follow NASCAR and monster trucks.
Owns a Ford • Eats at Hardee's • Spends 5+ hours a day on social network sites • Follows
monster trucks • Stays at Days Inn • Watches Boomerang • Listens to NASCAR
Contrast that with the "Connected Bohemians"
A collection of mobile urbanites, Connected Bohemians represent the
nation's most liberal lifestyles. Its residents are a progressive mix of tech
savvy, young singles, couples, and families ranging from students to
professionals. In their funky row houses and apartments, Bohemian
Mixers are the early adopters who are quick to check out the latest
movie, nightclub, laptop, and microbrew.
Owns a hybrid vehicle • Eats at Starbucks • Shops at Express • Goes skiing/snowboarding •
Visits the Middle East • Uses Yelp • Listens to Alternative
I'm not sure "buys a laptop" and "uses Yelp" somehow make someone more tech savvy than a person who spends 5+ hours a day on social media and games a lot -- it seems to be tiptoing around how some types of tech activities are coded to some socioeconomic levels.
posted by sobell at 1:14 PM on December 12, 2019 [7 favorites]


‘Comfortable with Excel’ ?
posted by clew at 1:32 PM on December 12, 2019 [1 favorite]


(My guess at what they’re tracking, not a phrase I found)
posted by clew at 1:33 PM on December 12, 2019


The description made me think of "Mama's Family", and all the Raytown-likes I've known. People sort of alienated from mainstream society, all up in each others' shit all the time; decisions become a weird kind of statement, often like, frugality signaling.
posted by fleacircus at 2:35 PM on December 12, 2019 [2 favorites]


WHICH. ZIP. CODES. GOD. DAMMIT.?
posted by soundguy99 at 2:48 PM on December 12, 2019 [5 favorites]


Colgate frozen meals (1983)

If only they'd had any flavour except spearmint …
posted by scruss at 2:50 PM on December 12, 2019 [6 favorites]


This is interesting stuff. I recognize people I've known in that description and it is interesting to think of them as the canaries in the coalmine of doomed products.
posted by Dip Flash at 5:31 PM on December 12, 2019 [2 favorites]


This sentence jumped out at me:

We see clear evidence that when households in harbinger
zip codes move, they move to another harbinger zip code.


Fascinating. Canaries move from one canary neighborhood to another. How do they spot other canaries?

This study is probably analyzing a specific socio-economic group that the study is being coy about.
posted by toastedbeagle at 5:58 PM on December 12, 2019 [7 favorites]


I mean, Bulgaroktonos, OS/2 Warp was pretty awesome.
posted by wintermind at 6:44 PM on December 12, 2019 [2 favorites]


Sorry to jump in to the thread so late. I was pounding-back a couple of Zimas, rocking out with my Zune, and I guess I just lost track of time...
posted by Thorzdad at 7:00 PM on December 12, 2019 [8 favorites]


Meanwhile the boingboing writeup is overflowing with the typo "habringers"

Yeah, we've covered this glorious word before.
posted by ricochet biscuit at 7:44 PM on December 12, 2019 [5 favorites]


Canaries move from one canary neighborhood to another. How do they spot other canaries?

I suspect that they're more driven by avoiding non-canaries, if you know what I mean.
posted by praemunire at 9:54 PM on December 12, 2019 [3 favorites]


Canaries move from one canary neighborhood to another. How do they spot other canaries?

I suspect that they're more driven by avoiding non-canaries, if you know what I mean.


That, and selecting a neighborhood the same way they are buying products, based on marketing and discounts. Racial or cultural biases should fit neatly into the mix.
posted by Dip Flash at 10:27 PM on December 12, 2019 [2 favorites]



Bear in mind when I floated the idea that the retailer they were looking at was Costco, BJ's or Sam's Club, it was based solely on the mention that shoppers pay an annual fee to shop there and are required to show a card. There may be other retailers who also require that with whom I'm not familiar!


There really aren't any mass market stores unless this study is way more niche than it is letting on. Also, generally the 'whiter' a neighborhood is, the more it appreciates, especially in a suburb. If I had to guess, given what they say, to find a Harbinger zip code, search (1) growing suburbs/exurbs within 5 miles of a Sams Club (the other two demographically just don't fit) with relatively inexpensive (comparatively) older homes. The Harbingers are buying the older homes as the former residents trade up into newer homes.
posted by The_Vegetables at 7:45 AM on December 13, 2019 [2 favorites]


I keep gnawing at this paper because one of the (most effective) things I learned in an undergrad lab class was to read any academic research paper with the question, “What were they really hoping to learn, and what did they turn that into?”

This is a lot of work to identify a consumer class that effectively acts as an anti-bellwether for successful products and politicians. Was the point originally to figure out how to sell to these “real” (i.e. white and non-college-educated) Americans? And the discovery is that surprise, surprise, it’s not as lucrative in the long run as it is to focus on other groups?

(Sidebar: I don’t watch Fox News so I don’t know if this was covered there at all but I can’t believe some enterprising pundit hasn’t read this research and concluded that “academia and elite retailers are labeling real Americans losers,” or some such thing to keep the White Entitled Grievance engine stoked.)
posted by sobell at 8:17 AM on December 13, 2019 [1 favorite]


From the study, what if that line about where they got the consumer data is just missing a clarifying parenthetical:

When customers in this study (but not necessarily all customers of MassStore) purchase at this retailer they must identify themselves with their membership card, which allows us to link each transaction to each customer.

That would widen the scope to include any stores that have optional rewards memberships, like Kroger/Smith’s, Target, Best Buy, etc... it would make the whole study slightly more suspect in my opinion, so
posted by Doleful Creature at 11:44 AM on December 13, 2019 [3 favorites]


Sorry to jump in to the thread so late. I was pounding-back a couple of Zimas, rocking out with my Zune, and I guess I just lost track of time...

What is your zip code, Thorzdad?
posted by Dotty at 11:47 AM on December 13, 2019 [1 favorite]


I suspect that actually I just crave novelty, so when I see a tst product, if it's not absurdly priced, I'll try it. And then it turns out I really like Crystal Pepsi, and 1893, or whatever. But they were always intended to be niche, or temporary.

This is me but I find the observation that "novel things I like get discontinued" makes me less likely to try new things lest I really like it and be sad when one can no longer buy it. I'm looking at you grape Schweppes ginger ale; which I now find out is still available regionally just not my region :(
posted by Mitheral at 11:03 PM on December 13, 2019


For what it's worth, I worked at the agency who did the advertising for Miller Clear Beer. I still have a full sixpack (in bottles) in the back of a closet. I would love to believe that this is a Collectable and therefore Worth Money to someone out there, but I suspect that it was so much of a failure that no one would even want it now. But I'm still holding on to it...

Meanwhile, since no one else has linked to it yet, Crystal Gravy
posted by Mchelly at 4:52 PM on December 14, 2019 [2 favorites]


« Older 44,000-Year-Old Indonesian Cave Painting Is...   |   No ancient emperor ever lived so well Newer »


This thread has been archived and is closed to new comments