Always think in absolute dollars and not in rates. $1000 is $1000 whether the cost per use is pennies or tens of dollars. Thinking in terms of rates is how people would fall for predatory installment purchases in the old days (you can have a new computer for just $2.5 a day!)
The other obvious problem is looking at cost absent of a market. It's not just how much it is worth it to you, but what the competition is offering it for. I personally pay less than $400 a phone and keep it for over three years. Paying $1000 for a phone just two years is not an improvement.
Edit: I spoke very dogmatically above. Using rates can be OK in certain situations where the analysis becomes equivalent to using absolutes, but it's not at all rare for rate analysis and absolute analysis to differ, and people make the mistake of using rates there. You will never, ever, go wrong by looking at absolute amounts, though - and it's usually the easier way to analyze.
In my engineering undergrad, Engineering Economics was a required course. I remember a fellow student said "I often return other engineering textbooks to get some cash, but this book is more useful than most of my engineering ones." Decades later, I can say he was definitely right. All too often I see engineers making poor financial decisions based on heuristics when they're quite capable in doing the (simple) math. But a lifetime of being exposed to flawed ways of thinking of money provides a convenient shortcut.
So my friends call me the cheapest person they know, but I also think like the author. If there is something I am going to use a lot and the added value helps me, then I'll shell out more for it. Really we're talking about Boots Theory[0]. If you buy cheap rubbish then you'll buy frequently. If you buy quality you may only have to buy once a lifetime, even if it is 3-5x. With your phone, if all you care about is texting and making calls, yeah, you're not getting any added value paying more. If you're a person that takes a lot of pictures and values the camera, you will get added value (don't come at me with the "buy a dedicated camera" because you're not carrying that in your pocket everywhere and thus can't capture the same moments).
So if you wear boots every day, it's better to shell out for the boots that will last you a lifetime rather than ones that will last you a season.
If all you care about is texting, making calls, playing the occasional game[1], browsing the web, doing mobile banking, listening to music, watching videos, using social media, responding to work emails, viewing the occasional PDF, word processed document or spreadsheet sent over email, screencasting to a TV, using the device as a mobile hotspot, testing software you write for phones ... and a few more less frequent things (Like using employers paging app, employers internal systems) ... then you don't get any additional value spending more than $400 on a phone.
Don't ask me what you can do with a phone that costs more, because I haven't yet seen anyone do something on their phone that my <$400 phone cannot do.
[1] Wordle, for example, is insanely popular right now.
If that's all you care about, paying $400 like I do is silly - you can get it much cheaper.
The $400 phone I get takes fairly good photos (it's one of my main criteria in deciding on phones). Sure, if I want great photos, I'll have to pay more. I've decided that an extra $1100 over 3 years is not worth it for the marginal improvements in photo quality. But I know for some people it is, and that's fine.
Regarding the boots, we're both saying the same thing, which I think is not what the author is saying: You're fixing a time interval (e.g. 30 years), and calculating that it's cheaper to get the expensive boots vs the cumulative cost of cheap shoes in that time period.
When you fix the time interval (e.g. 30 years), comparing rates vs comparing absolute totals is equivalent.[3] What I typically see is that people take this equivalence and begin to compare against different time intervals (option X is for 2 years, option Y is for 5 years). This is what leads to so many silly blog posts saying you'll never get a better investment than your employer's ESPP benefit, because you get a huge return.[1][2]
In any case, the author isn't using the Boots Theory. He's looking at cost per use - and not utility/cost as a whole.
[1] https://thefinancebuff.com/employee-stock-purchase-plan-espp...
[2] I'm pro ESPP - but looking at those claimed return numbers is almost useless. It's trivial to make more money with, say, a mere 5% annualized return.
[3] When comparing investments, you also need the same input. It starts getting messy - which is why the standard advice is to convert to absolute dollars and compare.
Technically, you should think in probability weighted log-dollars to maximise your economy over the long term. (The Kelly criterion.)
In some cases this is equivalent to thinking in terms of fractions of your total wealth -- this is how the Kelly criterion is popularly presented.
For small ("everyday", as Bernoulli put it) amounts, this is equivalent to absolute dollars. So you're right, but it's a special case of a more general principle.
* $/use isn’t really a rate in the way you’re describing with predatory financing. It’s not a cost hiding measure but one of many ways of estimating utility. Optimizing for price alone and not $/utility will bite you for lots of goods.
* A $400 phone used 100 times is $4/use and a $1000 phone used 100 times is $10/use so markets still have plenty of an effect.
E.g. I should start a new hobby because I am definitely going to keep at if for 5 years! And then you don't.
I have a budget of 500€ for sports per year. I use this on gyms, running shoes and what not.
I wanted to get into skiing so I bought some skis this year. It was about 700€.
Normally this doesn’t fit my budget. But I’m gonna keep these skis for at least 4 years.
Now the price is 175€/year. And I still have a bigger portion of my budget left for other sports.
Or SaaS.
It's essentially a reworking of the "Boots" theory or poverty. Cheap things will typically wear out faster and need to replaced often.
Like in your phone example. You actually based and compared them on rate. $1000/2yrs is worse than $400/3yrs. Hell, even $1000/3yrs is worse than $400/2yrs.
That $2.50/day computer is going to wind up being $2.50 * number of days owned when it is finally replaced. It will take 800 days to match the value if the computer costs $2000 initially. If you keep that computer longer than 2 years, 2 months, and 9 days, you're now paying more than the value of the computer.
I think the author's suggestion would be to just pay the $2000 up front. It's a lot of money right now, but if you keep the machine for 3 years, that's like $1.83 per day you paid for the computer.
An example of where looking at the cost per use makes sense is a gym membership. --If I consider the gym is worth, say, $5/visit and it costs $50/month, then I know I have to go at least 10 times a month to feel like I'm getting what I'm paying for.
On the other hand, yeah, my phone is not something that I'm going to try and figure out a cost per use, not the least because each use has a different intrinsic value to me (contacting someone is more valuable than checking the news, at least for me).
This is also dangerously close to going backwards about it. Figuring out how much you need to do something purely to "get your money's worth" out of it is the sunk cost fallacy!
It's better to start from how many times you want to go to the gym, multiply it with your desired cost per visit, and then see if this total exceeds the cost of the membership.
I know that's what you were suggesting, but the way you presented it seems to me, well, backwards!
It depends on the opportunity cost. If you can get a house for 3600 a month, and the house appreciates in value in excess of the discount rate, then it's worth it to purchase that house.
To add to this, there is also the opportunity cost of buying the more expensive option because the difference is only $1/day. Spending $365/year on the more expensive option is missing out on getting some other item[1] for $365/year.
Like I keep telling people who ask me why I don't pay for $FOO instead of using $BAR (where $FOO is "Windows", or "Office365", or some development tool and $BAR is OpenOffice, Vim/Emacs, git from the cli, Linux, etc).
The answer to any "Well $FOO only costs you $10/m, and if you are a professional using it as part of your profession, then you should be able to afford it" is "I can afford lots of things, doesn't mean I am necessarily going to buy them".
[1] Sticking with the phone example, taking a phone that works out to be cheaper by $365/year means that you could spend the extra money for redundancy (if your phone breaks, simply buy a new one in 30m), you get a drawing tablet (suddenly, whiteboarding during zoom meetings saves 30m - what's your hourly rate?), or you put it into a index tracker or stock or debt, and get a positive return.
In this case, the absolute value analysis yields the same result, so it doesn't really "depend". I'm not saying you will always go wrong thinking in terms of rates. I'm saying you'll never go wrong thinking in terms of absolute values.
I think any introductory textbook on Managerial Economics would cover the same material. It's just that Engineering Economics books tend to use industrial plants and machinery for most of their examples.
This one[1] is probably decent (looking at its table of contents):
https://www.amazon.com/Engineering-Economic-Analysis-Donald-...
It's not a sophisticated topic. If you're a non-CS engineer who knows calculus and algebra well, this is one of the easiest topics in the curriculum.
Even if your calculus is poor, simply using the concepts with a spreadsheet will get you almost all the way.
My income is a rate so I find normalizing on time to be pretty useful. I think $/use is pretty similar to normalizing on time
Simple: $10K is greater than $8K. Analysis is complete.
Is 10% better than 8%? Not always. It depends on how much you put in and what the duration for those investments are. This is trivially true, yet so many people ignore this aspect.[1] People also get confused between simple and compound rates.
In a lot of real world investment situations, the inputs and timelines aren't the same. In real estate there are plenty of investments that require a minimum investment and have a fixed timeline (cannot withdraw in that time period). Quickly: Is a $50K investment at 8% compound rate over 7 years better than a $40K investment at 10% simple rate over 5 years?
There are also investment strategies that require periods of holding cash. It's not trivial to compare those with, say, an S&P 500 investment using only rates.
The standard taught in the engineering economics classes:
1. Fix your time interval to whatever is relevant for your situation.
2. Compute the cash you will have at the end of that time interval (i.e. absolute amount)
3. Go with the higher one.
[1] Yes - even in the absolute analysis, you should have the same timeline, and technically it's equivalent to a rate comparison in that sense. But I've found that when people do the absolute analysis, they never seem to forget to account for the timelines being the same.
Either you aren't using your phone much or you are needlessly suffering with inferior product. Wearing cheap shoes for years when you could afford comfortable high quality shoes is not a virtue. Same goes with tools we use.
I have not. What has happened? Have SUVs stopped being the most popular class of passenger cars being sold?
But if you're comparing using a computer for 5 years vs eating cheeseburgers over 10 years, using rates become more challenging, and is prone to faulty analyses.
My point, though, is that cost per use is not usually a helpful metric.
I've also estimated my cost to commute to my tech job in the suburbs (thankfully full-time remote now), and it was staggering how much that was including insurance, maintenance, tolls, electricity, etc., compared to free options/transit/corporate bus/remote work/my daily mortgage cost.
Cost-per-use by itself is fairly meaningless, costs needs to be compared against value-per-use.
What is the worth of: a visit to my friends; taking a surfboard to the beach; taking a load of firewood to my parents; an emergency 2AM visit to a suicidal acquaintance; taking my nephew and his friends to the skatepark; helping a friend move; picking up a Craigslist chair.
Other transport options usually have severe limitations on availability, flexibility, cost, destination, load, etcetera. Optionality and peace-of-mind has huge value too (i.e. availability if urgently required, even if it happens that nothing urgent comes up).
Last time I mentioned on HN how useful a car is, someone suggested hiring a car instead, which is ridiculously more expensive and extremely inconvenient (at least where I live).
Disclaimer: I am carless at present. The non-financial costs are extremely high.
In urban Europe (the one I'm familiar with, at least), I find public transit, then renting cars/vans and straight up calling cabs/Uber when needed is still absurdly cheaper, particularly with European gas, insurance and maintenance prices, and somewhat limited parking space. Sure, the odd 100€ cab fare to urgently go somewhere far hurts in one shot, but according to my own numbers for where I live, just having a used car lying around "just in case" without any mileage is 304€/mo.
That implies the only alternative is not doing things? You can just order a taxi?
I think for many car users the total 'cost per use' of their car exceeds the average taxi fare they'd pay for those journeys.
Nevermind considering public transportation.
How many years will I own my car? How do you know?
Maybe some folks buy a new car on a regular basis, but others will keep a car as long as they can. The denominator isn't known for everyone here.
And then there's inflation, recession, etc. which can make the numerator unclear too.
You amortize the cost over the planned lifetime. If you keep the car for longer than planned, great, you get a free usage extension! If you need/want to replace the car sooner than planned, you need to factor the sunk cost of the previous purchase into the cost/benefit equation of your next.
Also vehicles aren't lumps of gold. They rot while sitting still, even in a garage - which fewer and fewer of us do.
They do think about this, which is why the second-hand market for cars is so healthy.
To be accurate, you should restrict your assertion to those people who only buy brand new cars, which is necessarily a far small population of people who buy second-hand.
There are more people who think about this (they buy second-hand) than people who don't.
$8K is $8K - it doesn't matter if it's all paid in one day, or spread over 10 years.[1]
I always notice how they like to look at the cost per day and not, say, the cost per year. I suppose if the daily cost still looks high they'll calculate the cost per hour. Or as in this case, the cost per use just because it makes the number even smaller.
[1] If you exclude opportunity cost. If you include it, investing $8K up front will likely give you better returns than investing $2.20 a day.
So a car may be purchased based on pure utility, or for uncommon but plausible capabilities, or for luxury or status.
You haven't lost $700 by buying a pair of skis. You have exchanged $700 in cash for at least $500 in a pair of skis that is now previously owned.
So you might lose, say, $200 immediately for the privilege of buying something brand new. And then another $100 if we include up-front the effort to sell them to someone else.
But the rest of the money is still there, in the shape of skis.
As you use the skis they will depreciate in their second hand value -- this is when the money is actually lost, bit by bit. You can estimate how much the second-hand value of the skis are with 10 uses, 100 uses, 1000 uses, and so on. This is the real cost per use.
I also don't know anything about designer tables or design furniture so I probably would get ripped off on buying first and then second time when selling.
In the end I don't have any patience for selling things so I would even low ball the price just to get rid of stuff. Heck even giving stuff away for free is sometimes annoying.
If I would plan not to move anymore or be sure to live somewhere for 20+ years I could consider designer furniture even if I would pay too much for it :).
I encountered this first in a forum discussion about boilers - someone pointed out that the largest cost component isn't even the unit itself, but all the space it takes.
Anyway whenever I think of buying a treadmill, I remind myself that the actual cost is an additional €4500 just for the real estate and perhaps a gym membership would have been cheaper, but since I still haven't gotten one, perhaps it's less important to me than I think.
Also, now I have two kids, I always think how much its going to cost my kids. Sure I could buy a $1000 phone now, or I could leave my kids $10,000 extra in the Will. :)
It’s a blessing and a curse, because it leads to wealth but can end up in some weird end states. Like having an uncomfortable lifestyle that the money-controller can tolerate, but other members of the family chafe at.
Be careful to value the compounding interest in your own happiness and the compounding value of the happiness of those around you. Through that lens, a marginal $10k (on top of an already secure retirement fund) might be worth far less than a marginal $1k investment in the happiness of yourself or your friends and family.
Money is not a goal in itself after all, it exists to further our true goals in life. In this day and age it’s all to easy to get distracted by the money game, and forget what our true goals are.
In periods when the risk-free rate is 2 %, that is what any investment costs you in addition to the principal, as Deming put it poetically, "every day, rain or shine, even on Sundays."
If you assume you use a couch 50% of days--because you travel, sit other places, etc. you're talking about over 10 years which seems sort of reasonable. And OK you can try to find something used or something at Ikea, but $5K isn't a completely extravagant number for a couch, especially if you own a house.
My view is -- start from the cost of your use and buy it even if it is on the costlier side upfront. This is also akin to buying smaller number but of high quality item so they last longer and thus the life-time cost to usage is way beneficial than buying cheap that you need to repeat frequently.
Breaking down the cost down to the number of hours use per day, and thus if a Laptop lasts 3+ years, how much is the actual cost for each hour. Now, will it "spark joy" buying a better but costlier Laptop that will last years.
It's a really good way to evaluate how much you should pay for the second purchase of a tool - also whether the price is worth the warranty.
My Bosch laser Level for example cost $150 more then the alternative, but came with a 6 year as opposed to 1 year warranty. At a total price of $600, I was buying it at $100 a year guaranteed availability. Whereas the cheaper option actually cost me $450 for only 1 guaranteed year of service.
$2k bed / 10y = $200/y.
If that’s worth it for me to upgrade then it’s simple to make the call. We spend a third of our life sleeping so having a good bed is a worthwhile investment.
One would typically try them out at the showroom - it'd be really nice if they could let you sleep a whole night on one.
Later in college, I bought a used laserjet from the university surplus sale for $20. That little printer got me through several reams of paper before the toner went out. By that point I was in the pages per penny range.
I personally don't find the concept described in this blog post very useful. I prefer to think in terms of annualized/amortized costs, especially within the perspective of money-stress-time tradeoffs. I could see others finding it helpful though.
Yeah, it's called budgeting and being poor. Some of us try to maximize this satisfaction and lifelong memories while avoiding incurring significant personal debt on a low salary. Tradeoffs and opportunity costs abound. Of course not everyone should or would be better off being cognizant of these costs and quantities, but some of us have to be by necessity.
Its not the favourite unit of measurement of the cloud salesmen for nothing.
"Come to my cloud" harks the salesman .... "look, X is only $0.000000001 a go, Y is only $0.00095 a go .... its so CHEAP".
I think we know how it all pans out. You / your company ends up with an eye-watering cloud burn rate.
Beware the double-edged sword as they say, "cost per use" is not the panacea ... apart from for the seller, perhaps.
The absolute cost is very important because it relates to budgeting and total resources.
The cost per use is interesting for making decisions between exclusive alternatives.
Should you buy a hot tub (no)? It’s not just the initial cost but also the maintenance cost (chemicals). When you divide total cost by actual use in hours it’s clear it’s a bad deal.
I often think of entertainment cost in terms of cost-per-entertainment-hour. AAA video games, for example, provide a much lower cost per entertainment hour than movies or amusement parks, for example.
I often think of cost in non monetary terms as well.
When you think about the hours required to maintain something per hour of enjoyment, a lot of things that you can afford all of a sudden seem like bad ideas (many kinds of pets, hot tubs, etc.)
So thinking about usage in a buying decision is a very helpful tool in making good decisions, although it’s in addition to looking at whether you can afford the total cost.
In that case it's you, paying thousands for the hardware, to enable others to use you and the limited time you have on this earth.
On the one hand, we domesticated some cows to get milk, in another sense the cows have enslaved the farmers and gotten them work all day to pamper them like royalty, the farmers ensure there is always food in all seasons, they're made to build the cattle shelter from the weather, they go out and hunt predators, they shovel manure, they get up at ungodly hours to milk the cows, and so forth.
So who actually domesticated whom?
What? Why would you compare a "use" of your cellphone to a really big McDonalds meal? I mean, I should stop eating because there's no way I can eat a meal for less than the cost of using a cellphone one time?
Of course one is always able to do a post-facto "price per use" calculation, it's just an extremely myopic paradigm to be thinking in. Even $/month is too small minded - that's how salesmen push you into paying way too much and financing things. But thinking $/year starts to make sense, letting you take into account things like how long upstream software is going to be supported etc.
Also yes, most everybody knows the tabs for eating/drinking out add up to a considerable amount of discretionary income. Those too are better brought into the $/mo and $/year paradigm.
Today, I look at the math, on a cost-per-day or cost-per-use angle, and I'm so happy that I went for something nice that I enjoy.
I do feel like cost-per-day amortized is probably more interesting for a lot of decision making.
We probably over-estimate how often we'll use a product or service.
The decision if I should buy a phone for $400 or $1000 should come after answering the question "Are the extra features for $600 worth something to me?" - for some people they'll be worth - e.g. having better quality photos of their family, or having better FPS on the high-end MMORPG. Others don't need that, they don't have a ROI on the extra $600, so should skip the more expensive one.
Now, going out with an old friend and spending $100 on a night out - it's a huge ROI for some, for others not, although I think the latter wouldn't have gone out at first place. The same is valid for holidays and any kind of experiences - people will have huge ROI and will go.
The one McD menu - that's where I personally will never find ROI, so can't speak about it. Even if I am starving, buying a loaf of bread, butter and some potatoes for less than $14 will give me much more ROI.
As a hiker, I have spent a lot of time calculating the cost and weight per calorie of food, and the fuel to cook it.
TCO and ROI are far more important.
For example, inkjet printers are insanely expensive per use.
Generally, the larger and more expensive the laser printer, the cheaper it is per page.
30 years from now I will hopefully retain some my experiential (expensive) travel memories but may not remember/care if I used a value or flagship phone for a given 24 month period.
A computer, per hour of utility, becomes fractions of cents, coupled with the notion that it is practically indispensable to me then it is a no brainer.
My car on the other hand is incredibly expensive per hour, and with increasing public transit and cycling near me, its utility is also in question.
I’m not sure if it was usable life, or how long the user kept it.
Somethings don’t break, they’re just not very good, and it makes sense to replace them with a more useful version. Most things just break tho.
A McDonald's meal is an expense.
A mobile device, you can argue, is an investment.
The device can return daily joy*. A meal is, special events with friends aside, a soon forgotten one-off.
* The device can also be a distraction and a time-suck. That needs to be factored in.
The two are so different that it makes no sense to put them in the same comparison.
One should compare phones within the category of other phones. Compare food with other foods.
