> There's an oft told story about Jeff Bezos pausing a meeting to call his own customer service number - and waiting over 10 minutes for an answer.
One of my jobs was at a company that had developed at unhealthy amount of bureaucracy and politics. The product barely mattered to some because they were playing internal games of grandstanding, taking credit, and building their empires.
In meetings where were supposed to be talking about product direction and priorities I would some times pull out my phone and open the app to try to demonstrate some real problem with the service. The tone of the meeting would change to panic as certain product leads would try to do anything to stop me from showing what the real product did instead of their neatly prepared slide decks that showed a much nice story for the executives. I became the enemy for showing the actual product instead of their alternate world of KPIs and charts.
I work at a (government and extreme bureaucratic) organisation that builds apps used by field engineers.
I found out SSO was broken. They had to login to every app using the same account. Twice per day because the token live was 4 hours "for security".
I found out it was because they published these apps as PWAs, making them more isolated than normal apps.
I asked the product manager and he says the issue is "with Apple and Google", not his department. When asked why he chose PWAs for the apps he said this was easier to deploy, saves them developer accounts and such.
Since I can't force him to change I found a workaround: SSO works in PWAs if you use Edge on a recent Android version on a Samsung tablet. Lucky me we had bought Samsung tablets (this was not a requirement when purchasing I looked it up, just luck).
I asked the Intune manager about this and they said the field engineers should just set Edge as default in stead of Chrome.
When trying this on a company tablet it said: "Edge disabled by X group policy". That guys' department set the policy...
After they removed this I asked why it wasn't the default browser and he said this wasn't possible. I challenged him on this by Googling the Intune manual to set the default browser.
Later they said they had raised a support ticket with Microsoft for this.
On the internal Wiki I found a document describing the problem. It was dated 11 months before I joined.
I legitimately don’t understand how companies get to this point, especially when the C-suite is full of founders (or maybe that’s worse?). I can understand how people want to make their bosses happy, and that can cascade into constant bullshitting, but at some point why doesn’t the CTO / CEO / etc. say “I’m going to go have conversations with the workers to get their perspective?”
The U.S. Nuclear Navy, for all of its many flaws, gets this right. Generally at least once a year, the head of Naval Reactors - a four-star Admiral - tours every vessel, which may include a brief underway period. During this tour, the Admiral will talk to the engine room watchstanders, with all senior leadership removed. They’ll ask how daily life is, what they find challenging or annoying, what they like, etc. There’s obviously a lot of self-filtering (though sometimes not - Navy Nukes are not known for their social graces) that occurs, and also what a junior watchstander finds annoying may just be a required part of the job, but some useful signal is gathered.
Even outside of the nuclear program, one standout example was Admiral Zumwalt, who as Chief of Naval Operations implemented 70 different changes over his tenure as a direct result of talking to sailors, all of which were designed to improve quality of life, efficiency, or communication.
About 10 years ago now, when what would eventually become Oracle Cloud Infrastructure had just done an internal launch of the first availability domain, as we got ourselves read for public launch late in the year, several senior staff and engineers had to go do a presentation and demonstration of the product to Larry Ellison.
They did trial run after trial run, made sure trying to make sure there were no bugs in the demonstration path. They nailed it, presentation went smoothly, live demonstration just worked. Provisioned a bare metal instance, had it running hosting something within minutes of launch. Larry was suitably impressed, but the thing that most impressed him was that he'd been presented with an end-to-end live demonstration. It had never occurred to any of the folks involved to do it any other way, but apparently all too often, all he ever saw was slide shows from product teams, particularly when things were several months away from public launch.
I reflect on that situation from time to time, wondering at which stage you sort of go from expecting to see live demonstrations, to slide shows. I assume it just slowly slips away from you, one at a time until you're stuck in the land of "make believe we have a good product".
The later part of that quoted passage ends with a possibly rhetorical question.
> ... When was the last time the CEO of the above company called their own customer support line?
So Bezos definitely hasn't done that in a long time. Definitely not in India. So I would say the answer ought to be: a CEO does that or has to do that only until the company becomes too big or has captured a sufficiently entrenched large slice of the market.
At one meeting to build out a new service as a next generation to a flagship AWS service that I worked on, I got to meet all the product leaders and managers.
At that meeting, I realized most of them had never used the product and see their claim to leadership role due to their the ability to manage up and down.
I use the product on my personal projects and I hated it with a passion.
In my last role (engineering side) the VP of Product objected to me using the phrase "eating your own dogfood" because it was "gross" and she always interjected and replaced it with "drink your own champagne". I countered (privately) that was a feature; it's supposed to be a little unappetizing because you're early, building empathy and getting a different perspective (to a dog, dog food is delicious!). I think the difference in perspective succinctly illustrates the schism between trying to understand the customer experience asap, and the data-driven, kpi crowd where - conveniently - you can't discover these issues until it's too late.
I’m the kind of person that just assumes customer service is going to be bad. I gird myself whenever I have to call a company and just deal with their gauntlet with patience, knowing the trick is to outlast them. It costs them money every time you call. I’ll often tell them I know that and assure them I will continue calling until the matter is resolved. It’s not fun, it’s just the way things sadly are.
My old man, however, still feels some kind of righteous indignation when he spends his hard earned money and doesn’t feel he’s getting what he paid for. He loves to give a piece of his mind to the companies that mistreat him, and he always says “And I hope my comments are being recorded for quality assurance!”
I agree with the broader point, but I'm perplexed that the author is talking about dogfooding as a "sacred practice in the tech industry" in the context of customer support. Among big tech companies, customer support usually isn't seen a part of the product. If you work at Facebook, Google, or Microsoft, you don't try to go through the non-existent consumer support channels to resolve issues with your account.
"Eat your own dog-food and smell your own farts" is just crass enough that people will actually remember and retell it. Nice work by the author's Comparative Memetics Division.
"Eat your own dog food" == experience your own product.
"Smell your own farts" == experience your entire product, including things that are typically unmentionables like customer service and billing
The thing about dogfooding is that you have to force it in many domains, and if you aren't careful it can push you towards solving issues your users don't actually have.
We've found ourselves trying to find this balance on Tritium. It's a word processor for lawyers, so has a specific narrow domain that allows us to provide a differentiated experience from Word. But if we try to use it like Word, we end up wanting generalized features that don't fit that strategy. I wrote a little about what we've come up with here: https://tritium.legal/blog/eat.
This is one of the compelling rationales for closed-source / commercial software in certain B2B SAAS domains. It seems like you just cannot adequately test the happy and sad paths from a QA perspective in FOSS unless it's (1) insanely successful or (2) a dev tool.
This describes quite well the huge advantage small companies have vs big companies.
(Motivated) people at small companies "care", and what I mean with that is they are responsible and can see a large enough portion of the customer experience that - if something is broken - they'll see the pain and try to address it.
At a big company no one cares. They of course care about their job, but their job is such a small fraction of the overall customer experience, that seeing their work having an impact on their customer is exceptionally difficult.
That's why large companies need to encode customer feedback into a system to imitate feedback cycles. Mostly in metrics.
That's a very lossy way to capture signal, and leaves a lot to be desired, but so far it doesnt seem like anyone has come up with a better system.
To me, "dogfooding" isn't using your own product out of loyalty or some nondescript principle, as the author suggests. It's specifically about using your own product during its development as a means of testing while it's in its worst phase - literally eating dog food for a while.
It's the same horse manure, when architectonauts and developers aren't responsible for the operation of their Goldberg-inventions.
Another phrase that comes to mind is: no skin in the game.
To me "unaccountability" -- or whatever naming fits better -- needs its own circle of hell.
1) I call to cancel an insurance policy on a car I sold. I'm greeted by the IVR, press three to cancel a policy, we're off to a good start. Next follows a long speech about how I need to call a special number if I stuck in the middle east and need to get back home, general precautions I need to take and my rules and rights. All great information, except I've already indicated that I call to cancel a policy. The chance that I'm sitting in an airport in Bahrain, desperately trying to get home, yet I decide that now is a good time to go through and cancel unneeded insurance policies is absolutely zero. You already know why I'm calling, tailor the message to that.
2) Internet is out, for the second week. Customer service dude is typing in stuff, looking stuff up, trying to figure out why the case has been closed. "While we wait let me talk to you about our streaming bundles"... Dude, I know the boss is making you do this, but don't try to upsell a streaming bundle to a customer you can't even get online.
The doctors office is the worst though. Their entire system for guiding you through when to call and where to call take minutes for them to explain. The call it routed to the same people regardless. There are so many confusing and irrelevant messages from the system and in the end you are still routed to the same set of people.
Most of my calls to customer services is because selfservice online absolutely suck and can't do simple things. Every industry could save a fortune in callcenter costs if their websites was ever so slightly better. Often it's not even about being able to selfservice, it can just be providing the tiniest bit of actual information. Your call volume is larger than normal for the past five years, because your stupid website is getting worse every year.
Customer service has different meanings. Personally I prefer
good, simple and effective documentation. I hate having to use
a phone to explain problems. It's ok to do it in person, but
not my preferred way.
> It's all very well to experience your own product when it is working, but when was the last time anyone in the above organisation went through a "difficult" customer journey.
I kind of prefer companies that build products that never ever
need anything. Not even warranty calls, because the thing just
keeps on working.
What I noticed in the last few years was that we are too dependent
on google search. Now that it sucks, finding high quality information
has become harder - and AI trend is further ruining this, as everyone
just has the AI summarize stuff now, which does not always work either.
This resonates. We run a medical device manufacturing operation. We used an ERP for years that claimed to specialize in our industry. In practice it was a bloated system designed by people who'd never worked on a manufacturing floor. Bank reconciliation felt like threading a needle in the dark. Simple tasks required five nested menus.
So we built our own. Went live on it a few months ago. The bugs we catch now are ones we catch because we're running real production on it every day. There's no substitute for being the one who has to suffer through your own software.
Some years back I was introduced to UK service consultancy Vanguard (and their "Vanguard Method"), essentially a systems thinking approach to improving service (not to be confused with the enormous US investment management company).
In their world, "smelling your own farts" (ie. listening to and, more importantly, understanding what matters to your customers using normative learning methods) isn't primarily about empathy, it's about getting knowledge so you can understand how to intervene in your company as a system.
Put that way, it's not a waste for decision-makers to listen to customer phonecalls, it's in fact the only way for them to gain the knowledge they need to understand what to do to improve their service (assuming that's their goal).
I just assume a lot of product managers have never used their products anymore, and it shows across the industry.
Have you ever tried to play Xbox live after not being signed in for a long time?
Have you ever tried to use smart search terms in the Amazon search box?
Have you ever tried to use these tv channel apps on a Roku?
Do Fridgeaire reps own their own appliances?
I’m just convinced people don’t use their own products where the people that use them and care are so few and far in between things are only getting enshrined.
I know a person. Ex-Googly. Doing his startup. Spent one year on a crazy complex product. Investors do not get it. Users do not get it. He spends 99% of the time explaining why his ideas are so good. You ask to try the product, weeks pass and just slides and video demos. When you eventually try it, it's so confusing and nothing really works. I tried to make him understand that the constraint he needs to fit is users being able to understand rapidly why it's useful to them and that it should work. He does not care. He says it's about the story, and that the story will drive millions unlocking a super big team building his idea. I said that's cool but why in the meantime you don't just go with one thing that is useful and works and then procedurally evolve to your vision while you interact with the user and discover more about their problems. The answer was a one-hour speech on Google leveling system. Maybe he is right. Time will tell.
The age old problem. Customer service isn't supposed to be mind numbingly unfathomable, and responding to customers should be easy. Unfortunately, unless higher management engages (is that Bezos or Gates? I've heard both) things will get lost in the mix.
214 comments
> There's an oft told story about Jeff Bezos pausing a meeting to call his own customer service number - and waiting over 10 minutes for an answer.
One of my jobs was at a company that had developed at unhealthy amount of bureaucracy and politics. The product barely mattered to some because they were playing internal games of grandstanding, taking credit, and building their empires.
In meetings where were supposed to be talking about product direction and priorities I would some times pull out my phone and open the app to try to demonstrate some real problem with the service. The tone of the meeting would change to panic as certain product leads would try to do anything to stop me from showing what the real product did instead of their neatly prepared slide decks that showed a much nice story for the executives. I became the enemy for showing the actual product instead of their alternate world of KPIs and charts.
I found out SSO was broken. They had to login to every app using the same account. Twice per day because the token live was 4 hours "for security".
I found out it was because they published these apps as PWAs, making them more isolated than normal apps.
I asked the product manager and he says the issue is "with Apple and Google", not his department. When asked why he chose PWAs for the apps he said this was easier to deploy, saves them developer accounts and such.
Since I can't force him to change I found a workaround: SSO works in PWAs if you use Edge on a recent Android version on a Samsung tablet. Lucky me we had bought Samsung tablets (this was not a requirement when purchasing I looked it up, just luck).
I asked the Intune manager about this and they said the field engineers should just set Edge as default in stead of Chrome.
When trying this on a company tablet it said: "Edge disabled by X group policy". That guys' department set the policy...
After they removed this I asked why it wasn't the default browser and he said this wasn't possible. I challenged him on this by Googling the Intune manual to set the default browser.
Later they said they had raised a support ticket with Microsoft for this.
On the internal Wiki I found a document describing the problem. It was dated 11 months before I joined.
The U.S. Nuclear Navy, for all of its many flaws, gets this right. Generally at least once a year, the head of Naval Reactors - a four-star Admiral - tours every vessel, which may include a brief underway period. During this tour, the Admiral will talk to the engine room watchstanders, with all senior leadership removed. They’ll ask how daily life is, what they find challenging or annoying, what they like, etc. There’s obviously a lot of self-filtering (though sometimes not - Navy Nukes are not known for their social graces) that occurs, and also what a junior watchstander finds annoying may just be a required part of the job, but some useful signal is gathered.
Even outside of the nuclear program, one standout example was Admiral Zumwalt, who as Chief of Naval Operations implemented 70 different changes over his tenure as a direct result of talking to sailors, all of which were designed to improve quality of life, efficiency, or communication.
They did trial run after trial run, made sure trying to make sure there were no bugs in the demonstration path. They nailed it, presentation went smoothly, live demonstration just worked. Provisioned a bare metal instance, had it running hosting something within minutes of launch. Larry was suitably impressed, but the thing that most impressed him was that he'd been presented with an end-to-end live demonstration. It had never occurred to any of the folks involved to do it any other way, but apparently all too often, all he ever saw was slide shows from product teams, particularly when things were several months away from public launch.
I reflect on that situation from time to time, wondering at which stage you sort of go from expecting to see live demonstrations, to slide shows. I assume it just slowly slips away from you, one at a time until you're stuck in the land of "make believe we have a good product".
> ... When was the last time the CEO of the above company called their own customer support line?
So Bezos definitely hasn't done that in a long time. Definitely not in India. So I would say the answer ought to be: a CEO does that or has to do that only until the company becomes too big or has captured a sufficiently entrenched large slice of the market.
At that meeting, I realized most of them had never used the product and see their claim to leadership role due to their the ability to manage up and down.
I use the product on my personal projects and I hated it with a passion.
My old man, however, still feels some kind of righteous indignation when he spends his hard earned money and doesn’t feel he’s getting what he paid for. He loves to give a piece of his mind to the companies that mistreat him, and he always says “And I hope my comments are being recorded for quality assurance!”
"Eat your own dog food" == experience your own product.
"Smell your own farts" == experience your entire product, including things that are typically unmentionables like customer service and billing
We've found ourselves trying to find this balance on Tritium. It's a word processor for lawyers, so has a specific narrow domain that allows us to provide a differentiated experience from Word. But if we try to use it like Word, we end up wanting generalized features that don't fit that strategy. I wrote a little about what we've come up with here: https://tritium.legal/blog/eat.
This is one of the compelling rationales for closed-source / commercial software in certain B2B SAAS domains. It seems like you just cannot adequately test the happy and sad paths from a QA perspective in FOSS unless it's (1) insanely successful or (2) a dev tool.
(Motivated) people at small companies "care", and what I mean with that is they are responsible and can see a large enough portion of the customer experience that - if something is broken - they'll see the pain and try to address it.
At a big company no one cares. They of course care about their job, but their job is such a small fraction of the overall customer experience, that seeing their work having an impact on their customer is exceptionally difficult.
That's why large companies need to encode customer feedback into a system to imitate feedback cycles. Mostly in metrics. That's a very lossy way to capture signal, and leaves a lot to be desired, but so far it doesnt seem like anyone has come up with a better system.
To me "unaccountability" -- or whatever naming fits better -- needs its own circle of hell.
1) I call to cancel an insurance policy on a car I sold. I'm greeted by the IVR, press three to cancel a policy, we're off to a good start. Next follows a long speech about how I need to call a special number if I stuck in the middle east and need to get back home, general precautions I need to take and my rules and rights. All great information, except I've already indicated that I call to cancel a policy. The chance that I'm sitting in an airport in Bahrain, desperately trying to get home, yet I decide that now is a good time to go through and cancel unneeded insurance policies is absolutely zero. You already know why I'm calling, tailor the message to that.
2) Internet is out, for the second week. Customer service dude is typing in stuff, looking stuff up, trying to figure out why the case has been closed. "While we wait let me talk to you about our streaming bundles"... Dude, I know the boss is making you do this, but don't try to upsell a streaming bundle to a customer you can't even get online.
The doctors office is the worst though. Their entire system for guiding you through when to call and where to call take minutes for them to explain. The call it routed to the same people regardless. There are so many confusing and irrelevant messages from the system and in the end you are still routed to the same set of people.
Most of my calls to customer services is because selfservice online absolutely suck and can't do simple things. Every industry could save a fortune in callcenter costs if their websites was ever so slightly better. Often it's not even about being able to selfservice, it can just be providing the tiniest bit of actual information. Your call volume is larger than normal for the past five years, because your stupid website is getting worse every year.
> It's all very well to experience your own product when it is working, but when was the last time anyone in the above organisation went through a "difficult" customer journey.
I kind of prefer companies that build products that never ever need anything. Not even warranty calls, because the thing just keeps on working.
What I noticed in the last few years was that we are too dependent on google search. Now that it sucks, finding high quality information has become harder - and AI trend is further ruining this, as everyone just has the AI summarize stuff now, which does not always work either.
I actually have a recording of it (scratchy), but won't link it, because it's probably not worth it. It was a riot.
https://news.ycombinator.com/item?id=47427224
a happy coincidence.
In their world, "smelling your own farts" (ie. listening to and, more importantly, understanding what matters to your customers using normative learning methods) isn't primarily about empathy, it's about getting knowledge so you can understand how to intervene in your company as a system.
Put that way, it's not a waste for decision-makers to listen to customer phonecalls, it's in fact the only way for them to gain the knowledge they need to understand what to do to improve their service (assuming that's their goal).
Have you ever tried to play Xbox live after not being signed in for a long time?
Have you ever tried to use smart search terms in the Amazon search box?
Have you ever tried to use these tv channel apps on a Roku?
Do Fridgeaire reps own their own appliances?
I’m just convinced people don’t use their own products where the people that use them and care are so few and far in between things are only getting enshrined.
https://yuraa.com/2026/03/21/customer-service-as-a-product/