Hacker News Reader

Comment by dawhizkid | Hacker News Reader

All I can gather from this is Airbnb has way too many engineers and designers for the amount of actual work there is to go around.

At the end of the day, the day to day engineering problems Airbnb has are not very interesting. It’s a web app that can handle peak traffic with standard technologies (meaning nothing real time, live streaming, or millions of users at once).

I have no clue what most engineers would be doing there day to day.

View on HN

I don’t know, having spent a significant amount of time with Airflow and Superset, their analytics stack must take an army of engineers, product people, and analysts to use. They have to troubleshoot issues with visits in real time. They have to not double book properties and keep an audit log of every crazy thing hosts do, and detect all kinds of fraud, all while supporting global localization. The have municipal level compliance to manage, and I’m sure a huge back office to help their various legal efforts.

I’m convinced that any business that operates at global scale needs an army of engineers doing, training to do, and coordinating across hundreds of technical areas.

Or is hiring engineers somehow so glamourous that doing so outweighs the profit motive?

Sorry but are you really citing the problem of “not double booking” properties as one of the reasons Airbnb needs to hire an army of web developers?

“How many developers does it take to design a reservations schema?”

I would put open source projects like the ones you’ve mentioned as more evidence Airbnb has too many idle engineers. I think even Google uses Tableau for their BI analytics stack. There’s not necessarily any business justification for a private startup to have to build that from scratch, and the only way Airbnb could justify working on projects like that is if they had a ton of engineering capacity not already working on the core marketplace product.

Most if not all of the challenges you listed are non-engineering related (legal, compliance, host operations, localization)

Hmmm, then one has to wonder, how did AirBnb become so successful while so many others have failed? It's a mystery that will tease our imaginations until the end of time no doubt. If only there were some clue, some window into their company and culture that could help us explain this puzzle?

But yeah, seriously, how hard can it be to make AirBnb? It's just a couple of tables in a database, a HTML form and a jQuery image gallery.

That's a strawman. No one has stated that the success of AirBnB is easy to replicate, nor that a system of equal functionality is easy to create.

Isn't that just a transactional database and support staff? Sounds to me like any simple bookkeeping software with a web front-end.

In fact, were I inclined so, I could make an airbnb in a weekend. /s

I'll use an iceberg analogy: There are moving parts behind the scene making websites' wheels turn that you're not considering. I'm sure there are projects behind the scenes at Airbnb that are interesting but aren't apparent from an outside perspective. Don't judge a product's complexity by their landing page.

Since you worked for Uber, someone could say Uber only has a mobile presence so it's less complex and less interesting than a full website.

I don't get this post? Does this mean only Apple & Google are allowed to create fonts? Are companies no longer allowed explore pursuits outside their immediate business model?

I can only imagine posters like you, on Usenet back in the days complaining that Apple has "too many engineers & designers" when they had Susan Kare design their icons.

No, there is a difference what a VC-funded startup should be burning money on. It's not a public company, yet.

But where do we draw the line? Should they be buying cheap Lenovo laptops and second hand furniture for their office too, out of fiduciary duties? I don't mean to be facetious, but my point is that on some level the investors is investing in the company (or more accurately its leadership) and trust they will do the "right" thing... its a heuristics thing - does sweating the details on things that other people may find irrelevant make it so? Not necessarily. Building the brand, culture, product the way that the company deems it best is kind of what its about no? (validated by paying customers, company growth and whatever metrics the VCs and others are looking at).

Whilst "what a VC-funded startup should be burning money on" is something that the investors should have input on, I don't think random people on the Internet get to be the gatekeepers for what a company is allowed to do with their investors' money.

It's not about being allowed or disallowed. It's about having priorities as a startup. And every random person on the Internet has the right to say that his priorities at this stage would be very different.

If you think people shouldn't be discussing what companies do with their money then maybe you're on the wrong forum ;p

As an example of what some of the engineering people do, that wouldn't be an immediately obvious need if you haven't been involved in running global businesses at scale:

https://medium.com/airbnb-engineering/tracking-the-money-sca...

A tiny part of what they need all these resources for.

Obviously running any sufficiently large business on a global basis is a tad more complicated than simply keeping a website up that can absorb usage spikes.

I worked on payments at Uber for 2 years, so I know there is unavoidable complexity in operating at a global scale. I would argue payments is harder at Uber especially on the remittance side compared to Airbnb but obviously I’m biased.

I would agree that money engineering is probably the hardest problem they have. But unlike Uber they don’t have any realtime/streaming/GPS-dependent tech.

So outside of payments it’s not clear what “interesting” engineering problems they really have, which means throwing engineers and designers to work on pet projects they’ve convinced themselves is really important to the future of the company but are just...pet projects.

As any company grows, I’d argue that solidifying their identity becomes increasingly more important. Hence, a custom typeface they can use to distinguish the brand.

Whether it will be successful or worth the effort and expense is a different discussion. It seems like a well-designed and cohesive set of fonts, but otherwise unremarkable (in my opinion).

Brand is all they have, since the app is so simple. Saying that a typeface in an app is wasteful is like saying it’s wasteful for Starbucks to commission their own chairs.

...and does Starbucks commission their own chairs? Serious question

Yes, as in their purple chair [0]

[0] https://news.starbucks.com/news/how-a-purple-chair-changed-t...

Airbnb handles hosts and guests paying each other in different currencies. They're also active in 190+ countries. I would say their payments system is more complex than most companies.

Setting up payment interfaces in each currency takes a while but matching isn't hard. FX streams are easy to get, the algorithms to convert currencies so that you avoid exposure are not complicated.

It's probably mostly keeping all interfaces to payment vendors working.

> All I can gather from this is Airbnb has way too many engineers and designers

This is a perfectly valid job for the design team (with the assistance of the engineering team, initially to make sure they are not missing technical requirements and latterly for implementation).

Presumably (I've not checked) they were previously using either very common font which doesn't stand out at all and/or a licensed font that was costing them money. Sorting the "stand out" issue is important because (much as I hate the fact) branding it important, and sorting any licensing issue could be even more so (licensing costs for typefaces can add up considerably at scale - those designers and engineers could have just saved the business a ton long-term).

Why did they not do this from the start then? That would probably have been premature optimisation. Early on in such a product/company's life-cycle there are growth issues to manage & engineer through/around and unexpected fires to fight, which are more important sinks for your engineering resource, but when the dust starts to settle a bit there is time to deal with important-but-less-urgent matters.

HN at its worst: using a post about creating a font as a cheap excuse to mock Airbnb's engineers and designers, their jobs and their reason for being employed there.

And somehow it's the top comment in the thread, as opposed to, say, actually discussing the font or the content of the article. Wtf.

it's a classic HN. "I can build that in a weekend!"

First of all, nobody mentioned building anything in a weekend. Second, generic responses like yours can also be considered "classic HN", so it's kind of like the pot calling the kettle black.

And third, If you've made the same font for Netflix, it apparently takes 3 months to make enough subtle changes to brand it "one of a kind".

Actually my response was - WHY build/work on this? I understand that licensing a font can cost a company a lot of money but there is no reason why a company of this size can't just use any of the thousands of open source fonts available. It is a waste of time/resources that is pretty characteristic of a big tech company (having worked at one).

In at least the cases I've seen - the reason why this happens is because design teams want to stand out and stamp their identity on the company. It is a reason to add imaginary value to internal "design studios" and promote designers to arbitrary titles.

And engineers, PMs, etc. mostly do the same. Create work, features, platforms that add no real value but simply to stamp their identity and pursue (effectively) self-promotion.

The original commenter makes a great point about saying that there are many other things to focus their energies. Okay, throwing engineers at a single problem is not a good way to do things.

But the AirBnB iOS app has been broken for ages. Various visual and functional bugs. Offline support is arguably the most important feature for an AirBnb app (if you're traveling and don't have service). But this hasn't been tackled at all.

Instead AirBnB engineers are going through all of their properties and updating their fonts, fixing issues with responsiveness/scaling, packaging on mobile platforms, etc. WHY?

The WHY is because they want a unique brand identifier. If they used a font anyone else could use then it would no longer be unique to the brand. Essentially they don't want you to feel comfortable or at home on a competitors site who may try to mimic their branding.

This expense goes under marketing not R&D.

Let's discuss the font then. It's boring and generic. Obviously it needs a 30-page puff piece written about it. Remind everybody the AirBnb had a super bowl ad, etc. This is not content marketing. Yada yada.

I'm making a point on how indulgent it is to spend technical resources on something of so little consequence, and then write 2(!) 10,000+ word blogposts (including https://airbnb.design/working-type/) on it and also produce a video to announce it to the world.

The second blogpost actually says After 18 months of creating, refining, testing, and integrating, we flipped the switch and released Airbnb Cereal across our brand and product. 18 months?! This is HBO Silicon Valley level insanity.

Spend 18 months figuring out how to stop real issues on the marketplace like discrimination against minority guests. The only solution Airbnb has offered to quell the #AirbnbWhileBlack controversies that started in 2016 has been to force hosts and guests to click a button pledging to not discriminate. That is an actual issue. This is throwing idle bodies at a non-existent problem.

>And somehow it's the top comment in the thread, as opposed to, say, actually discussing the font or the content of the article. Wtf.

Specifics of your comment aside (which I may agree with), on principle it's not necessary to always discuss the "content of the article".

The post serves as a starting point. It doesn't have to confine and frame the discussion. Especially if there are other more important stuff lurking or implied behind what a post says.

I don't like excessive complaining but isn't it valid to ask why a company such as airbnb needs to spend the money on their own font?

My guess is that their biggest technical challenge is fraud detection. High dollar two sided marketplace is ripe for it.

The designers do a lot of product development-- or it seems that way based n the narrative. The website is not AirBnBs product, they sell the entire experience. I get the sense it is a much more designer centric place than engineering.That said, i am sure the engineers are optimizing not only the site but the reams of inbound data that comes with a global business. Clearly they must defend against fraud, protect their users, do identity verification and other data intensive predictions. Its up to you whether these are "hard" or "interesting" but they probably do have a ton going on at their scale and while it is conparatively simple to google/fb i think you are dismissing a lot of challenges that may not be immeadiately obvious.

> too many engineers and designers for the amount of actual work there is to go around

Paging the gang over at Twitter.

It’s easy to say you’re overinvested in design, but many massively successful companies are separated from their also-ran competitors by seemingly minor design decisions, like 140 character limits or disappearing messages. In this context, design seems worth the money.

From Typology: Type Design from the Victorian Era to the Digital Age by Steven Heller and Louise Fili, page 185:

> They have also developed new faces based on these classical materials, but added those nuances and attributes that they lacked when current. Many of these faces are created for publishing, advertising, and computer clients who desire exclusive, proprietary alphabets. Like the king of France in the seventeenth century, these companies are patrons of an iconoclastic breed of type designer. The New Classicism is about both revival and invention, because there is no sense in remaking type that has existed before without improving upon it.

The font was commissioned from Dalton Maag, who specialize in this kind of work (as the second highest ranked comment notes). So if anything it is money that is being wasted, not employee time.

After 18 months of creating, refining, testing, and integrating, we flipped the switch and released Airbnb Cereal across our brand and product.

I'm not sure what you mean—but it is very important to note that the bulk of the work of creating a new, globally-internationalized typeface, is hard, tedious editing of glyphs and kerning tables etc.

Sure, AirBnB staff were involved, but they did not put in the huge amount of effort necessary to actually create the font. So that effort did not happen on AirBnB's property, so to speak, while AirBnB design staff "got to commission a font" as part of their career progress.

https://airbnb.design/introducing-airbnb-cereal/ is PR puff, it could mean they spent an afternoon or a decade of human-hours on this.

Look at it this way: if their technology cannot separate them from their competition, they effectively are a marketing/sales driven company.

I’m not saying that should spend time and resources on a font now, but I’m also not saying they shouldn’t.

A font can help their branding, and having a strong brand can help keeping their margins high. Typography works for Coca-Cola, for example.

Designers are creatives. It is better to nurture a great team of designers with projects such as these because of the intrinsic rewards the project creates for it's creators. You don't keep design talent around for long unless the work engages creative passions.

The same can be said about software engineers. What is code craftsmanship?

The argument was that there are too many designers and engineers. Not keeping the talent around for long is exactly the point.

Have you every written a scheduling app or a booking engine? At their scale, it’s hardly not interesting. Just that aspect alone is crazy hard to get right. On top of that they essentially have a CMS for hosts and very complex payment system. Security alone requires a pretty massive team.

Each component of their system could be a separate company. I own a small B&B in a tiny corner of Provence and I get dozens of inquiries a week — all of those inquires are spending time looking at images, checking availability, etc. now multiple that across almost the world and you will find some extremely interesting engineering challenges.

And site reliability has been exceptional. I’ve also never encountered a single bug in my normal use. That’s a big deal.

> I have no clue what most engineers would be doing there day to day.

Keeping up with the ever changing requirements of service design, marketing and sales, I suppose.

This is basically what I’ve gathered from spending time inside their main HQ as well. A bunch of insanely overpaid people drinking free beer and doing absolutely nothing interesting while supporting a Rails app.

Maybe they can use them to start complying with local laws.

Reply on news.ycombinator.com