Hacker News new | comments | ask | show | jobs | submit login
Facebook and Twitter Don’t Know What You Want (diegobasch.com)
36 points by diego on Aug 18, 2012 | hide | past | web | favorite | 12 comments



Reposting the comment I posted on the blog post here:

It seems to me that consumer tech startups have a default belief that selling the promise of sales is easier and more lucrative than selling something directly. It’s like a slightly fancier version of the “how to make money online” courses.

They also seem to look down on selling their products directly to users (“we don’t want to nickel and dime our users”). They have a strange attraction to portraying this benevolent change-the-world-for-free image (see http://help.kik.com/customer/portal/articles/654821).

I suspect there are big unexplored opportunities for selling software services directly to consumers – things which improve people’s lives noticeably, without them having to first do a ton of donkey work (checking in, writing reviews, etc). But, the approach to building such products would be to go deep on a specific problem in a small market and expand only when you’ve nailed that business. Unlike typical consumer startups, the focus should be on high quality not scale.

Anyway, I am working on some ideas along these lines. I shared my perspective to clear my thinking and also hear what you think about it.

Oh and thanks for introducing me to the delightful word Veeblefetzer!


Most consumers are not used to buying online services that don't have a physical component. We're used to free. And when there's a free alternative, not a lot of people will be paying for a slightly better paid version.

If you want to sell stuff to consumers, it's important to have a physical component. And it has to solve a very very crucial problem like losing weight.


This article is spot on in pointing out that search advertising is more valuable and more effective than display advertising, and that Facebook will most likely never be a major player in search. However the rest of the things this article is getting at is bogus, much like most of the talk I hear regarding Facebook's monetization.

While speculating about the crapiness of Facebook's ads system and bringing up numbers like phenomenally low click thru rates compared to search, people seem to forget that display, interruption based online ads work. Facebook can also do it better than almost anyone because of the 'interest' and 'likes' data they have. And I don't believe they are even using retargeting yet but that's another conversation for whether or not it's even possible for them.

Basically what I'm trying to say is you can't compare search ads (Google) to display ads (Banner ads, Facebook, basically everything else) by just stating the difference between the two. Such a comparison must be done scientifically through case studies or experiments. Don't bring up Facebook's average click thru rate in such a debate without mentioning that the billion dollar display ads industry (which has no end in sight) has the same low click through rates. Would it be nice for Facebook to know what you want? Yes it would be amazing, but that's not possible, and that's not a problem. They already know what you like, and it makes for a great (self-serve) display advertising experience. I'm not saying what they should do to make more money because I don't know, I'm just saying that their current ads system is not the problem.


The only problem with display ad is that you can't effectively measure ROI for small campaigns. The value that Facebook is supposed to bring to the table is a scalable platform for small, well targeted campaigns. As a result, big ad spenders are still spending their big money on tv, newspapers and billboards.

Here's an advice for FB. Make your targeting more relevant for photos. Why not do some object/brand/keyword extraction from photos and sell that inventory. Hire some computer vision scientists and let them loose on your photo collections. But of course, when the whole company is focussing on shipping code twice a day, there's no place for someone who can sit back and think.


Nobody says that display ads don't work, of course they do. The point is that they don't justify Facebook's hype or valuation. It's a matter of numbers, and it's perfectly fair to compare both. There's a reason Google is by far the most valuable company in the online advertising industry.


I think if FB and Twitter fail (compared to their expectations), that would be a good thing for the startup world. Right now we see a lot of consumer startups whose sole business model is online advertising - it doesn't matter if it's Adsense, display ads, sponsored tweets, what have you - it's the same thing. Online advertising. And even if you're optimistic, if you do the math, you realize one thing: it's not sustainable.

If there's 1 thing I like about App.net, it's that we're starting to see a shift from relying on ads as revenue to building a product users will PAY for, whether it's B2C or B2B. In a sense, relying on online ads is a semi-copyout. It's like you're throwing your hands up on the air, and saying "I got NO clue how to provide a service people will pay for here, I'll let YOU advertisers do that instead for a fee!".

Most people are just naive and see FB making billions on ads and think they can build a startup around advertising as well. The harsh reality is that only a few hundred companies can rely on advertising, and even most of them are having troubles these days, with the incredible high cost of creating high quality content, diminishing advertising rates, and avoiding Google penalties that happen every other week. The ROI is just not there.


Besides relevancy, another problem for Facebook and Twitter ads is that to be successful, the ads should not degrade the user experience. Speaking just for myself, promoted stories or tweets definitely feel like a degradation of user experience even if I understand why they do it.


I agree. That's one thing Google figured out early on. It's much easier to not degrade the experience when you can match an ad to a query, because you can make sure that it appears relevant.


Exactly. I cannot think of any other use cases for ads where they would not cripple user experience but search. When I am looking for something you are most welcome to show me your suggestions. When I am chatting with my friends, o reading updates from Curiosity Rover—stay away.


Ads do seem to be the "default" business model on the web, which is understandable but a bit sad. App.net has an amazing potential around this if they build out their API and developers can use it as infrastructure for inter-app communication. It'd be like an internet-wide "open" enterprise message bus. I'd very much like to see that, and would have liked to have seen Twitter go that route. They have outright rejected that though.

Facebook on the other hand is in a more difficult position short-term, conceptually they're not well positioned for things like payments, etc. But it's hugely encouraging that Zuckerberg is so young and has such a long-term view on Facebook. I'm fairly sure they will figure it out and continue to grow Facebook into something even more amazing than it is now, if the bottom doesn't fall out from beneath them (read: their engineers leave in frustration or boredom).


I think what he's trying to say is that Facebook lacks intent data (which is the richest at a search engine like google).

Open Graph actually does give Facebook a great deal of behavioral data that includes intent. But AFAIK it's just an arbitrary policy decision that Facebook is not using off-Facebook data within Facebook, and also not using Facebook data outside of Facebook. I didn't mean to write a tongue-twister, sorry there.

If they change their mind on the arbitrary policy decision they'll have a lot more data to work with.


I know one thing I don't want, and that is clicking ads. So I don't.




Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: