Leaving the merits of advertising itself aside for a bit, this post brings up a bigger topic that I think anyone running a startup needs to think about:
When you're getting involved in a startup, it is incredibly important that you be honest with yourselves and your partners about what the goal really is (this is to the point made the other day about self-awareness) and what you're really building.
A Feature: A piece of a piece of a puzzle. "It's like Facebook's News Feed, but instead you get audio updates instead of text updates!". Hackers can be guilty of this often - focused obsession on improving the world (or a piece of it) by tackling even the smallest problems. Not necessarily a bad thing, but not necessarily a guarantee of success.
A Product: Instagram is a product. It's full-featured, it's beloved (or was ;), it's free-standing, it works, and people use it. But do people pay for it? Would they? Would they make money by themselves or only in the context of a bigger company (like Facebook). Twitter is also a product, along these lines.
A Business: 37signals is a business. People pay them to use their products. They are free-standing, they are profitable, and they don't _need_ outside investors (need vs. want is different, btw - sometimes it makes sense to take strategic investment).
There's nothing inherently wrong with any of the above paths, btw. But you need to understand which path you're on, because you'll do a better job of it that way. If you decide to be on the feature/product path and hope to get acquired/acqui-hired, that's fine - it's a legitimate approach (that works, at least for now) and folks have made $billions doing it (see: Instagram & everyone else).
But it's a different path than starting a business. Goals are different, operations are different, and outcome is different.
The point, in the end, is that often the response - "well, I guess we'll sell ads" is what people who are starting features or products often say to convince themselves or others that they're on a worthwhile _business_ endeavor. And that's often dishonest (intentional or not).
Understand the path you're on and I'm willing to bet success will be more likely.
What would you say is the biggest difference between a product and a business though? Is a business just a company with a lot of successful products? Following your terminology, I'd say people create products and if they are successful turn into businesses. I totally get what you mean by features though and sometimes I wonder if tackling these tiny problems with full out services is a legitimate business strategy or just a fad.
The product is what you sell. The business is the rest (marketing, sales, etc.). A product itself is not a business, because products don't sell themselves. No matter how good they are, someone needs to market them in any given way. Problem with hackers is that they falsely believe that building a great product is enough. Nope.
Ever see the mountain of downright awful products out there? How do people buy such such garbage? Because those products are marketed well. If people sell garbage with good marketing imagine what a great product and great marketing can accomplish. One word: Apple. Billions of dollars due to great marketing and great products.
Instagram the social network/app are the product. The business is all the social engineering being done to market it, the infrastructure to handle all those booty shots, the HR dept., etc.
IIR, 37signals originally built websites (or something like that). They pivoted the company into SaaS around their internal project management solutions.
Thanks for the advice, it makes sense. However shouldn't the choice be between a feature and a product as they could develop into a business? You can't just make a business like a product, it takes time. A feature/product evolves into a business since no one can be profitable from launch day.
This is a good distinction, I find that businesses start with the notion that they will make money so the question is "What can I charge for it?" or "Can I monetize the tool/data/views/Etc?" and a feature starts with "How many people would find this useful?"
I find that businesses start with the notion that they will make money...
There is a distinction to be made. There are many startups that are the product itself. Their product is just a coy to make it seem as if they are successful. Their only aim is to sell the success to investors/people with money so they can cash out. In this case, the startup/product provides a very different set of usefulness. Yes, these types of businesses are useful. They generate traction for the buyer, their engineering talent put to better use, etc.
Though these days I just start with the question: Who will buy this? If I can't answer that then I move on. No sense in trying to mine a business model that does not exist.
This post has a very naive understanding of online advertising. Here's a breakdown:
- If you're thinking of advertising in terms of just slapping on some Google Adsense then you need high amounts (millions of users per month) of traffic to make reasonable money.
- If you're thinking of specific kinds of targeted advertising, this can be very powerful depending on your niche. If you can hire a sales team to push advertising in those niches, you may be very successful, especially if you can show quality ROI to advertisers. eg. a cancer hospital advertising on a site with cancer related information.
- If you're thinking of other kinds of advertising, business partnerships in terms of 'upsells' to a free product work really well too.
The above is a gross generalization but claiming that 'advertising' is a broken business model is laughable. There are definitely certain use cases where online advertising can be your business model and of-course others where it would absolutely not work.
The main downside of advertising supported business is that advertising can go in and out of demand, so you add an additional issue/unknown/risk in terms of product revenues.
TLDR; most high traffic sites can make reasonable money from online advertising.
I think the reason that most people in this story laughed at it when advertising was brought up is simply because advertising is the "default" answer from people who have not thought one bit about how to monetize.
And who, because of that, are not just answering "advertising", but are answering advertising without having a credible story of much about it in general as well as without having the faintest idea of what kind of model would be realistic to model their idea's income potential with advertising.
So really, "advertising" was a proxy for "haven't got a clue".
"This post has a very naive understanding of online advertising"
Not really. The post talks specifically about "just putting ads on", which means we're talking about something like AdSense and not the more lucrative options you mentioned which involve a sales team.
The fact is, you'll need a really high amount of visitors to make any decent amount of money through AdSense - especially if you have several employees and investors who need to get their share.
The post is right on the money. No doubt there are certain forms of advertising that are effective - but the point made by the post is that slapping AdSense on your startup's pages isn't going to cut it.
I have a site that has around 4-5m visits per month, yet all attempts to monetize it have failed. I'd love to use Amazon's referral program or Google Ads, but I can't because the site's content is somewhere between reddit and 4chan - deemed too shady for anything else than porn ads.
So, the site still remains a hobby project - exactly as when I launched it 6 years ago.
Recently 4chan had an excellent monetization idea:
Require passing a CAPTCHA in order to post.
Allow users to buy a code that will let them post without passing a CAPTCHA.
Accept bitcoins to avoid payment providers' prejudices and help users remain anonymous if desired.
Of course, you want to add some limits to discourage people from publishing their codes. For example, a rate limit on posts (with a slightly more generous burst allowance, e.g. you get 10 CAPTCHA-less posts / hour, but if you don't use them, up to 100 CAPTCHA-less posts are automatically "saved" in your token's reserve). Or a limit on the IP addresses, e.g. whenever a code is used on more than one or two unique IP's within a short time, lock it out. And of course the gold standard: Manual monitoring for spam, illegalities or other ToS violations, punished by account cancellation (and referral to law enforcement for extreme cases).
You could sweeten the deal with custom markup: Make posts by paying users be a different color or something, to show they support the site. Or if users have a traditional forum account (as opposed to anonymous posting), add an icon or something to their profile.
If I was implementing this, I'd have three tiers: The "cheap tier" (around $5) for just CAPTCHA-less posting. The "medium tier" (around $15) for CAPTCHA-less posting and a note in your profile that you're a supporter. The "high tier" ($40-$60) for different-colored posts.
If you want to know why I suggest three tiers, Joel on Software has a post that explains it very well [1].
If it's a forum style thing then discussion systems with a high number of pageviews per unique visitor (vbulletin/reddit/etc) will generally have very poor advertising revenue.
Maybe you need to look at a different model though ? Sort of like reddit gold, premium memberships, or some other services that may somehow add value to your site ? It's often surprising how many people are willing to pay $2-$10 per month for something!
If you really mean 4-5m million visits (and not pageviews) I'd love to take a look and see if I can help you figure out some sort of solution. Send me an email (addrress is on my profile).
The site is http://pr0gramm.com/ (NSFW) - mostly german visitors and yes, it's 4-5m visits, not pageviews. It's kinda impossible to define what a "pageview" is on that site anyway.
Have you considered partnering with a company that would play off the kind of irreverent content IRC carries? The company TshirtHell.com came to mind as (IIRC) they've fostered controversy in the past. You do something like every xth image on your site is the image from one of their Tees with a referral link to buy as the source. You could tweak x to find the right balance between annoying and profitable.
I run a similar semi-pornographic site. The problem is payment gateways like Paypal do not like sites with adult content, and traffic typically isn't high enough to justify splurging on a merchant account.
(Obvious suggestion) look at how similar content sites are being monetized. A lot of ad networks advertise on PirateBay. Maybe you can contact them and include similar ad units.
In my personal experience 90% of people with ideas think about advertising, when in reality, only 5% (I'm guesstimating) make serious money with it. Or maybe even less.
You're right, and the problem here is a naive understanding of advertising as a business model rather than it not being a good business model.
I happened to go the opposite way. My first major project was an online advertising success. I naively applied my 'learning' to my 2nd project. That obviously didn't pan out. That led me to a better understanding of business models and how and where they may work. So I guess it can go both ways.
Most people are only familiar with 'big names' like techcrunch, lifehacker, facebook, etc. which are very much advertising supported platforms. That's probably why they think 'advertising' is the answer even if they're not building a high traffic web app.
> If you're thinking of advertising in terms of just slapping on some Google Adsense then you need high amounts (millions of users per month) of traffic to make reasonable money.
Funny story that will probably get buried since I'm 2 hours late to this thread:
When we were first working on Kazoo we had thought our monetization would come from a service we called "promo calling". Basically we would detect the caller ID of where you were calling and place an ad for a competitor on your phone display with a coupon and a call to action of rerouting your call to the alternate vendor.
It just pissed people off.
We ultimately settled on a premium SaaS model but my point is this: try lots of stuff. A lot of it will fail, but failure is simply invalidating one incorrect hypothesis. Try lots, fail fast if you're wrong and roll to something else.
The key for us is that our core values didn't change. We still want to build amazing telecom systems, it just took us a while before people would pay for them.
Advertising is a beautiful business model that can work very well. I mean, it's obviously working for a LOT of people. And very well.
The only reason I don't want to do ad-supported products is because I believe in delivering value that users are willing to pay for. That's the kind of products I want to build. Simple as that.
Something something, business and user interests aligned.
> If the answer involves advertising, think again, chances are it’s a poor business model. Now you might say that [insert your favourite startup here] did it and succeeded, but let me warn you, those are the exception to the rule.
Perhaps. But the startup that succeeds is the exception to the rule, REGARDLESS of their model for making money. And I have seen quite a few that worried first about getting customers and only later about how to make a profit off them.
It's not that making a profit is unimportant (it is incredibly important, and also very hard). It is that making something that is successful is ALSO incredibly important, and is even harder. If you can solve the harder problem (being successful), perhaps later you can solve the problem of being profitable also. If you can solve both from the start that's even better, but not everyone can do that.
I don't think you can just separate these two issues. If you do not monetize your solution it will often be more attractive and thus the traction you get will be misleading. For example, I'm sure I could make a very popular product by simply giving away terabytes of free storage and bandwidth exposed through a simple API and web interface. But that would do nothing but attract the kind of takers that would never pay for anything while it hemorrhages money.
If you hate drunken frat boys, don't open an Irish pub.
Everything you do attracts a certain kind of person. Better do it deliberately, or you'll screw yourself.
Actually I have a post planned for the market fit as well - but yes, I mostly agree with you. But the problem arises when you try to fundraise (and many startups do) - then you'll have to have a very clear idea about making money. Based on my experience, it's pretty much the only thing investors care about.
What you say is both very true and very false, it depends a lot on the product.
Advertising can be an amazing way to get money and sometimes, even if you could directly sell to your users, advertising can perform WAY better.
The real important factor to decide wether you should put ads or not is "will I probably get a few paying customers or just a lot of users ?".
Advertising is all about the volume, the bigger it is, the more money you'll make and a good relationship with your advertisers can lead to really good revenues and a real business model.
When we launched our first product with my company, our only competitor was a paid app. We thought "f* it, we'll make it free!". We now have 600k monthly active users and make way much more than our competitor. So maybe not every business should "sell". With social media &cie, it's starting to be really easy to get a lot of users and ads can be an amazing way to monetize it !
This advice is reasonable for most types of startup though not particularly those in media where advertising is still a leading and viable form of monetization (principally because most media companies start from and focus on the question: how do we serve advertising efficiently and profitably? It's not a side project.)
Slapping ads on a regular webapp will rarely yield much income without high amounts of traffic, but savvy media companies who can target and deliver high impact advertising aren't crazy to lean on it as their primary form of income (although diversifying into subscriptions and products ASAP is a good idea, nonetheless).
(principally because most media companies start from and focus on the question: how do we serve advertising efficiently and profitably? It's not a side project.)
That right there is why a lot of programmers these days fail tomake any money with advertising. They see it as an add-on, rather than the product itself. Advertising is the product.
The users are the product. Advertising is how you sell them. If you think about the advertising model as selling your users/visitors then the high volume required becomes more obvious. You don't sell an ad slot for $1k. You sell a user for a tenth of a penny.
Fred Wilson covered the validity / opportunity / challenges associated with advertising as a business model in a recent post. The whole thing (and the comments) are informative but here's my high level summary;
- Huge difference between ads that are sold (niche) vs ads that are bought (scale).
- It takes millions of users a month to be successful at scale. It takes a great sales team to be successful in a niche.
- The absence of scarcity in online and mobile advertising creates structural downward pressure on CPM.
Another issue cropping up is eyeballs moving from desktops to mobile devices. The CPM for mobile is significantly lower than desktop which puts more financial constraints on a company that is only ad powered. These are real pressures affecting $FB and $GOOG right now. Based on KPCB (1) research, CPM for desktop is $3.50 and for mobile it's $0.75.
Why was this guy listening to an audience to begin with? People don't know what they want, and that's entirely different from what they will actually pay for. The only measure is the cash register, i.e., the reconciliation report from the credit card processor(s).
Make something awesome that clearly takes away more pain than the price incurs, more cheaply than you can produce it. If you can't make it awesome or profitable, then it's not a business.
The most important thing when creating your business model is to understanding precisely who you're customer is. Often, someone creating an advertising thinks their customer is a the person viewing the page. In fact, this is just a sourcing effort.
This post is right on, but it's kind of telling that a post where the TL/DR is "Your business idea is not good unless it makes money" can make it this high on Hacker News.
The success of advertising in your app depends on how valuable a service you provide. As value increases, so does the viability of an ad based revenue model.
FXXK ads. Ads will give no money to you!!! Nowadays, a 3 years old child knows how to use Adblock to block ads. People do not care about how beautiful your ads are, they just want a free website without ads. If your website or application could not sell itself, just give it up. Ads really do not help.
When you're getting involved in a startup, it is incredibly important that you be honest with yourselves and your partners about what the goal really is (this is to the point made the other day about self-awareness) and what you're really building.
A Feature: A piece of a piece of a puzzle. "It's like Facebook's News Feed, but instead you get audio updates instead of text updates!". Hackers can be guilty of this often - focused obsession on improving the world (or a piece of it) by tackling even the smallest problems. Not necessarily a bad thing, but not necessarily a guarantee of success.
A Product: Instagram is a product. It's full-featured, it's beloved (or was ;), it's free-standing, it works, and people use it. But do people pay for it? Would they? Would they make money by themselves or only in the context of a bigger company (like Facebook). Twitter is also a product, along these lines.
A Business: 37signals is a business. People pay them to use their products. They are free-standing, they are profitable, and they don't _need_ outside investors (need vs. want is different, btw - sometimes it makes sense to take strategic investment).
There's nothing inherently wrong with any of the above paths, btw. But you need to understand which path you're on, because you'll do a better job of it that way. If you decide to be on the feature/product path and hope to get acquired/acqui-hired, that's fine - it's a legitimate approach (that works, at least for now) and folks have made $billions doing it (see: Instagram & everyone else).
But it's a different path than starting a business. Goals are different, operations are different, and outcome is different.
The point, in the end, is that often the response - "well, I guess we'll sell ads" is what people who are starting features or products often say to convince themselves or others that they're on a worthwhile _business_ endeavor. And that's often dishonest (intentional or not).
Understand the path you're on and I'm willing to bet success will be more likely.