6 ms·

Page views required to generate $1M in ad revenue?

60 points by thinkzig 17y ago

heed 17y ago
The thing that is difficult about ads, not many people like to be told what to buy. And I think it might be the ads themselves that are lacking, not necessarily how they are delivered. For example, look at this award winning Pringles online ad: http://awardshome.com/cannes2009/pringles/can-hands.html http://awardshome.com/cannes2009/pringles/can-hands.html The reason why this one is great, is because it isn't pushy, it has humor, and it uses normal, honest language that we can all relate to. Maybe it's these type of ads that need to be displayed in the social networks.
- il 17y ago
  Your comment is so very incorrect. It's not that _ads_ as a whole don't work, it's that untargeted, random banner ads are very ineffective at driving conversions. Ads on search, which frequently tell you what to buy, and don't have room to be artsy of humorous generate billions of dollars for Google and probably much more than that for its advertisers. I wouldn't be surprised if Google gets $20-$100 CPM on its pageviews. Award-winning ads are frequently not the most effective. The most effective ads are often bland, boring ads that tell you what to buy but take into account segmentation, positioning, targeting, behavioral trends, and so on.
  - heed 17y ago
    I don't think a downvote was warranted here, but do what you may. Could you provide a source to your last sentence? I'd like to read more about this. Thanks.
    - jimboyoungblood 17y ago
      Take a look at any trade magazine or other specialty publication. In general the ads you'll find there have very low production values. Award-winning ads and creative branding are necessary when trying to sell highly undifferentiated, commoditized products (e.g., potato chips). But if you have a product that is obviously differentiated from the competition (a laptop with 100 hour battery life, an electric sports car, etc.), informing consumers of your existence is often enough.
      - imp 17y ago
        I got the impression that ads in specialty publications and trade magazines only have low production values because the companies advertising have small budgets.
lallysingh 17y ago
Hmm, not even counting the cost of content creation, how much is it going to cost to produce 1M in ad revenue? They mention 2.5 billion views for $1M. Being very, very conservative in data transfer, that's several terabytes a month. 4k/page is 10 TB (3.858 MB/sec sustained for a month), 32k is 80 TB (30.864 MB/sec for a month), etc. A single banner image off of gamasutra.com (not even flash) was 149k. The netlink, server hardware, and infrastructure (bought or rented ala rackspace) isn't exactly cheap.
- diego 17y ago
  This is highly dependent on what your site does. If you could somehow get that much traffic on a blog, the cost of bandwidth and servers would be very low (tens of thousands at the most, say 5% of your revenues). On the other hand, if every page view is extremely heavy with a long tail of uncacheable content (e.g. Youtube) or computationally expensive (web search engine), then bandwidth and servers become a significant percentage of the revenues.
- vaksel 17y ago
  if you are making a million/mo in revenue, you can afford to spend $100K/mo on servers/bandwidth.
- shpxnvz 17y ago
  If you are doing 2.5 billion page views a month, very little of your content is likely to be coming from your own network. The vast majority is likely to be delivered from a content delivery network like Akamai, at a much lower cost than hosting it all yourself.
staunch 17y ago
> But, based on a survey of social network sites let's assume an average CPM of $0.40. You would need 2.5 Billion page views per month to earn $1M in ad revenues. That is 2,500,000,000 page views...and how many sites can sell out all their page view inventory? This makes the mistake of assuming that one page impression = one ad impression. That's not the case on most sites. Most sites can quite easily have 2-4 ad impressions per page impression. Also, it's not hard to sell out ad inventory at ridiculously low CPMs. There's always someone willing to pay something for a genuine ad impression. Most sites can probably sell a little bit of their inventory for a high CPM and the rest for a very low one.
pjhyett 17y ago
Trying to make that sort of money using something like an Adsense program is going to take a hell of a lot more pageviews over having a sales force capable of selling high-dollar ad packages. That isn't to say that sites like PlentyofFish haven't been able to do it, but your CPM has to be fairly high if you're not getting billions of pageviews a month. Sourceforge doesn't get nearly the level of traffic that Facebook does, but they brought in $5.4 million in media revenue from 36 million uniques in their first fiscal quarter of '09[1]. 1. http://www.globenewswire.com/newsroom/news.html?d=155376 http://www.globenewswire.com/newsroom/news.html?d=155376
- blasdel 17y ago
  Isn't a ton of Sourceforge's ad sales direct to Microsoft, for their classic "I lowered my TCO with Server .NET Live 200X Enterprise!" campaign?
- int2e 17y ago
  Be careful not to confuse uniques with impressions. Assuming 10 impressions per unique, that's still a very respectable CPM of $15.
  - pjhyett 17y ago
    That's the point, though, $15 CPM is enormous compared to the article's claim that the social network average is 40 cents.
    - jacquesm 17y ago
      I can confirm that claim. This is over a sample of about 3 million uniques / month.
pmichaud 17y ago
I think it's pretty clear that the best way to make money from a site is to sell things. For money.Crazy, I know.
- jpwagner 17y ago
  Granted I'm sure your comment is facetious, but I do want to point out that it's a bit of a naive view. The best (outside of unethical methods) way of making money is to generate true value for a customer and allow a medium for that customer to transfer value (in the form of money) to you. In some cases that customer is the end-user, but in many it is not: If you make a tool for students, it is the parents or teachers paying for it. If you make a tool for employees, it is the employers paying for it. If you make a tool for people to surf the web, it is the advertisers paying for it.
  - joez 17y ago
    This is not my mind's definition of true value. True value would be if students wanted to pay for it themselves, employees themselves, people themselves. Price something correctly (not cost prohibitive) and if you really have a true value building product then people should be willing to purchase it. If you had a better version of Excel, you could sell it to companies for lots... or sell it to the workers. I'd say you had a better product if you had a product capable of selling to the workers. And while this may not be as much of a cash cow as large corporations, I'd think this was closer to True Value. But even that is not "true value." I'd think true value is something where you are generating a win-win. An example that came to my mind was Mint.com. Their business model (along with more retail advertising) involves giving money saving suggestions (credit card offers, better mortgage offers) and then charging the financial institution less than what it is charging them to generate a new referral. A "true value" is a win-win... not something like, "well I guess I could use a male enhancement pill, Google... I'll click your ad."
    - jpwagner 17y ago
      By "true value" I mean that the product creator has a positive ROI on his time, and the product consumer has a positive ROI on his money. This does not mean that the user must be the buyer. A clear example is a productivity tool (product ABC) that Employer X buys for its employees. The employees save enough time to offset ABC's cost and therefore Employer X saves money by buying the tool. The creators of ABC used sufficiently small hours creating a reproducible product and therefore make a profit. The less clear example is that of advertising. Google sells real-estate and guaranteed views and has a positive ROI. The ad company pays for the real estate and gets sufficient hits and hopefully has a positive ROI. You get to use Google's services with some distraction which you find to ultimately save you time, so you also have a positive ROI.
russell 17y ago
300 million in page views net $15K in revenue. Not a business I want to be in. Google makes a very nice living because people who are googling are looking for information and the ads provide links to information. People who are socializing are not looking for something, they are already there. Most of the time the ads are even below the level of annoyance. (Mouseover videos are definitely over the threshold.) CPC vs CPM doesn't address the issue. Radio and TV ads work because you cant avoid them. Print ads works because people do scan newspapers in search mode. The real issue is how to engage people when they are socializing. The advertiser needs something more engaging the current social event. Sex sells and flash sells, but the post is talking about something beyond that. I dont think target advertising is much more than the same old same old, but more refined. The break through will be something that enhances the social experience. Maybe something like a chat about Paris brings up a live feed of a Paris cafe complete with tables for your avatars, with an urchin trying to sell you a travel guide. Whatever it is it needsto be engaging an not annoying.
- vaksel 17y ago
  pretty bad example, considering the people might be chatting about Paris Hilton.
  - jasonlbaptiste 17y ago
    pretty sure sites like perez hilton are getting great CPMs and advertising offers. It may be annoying/not to your liking as a site owner, but tv shows "reskin" the site on premiere nights. That costs a pretty penny.
    - vaksel 17y ago
      I was talking about contextual ads
    - fallentimes 17y ago
      You're right - Perez Hilton is making an absolute killing: http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revision_create_shortcut?persistent_uid=3e759f53402fc5bafd783046580df71f http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revisi... http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revision_create_shortcut?persistent_uid=2bf3ca8b27a840dd60c8a1170cc175aa http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revisi... And there's multiple occurrences of those ads. 40k-75k PER week isn't that far of a stretch
brc 17y ago
Google CPC works because it ties in buyer behaviour with advertisers. The solution to driving revenue from pageviews must either connect up existing behaviour with advertisers (where can I buy an X), or create a compelling new behaviour. For example, in the early days of newspapers, it wouldn't have been obvious to look in the back for classified ads, but buyers learnt this and now know to look in the back for the ads. The power in social networks is the group of trusted friends. I'm sure some sort of group question along the lines of 'do you guys think I should buy an LCD or Plasma TV' with voting, or something similar, would develop new behaviours, particularly amongst consumers of products they aren't completely familiar with (anything from cars to financial products). Once you develop this behaviour in a social network, selling ads into it is a cinch.
Ardit20 17y ago
Last time I analysed revenue from ads per visitor the figure was between 0.0025 and 0.004, so 1000 views means 2.5 dollars at the low end or 4 dollars at the high end, so maybe the problem is limited to social networking sites alone. How can they make money? Perhaps freemium is the best way? I mean these people, well some of them, well most of them are addicted to Facebook and all the rubbish in it such as quizzes and friend poking, so if it costs say £2 to I don't know, send your friend the perfect girl questionnaire, sure some people will pay and if it goes hot and viral then all of them will pay, they have to.
- bmelton 17y ago
  Where are those numbers from?
  - Ardit20 17y ago
    just my own website
    - NonEUCitizen 17y ago
      which ad provider do you use?
pmorici 17y ago
What sites like Facebook and Twitter need to do is infer peoples needs from things that are happening to them in the moment. For example, if a user posts to their feed "My POS car broke down" They should be shown ads for new cars, repair shops, car rental and AAA. If a FB user invites 30 of their friends to a Halloween party they should all be shown ads for costumes and other related things. FB and Twitter have the ability to preempt Google because they could predict your needs from your updates before you even know them yourself. As it stands though Facebook advertising strikes me as being about as effective as putting up a billboard on the side of a highway.
- seldo 17y ago
  For certain types of advertising, I think Facebook's ability to target by demographic must make it significantly more cost-effective (and hence popular) than keyword searches -- products for a particular age range, or relationship status come to mind -- despite the lack of intentionality (i.e., a keyword search is somebody looking for something, but a Facebook user is not).
eli 17y ago
Uh.. audience matters. Like, a lot. If you think all impressions are equal, then you will only be able to attract lowest common denominator advertisers who will only pay you commodity prices. Why do you think washingtonpost.com has ads for asbestos lawyers? If your site is for luxury yachts owners, or people looking to buy Enterprise CRM systems, or is specially tailored to Fortune 1000 CEOs, you can charge CPMs that most sites only dream of. With a dedicated ad sales staff You could probably clear a million with a few hundred thousand page views a month.

60 points by thinkzig 17y ago

heed 17y ago

The thing that is difficult about ads, not many people like to be told what to buy. And I think it might be the ads themselves that are lacking, not necessarily how they are delivered. For example, look at this award winning Pringles online ad: http://awardshome.com/cannes2009/pringles/can-hands.html http://awardshome.com/cannes2009/pringles/can-hands.html The reason why this one is great, is because it isn't pushy, it has humor, and it uses normal, honest language that we can all relate to. Maybe it's these type of ads that need to be displayed in the social networks.

il 17y ago
Your comment is so very incorrect. It's not that _ads_ as a whole don't work, it's that untargeted, random banner ads are very ineffective at driving conversions. Ads on search, which frequently tell you what to buy, and don't have room to be artsy of humorous generate billions of dollars for Google and probably much more than that for its advertisers. I wouldn't be surprised if Google gets $20-$100 CPM on its pageviews. Award-winning ads are frequently not the most effective. The most effective ads are often bland, boring ads that tell you what to buy but take into account segmentation, positioning, targeting, behavioral trends, and so on.
- heed 17y ago
  I don't think a downvote was warranted here, but do what you may. Could you provide a source to your last sentence? I'd like to read more about this. Thanks.
  - jimboyoungblood 17y ago
    Take a look at any trade magazine or other specialty publication. In general the ads you'll find there have very low production values. Award-winning ads and creative branding are necessary when trying to sell highly undifferentiated, commoditized products (e.g., potato chips). But if you have a product that is obviously differentiated from the competition (a laptop with 100 hour battery life, an electric sports car, etc.), informing consumers of your existence is often enough.
    - imp 17y ago
      I got the impression that ads in specialty publications and trade magazines only have low production values because the companies advertising have small budgets.

lallysingh 17y ago

Hmm, not even counting the cost of content creation, how much is it going to cost to produce 1M in ad revenue? They mention 2.5 billion views for $1M. Being very, very conservative in data transfer, that's several terabytes a month. 4k/page is 10 TB (3.858 MB/sec sustained for a month), 32k is 80 TB (30.864 MB/sec for a month), etc. A single banner image off of gamasutra.com (not even flash) was 149k. The netlink, server hardware, and infrastructure (bought or rented ala rackspace) isn't exactly cheap.

diego 17y ago
This is highly dependent on what your site does. If you could somehow get that much traffic on a blog, the cost of bandwidth and servers would be very low (tens of thousands at the most, say 5% of your revenues). On the other hand, if every page view is extremely heavy with a long tail of uncacheable content (e.g. Youtube) or computationally expensive (web search engine), then bandwidth and servers become a significant percentage of the revenues.
vaksel 17y ago
if you are making a million/mo in revenue, you can afford to spend $100K/mo on servers/bandwidth.
shpxnvz 17y ago
If you are doing 2.5 billion page views a month, very little of your content is likely to be coming from your own network. The vast majority is likely to be delivered from a content delivery network like Akamai, at a much lower cost than hosting it all yourself.

staunch 17y ago

> But, based on a survey of social network sites let's assume an average CPM of $0.40. You would need 2.5 Billion page views per month to earn $1M in ad revenues. That is 2,500,000,000 page views...and how many sites can sell out all their page view inventory? This makes the mistake of assuming that one page impression = one ad impression. That's not the case on most sites. Most sites can quite easily have 2-4 ad impressions per page impression. Also, it's not hard to sell out ad inventory at ridiculously low CPMs. There's always someone willing to pay something for a genuine ad impression. Most sites can probably sell a little bit of their inventory for a high CPM and the rest for a very low one.

pjhyett 17y ago

Trying to make that sort of money using something like an Adsense program is going to take a hell of a lot more pageviews over having a sales force capable of selling high-dollar ad packages. That isn't to say that sites like PlentyofFish haven't been able to do it, but your CPM has to be fairly high if you're not getting billions of pageviews a month. Sourceforge doesn't get nearly the level of traffic that Facebook does, but they brought in $5.4 million in media revenue from 36 million uniques in their first fiscal quarter of '09[1]. 1. http://www.globenewswire.com/newsroom/news.html?d=155376 http://www.globenewswire.com/newsroom/news.html?d=155376

blasdel 17y ago
Isn't a ton of Sourceforge's ad sales direct to Microsoft, for their classic "I lowered my TCO with Server .NET Live 200X Enterprise!" campaign?
int2e 17y ago
Be careful not to confuse uniques with impressions. Assuming 10 impressions per unique, that's still a very respectable CPM of $15.
- pjhyett 17y ago
  That's the point, though, $15 CPM is enormous compared to the article's claim that the social network average is 40 cents.
  - jacquesm 17y ago
    I can confirm that claim. This is over a sample of about 3 million uniques / month.

pmichaud 17y ago

I think it's pretty clear that the best way to make money from a site is to sell things. For money.Crazy, I know.

jpwagner 17y ago
Granted I'm sure your comment is facetious, but I do want to point out that it's a bit of a naive view. The best (outside of unethical methods) way of making money is to generate true value for a customer and allow a medium for that customer to transfer value (in the form of money) to you. In some cases that customer is the end-user, but in many it is not: If you make a tool for students, it is the parents or teachers paying for it. If you make a tool for employees, it is the employers paying for it. If you make a tool for people to surf the web, it is the advertisers paying for it.
- joez 17y ago
  This is not my mind's definition of true value. True value would be if students wanted to pay for it themselves, employees themselves, people themselves. Price something correctly (not cost prohibitive) and if you really have a true value building product then people should be willing to purchase it. If you had a better version of Excel, you could sell it to companies for lots... or sell it to the workers. I'd say you had a better product if you had a product capable of selling to the workers. And while this may not be as much of a cash cow as large corporations, I'd think this was closer to True Value. But even that is not "true value." I'd think true value is something where you are generating a win-win. An example that came to my mind was Mint.com. Their business model (along with more retail advertising) involves giving money saving suggestions (credit card offers, better mortgage offers) and then charging the financial institution less than what it is charging them to generate a new referral. A "true value" is a win-win... not something like, "well I guess I could use a male enhancement pill, Google... I'll click your ad."
  - jpwagner 17y ago
    By "true value" I mean that the product creator has a positive ROI on his time, and the product consumer has a positive ROI on his money. This does not mean that the user must be the buyer. A clear example is a productivity tool (product ABC) that Employer X buys for its employees. The employees save enough time to offset ABC's cost and therefore Employer X saves money by buying the tool. The creators of ABC used sufficiently small hours creating a reproducible product and therefore make a profit. The less clear example is that of advertising. Google sells real-estate and guaranteed views and has a positive ROI. The ad company pays for the real estate and gets sufficient hits and hopefully has a positive ROI. You get to use Google's services with some distraction which you find to ultimately save you time, so you also have a positive ROI.

russell 17y ago

300 million in page views net $15K in revenue. Not a business I want to be in. Google makes a very nice living because people who are googling are looking for information and the ads provide links to information. People who are socializing are not looking for something, they are already there. Most of the time the ads are even below the level of annoyance. (Mouseover videos are definitely over the threshold.) CPC vs CPM doesn't address the issue. Radio and TV ads work because you cant avoid them. Print ads works because people do scan newspapers in search mode. The real issue is how to engage people when they are socializing. The advertiser needs something more engaging the current social event. Sex sells and flash sells, but the post is talking about something beyond that. I dont think target advertising is much more than the same old same old, but more refined. The break through will be something that enhances the social experience. Maybe something like a chat about Paris brings up a live feed of a Paris cafe complete with tables for your avatars, with an urchin trying to sell you a travel guide. Whatever it is it needsto be engaging an not annoying.

vaksel 17y ago
pretty bad example, considering the people might be chatting about Paris Hilton.
- jasonlbaptiste 17y ago
  pretty sure sites like perez hilton are getting great CPMs and advertising offers. It may be annoying/not to your liking as a site owner, but tv shows "reskin" the site on premiere nights. That costs a pretty penny.
  - vaksel 17y ago
    I was talking about contextual ads
  - fallentimes 17y ago
    You're right - Perez Hilton is making an absolute killing: http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revision_create_shortcut?persistent_uid=3e759f53402fc5bafd783046580df71f http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revisi... http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revision_create_shortcut?persistent_uid=2bf3ca8b27a840dd60c8a1170cc175aa http://web.blogads.com/adspotsfolder/ba_adspotsfolder_revisi... And there's multiple occurrences of those ads. 40k-75k PER week isn't that far of a stretch

brc 17y ago

Google CPC works because it ties in buyer behaviour with advertisers. The solution to driving revenue from pageviews must either connect up existing behaviour with advertisers (where can I buy an X), or create a compelling new behaviour. For example, in the early days of newspapers, it wouldn't have been obvious to look in the back for classified ads, but buyers learnt this and now know to look in the back for the ads. The power in social networks is the group of trusted friends. I'm sure some sort of group question along the lines of 'do you guys think I should buy an LCD or Plasma TV' with voting, or something similar, would develop new behaviours, particularly amongst consumers of products they aren't completely familiar with (anything from cars to financial products). Once you develop this behaviour in a social network, selling ads into it is a cinch.

Ardit20 17y ago

Last time I analysed revenue from ads per visitor the figure was between 0.0025 and 0.004, so 1000 views means 2.5 dollars at the low end or 4 dollars at the high end, so maybe the problem is limited to social networking sites alone. How can they make money? Perhaps freemium is the best way? I mean these people, well some of them, well most of them are addicted to Facebook and all the rubbish in it such as quizzes and friend poking, so if it costs say £2 to I don't know, send your friend the perfect girl questionnaire, sure some people will pay and if it goes hot and viral then all of them will pay, they have to.

bmelton 17y ago
Where are those numbers from?
- Ardit20 17y ago
  just my own website
  - NonEUCitizen 17y ago
    which ad provider do you use?

pmorici 17y ago

What sites like Facebook and Twitter need to do is infer peoples needs from things that are happening to them in the moment. For example, if a user posts to their feed "My POS car broke down" They should be shown ads for new cars, repair shops, car rental and AAA. If a FB user invites 30 of their friends to a Halloween party they should all be shown ads for costumes and other related things. FB and Twitter have the ability to preempt Google because they could predict your needs from your updates before you even know them yourself. As it stands though Facebook advertising strikes me as being about as effective as putting up a billboard on the side of a highway.

seldo 17y ago
For certain types of advertising, I think Facebook's ability to target by demographic must make it significantly more cost-effective (and hence popular) than keyword searches -- products for a particular age range, or relationship status come to mind -- despite the lack of intentionality (i.e., a keyword search is somebody looking for something, but a Facebook user is not).

eli 17y ago

Uh.. audience matters. Like, a lot. If you think all impressions are equal, then you will only be able to attract lowest common denominator advertisers who will only pay you commodity prices. Why do you think washingtonpost.com has ads for asbestos lawyers? If your site is for luxury yachts owners, or people looking to buy Enterprise CRM systems, or is specially tailored to Fortune 1000 CEOs, you can charge CPMs that most sites only dream of. With a dedicated ad sales staff You could probably clear a million with a few hundred thousand page views a month.