[ulvund]

On Facebook you buy ads for ~$1+ because you can set target demographic to something as specfic as "women; 42-47 years old; looking for relationships with other women; with a degree in biology or computer science; attened Harvard; Works at IBM; who likes horseback riding or skiing"

Imagine if google knew BOTH your search term and your complete personal history. Then the ad price and conversion would increase enormously.

That is what Google is trying to get a piece of.

[cdavid]

I am not convinced it is that useful to have explicitly targeted audiences for advertisement. It may be better to understand and interpret an ad campaign, but generally, what works best when doing statistical prediction is not what is the most intuitive.

Also, even though your example is obviously not meant to be taken as is, it shows that targeted demographics quickly don't have a lot of data behind them. Successful stories in AI usually involves lots of averaged data with only little "focused" data to adapt your model quickly (e.g. as done in speech recognition where models are estimated on 1000s of hours from many speakers, and the model is then adapted for the one speaker to be recognized).

I think the value of the so called social graph for advertising is overestimated. IMO, what's interesting about facebook is more the amount and diversity of data than its personalized nature. But then, I have little knowledge about algo for advertisement targetting, maybe the situation is different than the domains I am familiar with.

[redxaxder]

> "I am not convinced it is that useful to have explicitly targeted audiences for advertisement. It may be better to understand and interpret an ad campaign, but generally, what works best when doing statistical prediction is not what is the most intuitive."

If you've collected marketing data that tells you your conversion rate among pet owners is 2% when your conversion rate among the general public is 1%, then you know that every ad you show to pet owners is going to have twice the impact on your bottom line.

Ad targeting enables you to collect this kind of data and to exploit it once you have it. If you find the right audience, your advertising dollars can go more than twice as far.

[onemoreact]

Unless it cost your more than twice as much for that advertizing or your showing the same add to the same person several times (ex: hulu).

But, the real problem with facebook users is you are far less intrested in adds while on a social website and a lot of clicks are simple mistakes.

[uxp]

Buying a sponsored AdWords ad on Google results in people viewing it while they are actively looking to solve a problem or buy a product.

Buying an ad on Facebook that happens to be related to something they once had interest in, and may still do, will result in it being shown while they are not always looking to solve a problem or buy a product, but to entertain themselves, and most importantly may not care about that subject at that moment.

Online advertising is not a cut and dry field. You can get great results from either place, but you won't get the same results from both, at least in my experience.

[juliano_q]

The thing is that Google dont kneed to place ads on Google+, but they will know your personal information and place better targeted AdWords for each person.

[janesvilleseo]

Ding, ding, ding.....we have a winner

[kingnothing]

From my experience as a Facebook app developer for household brands, Facebook ads are one of the most effective methods of marketing today. Across all of the IP we manage, our company is behind over 800 million fans, likes, and shares, and a significant portion of that was generated from those ads.

[jasonparallel]

Have you been able to calculate the value of a like and a fan?

[rachelbythebay]

Who says it's about effectiveness of the ad? It might just be effectiveness of getting money from the advertisers. More data means you can do more hand-waving to say that this ad will be the one that really really reaches people this time.

[JanezStupar]

Because in the end if ads are not effective, you won't be getting money from the advertisers anymore. Marketers can't know in advance which marketing channels will work - so they kinda carpet bomb with their campaigns. However once they start getting data, they sure as hell kill of the channels that don't work.

Unethical behavior does not scale in business, no matter how much your calculations rationalize it. In the end people just go away from sleazy businesses. GM and Ford might tell you a story about it and millions of others short term profit oriented businesses also.

Theres only so much bullshit you can pile up before your reputation catches on to you.

[tokenadult]

The classic statement on advertising is attributed to John Wanamaker (1838 - 1922), a businessman who pioneered department stores in the United States: "Half the money I spend on advertising is wasted; the trouble is I don't know which half." As long as tools to analyze which advertising money works and which doesn't in an overall advertising campaign leave room for hand-waving, merchants will spend too much on the advertising arms race.

[mtts]

As much as I'd like to think you're right, you're probably wrong. The advertising industry is making very real money, which gives people an incentive to keep it going even though it delivers close to nothing.

Why do people buy Coca Cola? Is it because of the tremendously expensive advertisements they put out or is it because the stuff is ubiquitous, tastes fairly good and is usually not surrounded by a lot of competing products at the places where it's sold?

Don't believe me? What about if you're in a restaurant that sells only Pepsi. Do you leave the restaurant in search of another that sells Coca-Cola? Most people probably don't, thereby proving that all the so-called "positive emotions" advertisements instill in consumers do absolutely nothing when the time comes to make purchase. Theoretically, yes, in a situation where you can choose either Pepsi or Coca-Cola for (roughly) the same price, Coca-Cola's advertising campain may have an effect, but such situations are exceedingly rare. In most purchasing decisions there are variables that have much more significance than the advertising history of the products.

So ads are, probably, not effective. And despite this fact, there has been an industry selling them for over a century and that industry has been getting richer and richer, thereby proving that you can perfectly well build a business on scamming people and keeping it up indefinitely. You do, however, have to keep playing around with some superficial parameters of your business model to keep it convincing. Which is exactly what Google is doing here.

(btw, I use - and love - Google's services, but never in my life have I clicked one of their ads or even looked at one for more than a split second)

[JanezStupar]

I was talking of marketing in general. Advertising is only a single branch of marketing.

And its obvious that Coca-Cola's marketing has been waaay better than Pepsi's. And also I believe that Pepsi is quite comfortable in its #2 position. As is Burger King in its own.

Don't forget, that second biggest player in a huge market is usually still a friggin big behemoth. And that for Pepsi and BK being #2 is a core business strategy. Less upside, but also less downside. These companies basically use their bigger competitors as a hedge against market change.

[mtts]

Ok. I guess I was confused because the grand parent talked about advertising (Google's targeted ads).

Marketing in general, I agree, has some non-bullshit aspects to it. Not many, though.

But certainly the marketing that causes Coca Cola to be sold in way, way more places than Pepsi has been very effective.

[colomon]

Errr.... you have occasionally been in convenience stores and supermarkets, yes? Restaurants that serve both Coke and Pepsi are rare, yes, but stores where you can buy either are an everyday thing for most of us...

[mtts]

I'm guessing here, but I don't think the bulk of cola purchases are made in convenience stores and supermarkets.

[mitcheme]

At one point my college campus was a "Pepsi campus" and you could not buy anything but Pepsi. I think I bought it once, and I drink far more Coke than is probably healthy. That stuff tastes bad. I would bring my own Coke from home or buy water. If I'm at a grocery store and they're out of Coke, I will take the red label no-name cola over Pepsi or blue label cola. While we're at it, if people didn't care that much about the difference, I doubt my grocery store brand would make both kinds of cola.

[txxxxd]

It sounds like you're not very familiar with the industry.

One of the reasons Google's advertising programs are so successful is that they are measurable. A site can compute return on investment and know that for every $X spent on Google ad clicks they are generating $Y in sales.

[mtts]

I'd love to hear some figures on that, because the ones I've seen from various (large) online retailers were abysmal. Setting up adwords campaigns for these retailers was a fun diversion, in some cases also a good excuse to huff and puff about the future of online marketing, but not at all a significant source of revenue. It was done mostly "to keep an online presence", probably out of fear that if they didn't buy those adwords someone else would.

The retailers I'm talking about may not be representative, however: they already had a strong market presence and strong brand recognition before they went online. Maybe there are businesses that have started from scratch online that did see significant revenue from adwords, but I doubt it.

[msg]

The question Coke's ads answer is not just Pepsi Or Coke, it is To Coke Or Not To Coke. You always have the option of drinking water. Water dominates Coke as a beverage for health, it (perhaps) loses on taste and texture. Hence a lot of advertising stressing Coke's positives. How do you know that when you say it tastes fairly good that they haven't gotten to you too?

That's why we skip ads in my house, they are just too hot to handle.

[fttechfounder]

Excellent point here. There's a lot of advertising dollars to be spent, in fact too much. Superbowl advertisements certainly didn't help companies like Pets.com during the bubble, but man, did the NFL cash in on that one!

A lot of sales guys sell by telling stories, not necessarily by persuading metrics, and all the contextual data makes for GREAT stories.

[skarayan]

"I think the value of the so called social graph for advertising is overestimated."

I disagree, for certain businesses it is more beneficial then Adwords. I have had really good experiences with Facebook Ads.

[tdoggette]

I would guess that you were initially downvoted for a lack of specificity.

[skarayan]

I originally wrote from my iPad and I have a hard time using the touch keyboard.

Here's my 2 cents:

1) Cheaper than Adwords, much much cheaper. The ads tend to be expensive when you first get an estimate from Facebook, but in practice, this number goes down 5-fold. Also, in comparison to Adwords, it is at least 5-fold cheaper.

2) Easier to target people. This depends on what you are doing, but I believe it works better if you are trying to build a community with a specific interest. However, Adwords is probably better for targeting advertising for an e-commerce store.

3) If you advertise your fan page, you have a hook to constantly advertise to your fans. I like building my fan page, generating conversation, and sparking interest. It is important to give users good content, but it also doesn't hurt to self promote about 10-20% of the time. It all depends on what you are doing.

[scrollbar]

I am convinced that it is useful. I do it every day :)

Without going into specifics, think of this data as just another signal to use in ad targeting. Maybe users in certain regions are more likely to purchase a product? Maybe men versus women, people who have kids, people in certain income ranges. We ad people use all this info to eek out performance from campaigns.

Google is going for the holy advertising grail with +

[nivertech]

You think like a tech person:

- Google Engineer: If I'll better target demographics and increase conversion - I'll be able to raise my per ad price by 200%

- Ad Campaign Manager: better conversion rate? OK will pay max 15%-30% more per ad.

- Google AdSence Manager: we'll lose 80% of revenue if target narrow demographics - screw it

[eru]

If advertisement is paid per click, and not per impression, Google itself is interested in better targeting of what to display to whom, because it frees up inventory without impacting revenue.

[nivertech]

if you have less impressions, most likely you will get less clicks

[ordinary]

But you don't have fewer impressions. The number of ads you display doesn't change, you can display just as many targeted ads as you can untargeted ones. If each of the targeted ads has a higher conversion rate, the number of clicks goes up without ever changing the total number of ads you display.

[eru]

The number of impressions Google has to sell in total is determined by the number of pageviews. They still have to optimize what to show to whom. (And of course, make people view their pages more often.)

[rmc]

However, Google know what you're looking for, and all your past search terms. Although Facebook might have more demographic information, with google, you can target ads for people who are looking for dog worming tablets, regardless of who they are.

[officemonkey]

I just thought of something... With target demographics like that Facebook would be an excellent way to look for prospective mates.

[netcan]

Potentially, Google could get to the point where a single click is worth $100s.

[corin_]

As someone who works in digital advertising I would love that to be realistic, but I'm afraid it's not.

[netcan]

There are areas where it's already at close to that. In law & education for example, there are certain area/keyword combinations that can be turned into phone calls at a very high rate. Add in a few more filters (eg has 14 year old child) and that conversion rate goes up. Gogle already translates conversion rates in active markets into click prices very efficiently.

It would really be expanding the number of $100 clicks available and I think thats almost inevitable.