Monday, August 17, 2009

Can we make make all advertising useful, relevant, and helpful?

I have a new post at blog@CACM titled, "Is advertising inherently deceptive?"

It discusses some of the moral and ethical qualms I have when working on personalized advertising. It attempts to start a discussion around the question of whether personalized advertising will be used for good.

An excerpt:
Let's say we build more personalization techniques and tools that allow advertisers and publishers to understand people's interests and individually target ads. How will our tools be used? Will they be used to provide better information to people about useful products and services? Or will they be used for deeper and trickier forms of deception?

Is advertising an industry fundamentally fueled by deception? Or is advertising better understood as a stream of information that, if well directed, can help people?
If you have thoughts on this topic, please contribute to the discussion, either here or over on the full post at blog@CACM.

Update: About one month later, in the October 12 issue of the New Yorker, Ken Auletta has an article, "Searching for Trouble", that describes a 2003 conflict between the COO of Viacom and the founders of Google on exactly this issue, deception in advertising. An excerpt:
[You want] salesmanship, emotion, and mystery. [Viacom COO Karmazin said], "You don't want to have people know what works. When you know what works or not, you tend to charge less money than when you have this aura and you're selling this mystique."

The Google executives thought Karmazin's method manipulated emotions and cheated advertisers.


Daniel Tunkelang said...

(Also posted at blog@CACM)

Greg, I feel like we covered this last October when you wrote a post last year entitled "Google describes perfect advertising", leading me and Jeremy to engage in a heated discussion in a comments, as well as inciting me to write my own rant entitled "Search is Not Advertising".

Links to both:

My argument in a nutshell (excerpted from my post):

Advertising is about selling the user’s attention to the highest bidder. Google has done more than anyone to make that bidding process economically efficient. But any utility that advertising proves to users is a means to an end. Advertising is all about the advertisers, and the advertisers only care about providing value to users in so far as their interests are aligned. Absent alignment, advertisers naturally look out for themselves.

Sumit said...


Sounds like the dilemma of a nuclear scientist, who is wondering what his work would be used for. The difference being here its privacy at stake.

As an employee, you can only have so much say in the matter. Do you really think you can walk away, even if you knew that the personalization wouldn't be used in the benevolent ways you want?

Manuel Simoni said...

I think that Google's (monetary) success with advertising leads us to frame advertising in too simple terms, like page impressions, when in fact advertising (or more generally marketing) is a communication process between a vendor and a buyer, with all its complexities. And surely not inherently deceptive.

Dave Winer wrote the insightful equation that "perfectly targeted advertising is just information".

This is where we have to start to reinvent advertising for the internet age. And ads on web pages will turn out to be a small part of the puzzle that is corporate communication on the internet.

So what am I proposing: basically Winer's solution to advertising: Narrate Your Work. It hardly gets more personalized.

Christopher Smith said...

Daniel, I think your view ignores the role that the user's experience plays in all this. If the user consistently finds that their attention was warranted, they are far more likely to pay attention to the next ad. Consequently, there is value to the ad network in trying to ensure that each ad view is valuable not only to the advertisers but also the user.

In that context, the "perfect" add to show a user might be the one with the best ROI (where their attention is what they are investing) for the *user*, because the ad network is trying to maximize the long term value of *all* the ad views for the user.

Daniel Tunkelang said...

Chris, I thought I addressed that point: "advertisers only care about providing value to users in so far as their interests are aligned". I meant that generally as applying to the individual advertisers and the ad networks / search engines / content providers.

The latter group can't afford to alienate users and thus lose a long-term shot at their attention. But that is not the same as saying the advertisers have the users' interests at heart. There are a lot of long-term relationships that nonetheless have an adversarial component.

I maintain that, at best, advertising in its current form is a necessary evil. Users generally accept it as a preferred alternative to paying for services. But it's a trade: attention for services. Better targeting is a win-win: less annoyance for users, higher-value attention for advertisers. But let's not kid ourselves about the cold economic equations.

Greg Linden said...

That's a good point, Daniel. Under your framework, then, the question becomes what can we do to make sure advertiser and consumer interests are aligned.

For example, auction mechanisms that reward good reputations (such as Google's Ad Quality score) might be one thing we could do.

This specific post is more about the question of whether advertiser and consumer interests will be already aligned as we move into personalized advertising. Will individual targeting be used in a way that benefits consumers?

If interests are not already aligned, then I'd think that would suggest that, as we implement personalized advertising, there also will have to be a fair amount of attention and work on mechanisms that might push incentives into alignment.

Daniel Tunkelang said...

Here's a simple alignment test: provide consumers (and by that I assume we mean users in general) with an above-board opt out of ads--or, better yet, make the advertising opt-in. If consumers truly benefit from ads (as opposed to simply incurring them as a cost), then they will vote with their virtual feet.

Too naive? Perhaps. But arguably it will raise the click-through rates and improve targeting enough to make up for the loss of volume. It takes ad quality measurement to the consumer by asking if he or she wants advertiser interest to be part of the filtering / ranking function determining what content he or she sees in response to an expressed information need.

I'll go one step further. If such the unworkability of an approach would make a strong case that the interests of consumers and advertisers are simply not aligned.

Greg Linden said...

Thanks, Daniel. To clarify, are you arguing that advertising has negative value for consumers? Or just the softer claim that advertiser and consumer interests often are not fully aligned?

On the opt-in for advertising, there already are opt-out mechanisms for advertising, the various ad blockers you can install. But the install base of those is pretty small. Why do you think that might be? How come everyone does not install ad blockers?

jeremy said...

Greg, Daniel:

Don't forget this bit of conversation that we had two weeks ago:

The gist of my argument is that having two separate ranked lists, one for organic results and one for ads, is by itself a deceptive practice.

Now, have patience with me for a moment, before you jump to conclusions. I'm not saying that *any single ad* is deceptive. The user may be looking for hot water heaters, and the ad is about hot water heaters. I am also not saying that there is any deception about whether or not a result is sponsored. I assume all ads are clearly labeled.

But those are not the only, and even the most important, possible deceptions. The biggest deception (imho) is what information is the most relevant.

And by splitting the results into two columns, you confuse (deceive) the user about what information is the most relevant.

For example, is the first organic result more relevant than the first ad? Is the fifth organic result more relevant than the first ad? Is the third ad more relevant than the second organic result? Who knows?

To deceive is to lead astray. A search results page with two separate lists succeeds in leading the user astray about which results are the most relevant, and in which order the user should peruse those results.

I fully understand that I am in the minority when it comes to this understanding of deceptive advertising. Most people say "oh, but it's clearly labeled" and stop thinking about it. But as an IR researcher, I am interested in the holistic user experience, the totality of the interaction with the search engine, from first query all the way to satisfaction of the information need. And from that holistic standpoint, having two separate columns is, clearly, deceptive. It misleads the user as to where the most relevant information is found.

Daniel Tunkelang said...

I'm willing to concede for the sake of argument that that advertising has positive value for consumers--at the very least, advertising allows consumers to trade their attention for services they would otherwise be unwilling to pay for. I'll even go further and concede that advertising is sometimes informative.

But I am certainly making the softer claim that advertiser and consumer interests often are not fully aligned. As for ad blockers, they're fairly exotic. I'm not aware of any that work with Internet Explorer--which, as some of us need to be reminded, is still the most popular browser. Also, many people have concerns, justified or not, about the reliability of ad blockers. For example, I've noticed that CustomizeGoogle doesn't always keep up with the latest versions of the Google UIs. Finally, though it might not be a factor for many, I suspect that all of these ad blockers violate sites' terms of service.

An interesting question is what would happen if opt-out were an above-board option (cf.

jeremy said...

Maybe we should perform an A/B test with ad blockers. Firefox, Safari, IE should all get together, and agree that, on half of the installations that get performed, ad blocking is off by default. On the other half, ad blocking is on by default.

Then, we can experimentally see whether or not people opt-in. And how many change their default, in both directions.

How 'bout it? Shall we petition these browser makers to do the test?

Rangachari Anand said...

When you get to the heart of the matter, blogs like Gizmodo are just one big advertisement for gadgets. However, I don't find such blogs objectionable in any way. Indeed, in your terminology, I find such blogs "useful, relevant and helpful". The things is that I voluntarily visit that site when I am in the mood to learn about new gadgets.

Daniel Tunkelang said...

Many magazines have the same characters: the ads--or the articles that often feel like advertorials--are the content. But I agree that they aren't objectionable, because that content is the reason you go to those information sources. You very much opt in to the ads.

That's not at all the case for most advertising associated with ad-supported search engines and web sites. If it were, then those sites would surely make advertising opt-in, knowing that everyone would opt in.

Joe McCarthy said...

I'm not sure how useful or relevant this is - seems like much of the discussion is focused more on search engine advertising rather than social network advertising - but a potentially interesting data point is provided in a September 2008 Washington Post article by Rachel Beckman entitled Facebook Ads Target You Where It Hurts (subtitle: "My Facebook thinks I'm fat"). I don't know useful or relevant the "muffin top" ad was to Ms. Beckman shortly after she changed her status to "engaged", but they were certainly not welcome.