Network News

X My Profile
View More Activity

The good old days of low-information advertising

One of James Fallows's readers is skeptical about the long-term economics of online advertising:

The ability to measure the way people engage with advertising online only makes it more worthless. Industry average for CTR (click-through-rate) is usually cited somewhere around .12-.15%. In my experience working at a network that had stricter than normal standards for intrusive ad designs, it usually proved to be about half that. Eight clicks per 1000 people that "viewed" an ad. Probably half of those were by accident. This is the metric that most buyers care about, and it just serves to emphasize how unengaging most of this stuff is. Though these metrics can't really used to compare online and print campaigns (since print is rarely trying to encourage some sort of direct action), but because this information is so readily available for online, it will be used.

What always interests me about this observation is the implication that offline advertising is similarly worthless. There are certain types of advertising that get treated as content: People scan the newspaper for department-store sales, for instance. But an ad for Tide? Or Intel? You notice it, but you don't care about it. But everyone placed those ads so everyone kept placing those ads. And hey, having your brand noticed was worth something.

Online, however, that "not caring" has become too undeniable. The tiny handful of click-throughs, most of them accidental, are almost an insult to the advertiser. The right comparison here is old medical techniques that were widely used until information arrived showing them worthless. Advertising worked for everyone so long as the limited supply made it look really high-value to advertisers and the absence of information let them fool themselves into believing it really effective. But at least with online advertising, there's so much supply, and so much proof that it's not very effective, that the prices paid by advertisers are coming into line with the amount of attention people pay to their advertisements. And that's grim news for industries that rely on advertising.

By Ezra Klein  |  May 14, 2010; 10:42 AM ET
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Tom Toles is worth a thousand words
Next: What will conference committee do to FinReg?


Click-thru is even more useless thanks to malware and redirects. Even if an ad interests me, I'm going to go directly to a company site by typing their address in the or if it's a company I haven't heard of before, do some basic google research before checking them out.

Posted by: koalatek | May 14, 2010 10:52 AM | Report abuse

This reminds me of a statement that was attributed to John Wanamaker (remember Wanamaker's Dept. Store in Philadelphia?) --
"I know that half of my advertising budget is wasted. I just don't know which half."

Posted by: RMBinGoldendale | May 14, 2010 10:54 AM | Report abuse

Returns to advertising can and have been studied (Google Scholar "returns to advertising"). It isn't hard to imagine how to do design an experiment that would be revealing (advertise differently in different markets, collect sales data, etc.)

I don't know the literature, but it is clear there is one. Thus, I doubt firms are operating in a zero-information environment off- or on-line.

Posted by: TheIncidentalEconomist | May 14, 2010 11:08 AM | Report abuse

Certain types of advertising are doing very well. Google Adwords, for example. Pricegrabber. I can't imagine that isn't pricing themselves out of the market, but they also performed when we used them . . . it's just that at .90 or 1.00 a click, they became too expensive to use. For a time, blog advertising and text-link advertising was good for creating "link juice" on a search engines like Google, but then Google changes their algorithms to ignore such stuff, and that sort of advertising dries up.

I've never thought much of banner advertising. All the banners I ever click on are niche adds on niche websites, because they advertise stuff that I'm interested in looking at already, otherwise I wouldn't be at the site (like, for example). I've clicked on maybe 3 ads on the Washington Post?

In any case, I've never bought anything from any of them. If I know I want to buy something, I go to or eBay or do my own research on The Google.

I pay very little attention to ads on television. They have never influenced my buying decisions to my knowledge--at least, not since I was 12. I don't recall ever having bought a car I saw advertised. I have been to some of the furniture stores I've seen advertised--but that's about the only case I can think of. Even when I use national brands, it's not because of their advertising. And I usually end up using a non-advertised brand, based on research. I'm interesting in entertainment stuff, so I usually know about movies and TV shows that interest me before I ever see any advertising. When I saw the ad for KFC's new double down, I thought they same thing I thought about their Barf Bowls (or whatever they are called): that sounds interesting, I'd like to try that. And I haven't done it. And, come this time next year, I probably still won't have done it.

I used to work at an ad agency, and everywhere I've worked (until I started workin' for the gummint), advertising was always a major issue. On the web, what worked best was consistently organic search (and low prices, which led to word-of-mouth promotions), and email advertising (pushing our low pricing). Unless we had unusually low prices to advertise, advertising didn't seem to do much good.

We also found that investment in good package design often payed off, because impulse purchases were heavily influenced by packaging. Indeed, many brands could benefit from investing in compelling packaging and getting good shelf space in a variety of stores, and forget media advertising entirely, and be better off.

Posted by: Kevin_Willis | May 14, 2010 11:09 AM | Report abuse

But the print ads have *zero* clickthrough. How can they justify their rates for print ads?

The claim that the value of online ads is wholly from their conversion rate is not true. Just like print ads, a good portion of their value is in establishing and maintaining brand awareness.

You don't necessarily look for a Tide ad in a newspaper, but you do see it, and then the next time you're in the supermarket, you may buy it. The fact that you've seen the brand recently encourages you to think it's a popular, healthy business and worth trying.

If Tide were to stop advertising, I bet you'd think when you see it in the store, "Tide? I thought they went out of business. This must be the supermarket's leftover stock. Let's try Whisk instead."

So it is with online advertising, to some extent. The fact that clickthroughs are possible doesn't mean that's the only way the ad does its magic.

Posted by: billkarwin | May 14, 2010 11:22 AM | Report abuse

I would only say that the premise is a bit flawed based on the source and narrow scope of online advertising being looked at. Yes, standard ad units have low CTRs, but an ad network person only sees that part of the advertising world because its all they have to offer (its like running a low end apartment complex and bemoaning the fact that a premium price can never be attained or sustained).

The online advertising world is much broader and more effective (in fact, ad networks really are the most brute force, low value offering out there compared to every other offering). Partnerships, search, etc. all play more advanced and effective advertising roles in the space. In my current work at WebMD we employ a very different model that has been very successful for our clients and partners...but its far removed from the world being described by this source and how the Washington Post might be monetizing its offerings.

Just a thought.

Posted by: chackney | May 14, 2010 11:22 AM | Report abuse

Klein and Fallows both make good observations on the topic of online advertising -- reminds me to keep reading both of them. I notice myself scanning the paper for department store sales, food sales, and restaurant specials.

From my own standpoint as one who advertises hospitality services (among other things), I've found that the greatest impact comes from ads placed in the freely-distributed weekly "happenings" papers that sit near the door of most bars/clubs/etc. From time-to-time, we do a clip-n-save coupon, which lets us know who's reading what... and it's amazing how many people pick up those weeklies, read them, and tear out & retain the parts of interest. It's the only form of advertising we've found that truly crosses metropolitan boundaries: a reader in DC, for example, might actually clip the coupon for the NY location and visit it.

There have been a few tries at online "happenings" publications, but even in the iPhone era most folks seem to turn to the printed paper product to help find the next club/bar/restaurant to visit. I've wondered why this is true and if it will continue -- if you're bored in a bar and have an iPhone, why not flip its pages rather than those of a paper?

Posted by: rmgregory | May 14, 2010 11:33 AM | Report abuse

"You don't necessarily look for a Tide ad in a newspaper, but you do see it, and then the next time you're in the supermarket, you may buy it. The fact that you've seen the brand recently encourages you to think it's a popular, healthy business and worth trying."




Click-throughs actually give advertisers a pretty decent gauge of the immediate impact of their ads -- something that can only be done indirectly in other forms of media.

However, there are considerations that matter beyond just that one immediate purchase (e.g. encompassing ideas such as brand-awareness).

Posted by: JPRS | May 14, 2010 11:40 AM | Report abuse

The thing ad buyers forget is that people notice only what they're interested in. My newspaper had no ads for appliances until my refrigerator broke, then it was filled with them. Of course, the ads were there all along, my brain just filtered them out as not worthy of attention.

And then as koalatek noted, the internet bad guys have made clicking on ads even more problematical. You never know if you'll end up in a mousetrap, and have to close your browser to get out. So, as s/he suggested, if you see an ad you're interested in, you're more likely to google it and come in that way than by clicking.

This is not to say that "brand" ads can't be effective. I used to sell local advertising. All my prospects would tell me that their customers found them in the yellow pages. The question then is, what made the customer pick them? People most often pick the name (or product) that seemed familiar. How do you make your name/product "familiar" for when that customer is finally interested in it? By making sure they've seen it before. Whether it's on a billboard, in a newspaper, on tv, or on the web, if they've seen it (and even if their brain didn't deem it worthy of attention *at that time*, it did note the information), it will be "familiar" when they're actually looking for what you're selling.

It's my guess that, other than ad words type advertising (where you know the person is interested in your product/service), web advertising is most effective as a "name recognition" promoter. Not measurable in terms of click through, but not useless either.

Posted by: KarenJG | May 14, 2010 11:51 AM | Report abuse

Newspapers (as opposed to websites) don't have to contend with AdBlock Plus. I haven't seen an ad online in weeks.

Posted by: wiredog | May 14, 2010 12:00 PM | Report abuse

Who doesn't block internet ads and own a DVR nowadays? The only ads I see are on Hulu, and I mute them and tab to another page if they over 15 seconds.

I actually get angry now when I even see adwords or other non-obtrusive ads on websites; because it has been so long since I have browsed with ads, pry 1998. The first time I had to deal with an annoying flashing ad I discovered how to deal with it by using a combination of an ad blocking proxy server and plugins for my browser which further reduces ads. Sorry, but the pop-ups, flashing banner and esp ads with sound have made more and more customers go ad-free instead of having to deal with such outrageous annoyances to view content.

Posted by: rarianrakista | May 14, 2010 12:40 PM | Report abuse

Unspoken here: many frontline ad buyers are not all that smart or experienced. Otherwise they'd be doing something else. So if you give them a number they think they can understand, it's like giving them rope.

(And yeah, I'm another one of the people who pretty much never clicks on an ad. I'll type a company's url into the address bar before I click on one of their ads.)

Posted by: paul314 | May 14, 2010 1:05 PM | Report abuse

Since the beginning of the Internet and the Web, advertisers and their agencies have assumed that online viewing was somehow like offline viewing, and was going to be susceptible to natural extensions of their normal practices.

That assumption was false then and is still false now.

The online experience — its speed, breadth and depth — is active. TV, radio, movies, newspapers, magazines are mostly passive. Viewing online moves quickly, as viewers follow the links, check the suggestions, search through Google, Wikipedia, Amazon — all in a constant, continuous flow. As much as multitasking is touted, people think and view in a linear fashion for comprehension. Interrupting that flow promotes confusion and dissatisfaction.

In such an environment, online advertising is seen as much more a nuisance than traditional 30-second TV commercials or 60-seconds radio spots were for their media. Once upon a time, TV viewers were willing to put up with three minutes of advertising for ten to fifteen minutes of uninterrupted show. The same with radio. But now TiVo and DVRs have upended that model, just as Last and Pandora and Sirius have altered the traditional radio model, and as news aggregation sites have withered the traditional newspaper experience. Netflix and Hulu and the like are doing the same for TV and the movies.

Advertising is simply to too slow for the online world — especially for the favored 18 to 35 demographic. Who among them wants to be dragged away by a banner or popup link from the page they've just navigated to? Cognitio interruptus.

Advertising techniques take time for cognition to work, if only seconds, which is why it excelled in passive media. Google Adwords work because they are short, direct, quick — and not very interruptive. A viewer can actually spend a few seconds contemplating and judging whether to follow on or ignore them. But either way, the suggested links are more closely related to what brought them to that page, and much less infuriating in the way they flag attention.

Until online viewers slow down — not very likely with the young and restless — it just won't pay to advertise there. Never has. Never will.

Posted by: tomcammarata | May 14, 2010 2:07 PM | Report abuse


What you say may be true with respect to pop-up ads and the like. However, on Ezra's blog I've seen ads this morning for Pfizer, Sprint, and Toyota next to the blog itself. I'm not sure how this is any different than other media -- especially print.

Now it may be that these ads are so unobtrusive compared to those in other print media that the awareness of the ads is lower -- something that can be measured via scientific random surveys.

However, I would be very surprised if the awareness level registered at zero.

Even if the level of awareness on a per person basis was lower than that of a print ad, it would still have some value for the companies paying for the advertising.

With respect to the Google AdWords ads, the real value is that they tend to be highly targeted based on the search terms a person uses. The only cost incurred by the advertiser occurs when someone clicks through an ad. With more passive ads on media websites you have color and designs and logos and the like, which provide a different kind of value (e.g. the focus is on brand awareness -- planting a seed in consciousness for a future purchasing decision -- over immediate conversion rate).

One way that online publications could probably increase the awareness of ads and the click-through rates would be to put the ads on the far left column of the screen, so that they are in a dominant position (the first place the eye tends to go in part based on the way we are conditioned to read). It seems almost uniformly that web publishers put ads on the far right column, which probably diminishes some of their value for advertisers.

Posted by: JPRS | May 14, 2010 2:36 PM | Report abuse

All ads are useless. Period. End of Story.

Posted by: mabkhar | May 14, 2010 3:22 PM | Report abuse


As you noted, Google Adwords tend to be highly targeted.

That's the key—relevance. In online advertising, it's the critical factor for slowing down the viewing process to take in the message and possibly interrupt it long enough to click through on the ad. But if the interruption was then a pop-up window that didn't break the viewing flow because it can be quickly closed, it might have a better chance of succeeding in capturing the interest and reaction of the viewer.

In the old days — i.e. pre-2000 say — relevance what a major function of media placement along with the ad's message. Advertisers sought media that was either wildly popular (scattershot in its demographics), or relevant in to consumers' interest (tightly targeted).

The only difference now is that online advertisers don't pay for the waste if their ads aren't clicked. Traditionally, media was bought based on CPM, or Cost Per Thousand viewers. Relevance of the media to the viewer's interests and the relevance of client's message increased odds of an ad "working". That made it worthwhile paying for ad space or time that was only going to be acted on by a fraction of the potential viewers. For the online world, relevance is even more important, because most websites are all about relevancy to some topic.

As you also point out, positioning on the left side of the page could increase notice of ads or ad words, as opposed to the right side. The problem here is that web viewing — while reading left to right — is also linear in the up and down direction. The viewer wants to stay focussed on what's being viewed, which means continuing to read down and down, and away from the advertising. If you notice the position of Google Adwords, they are often immediately below the page's main content, like Ezra's blog entry or on top of Google's search results, but either way vertically placed.

Maybe it's because I'm older (40 years in the ad biz), but pages cluttered with many columns and blocks, whether online or off, make it harder to deliver relevant content. That's why Google, for example, has always avoided cluttering up it's home page. It's the epitome of relevance. (Remember the Portal wars of the late '90s?)

Relevance — interest — can slow and capture attention, especially if it's delivered in a less cluttered way. But contemporary web design mostly ignores that in favor of delivering as much content as fast as possible. Where's a weary eyeball to go?

Posted by: tomcammarata | May 14, 2010 4:11 PM | Report abuse


With respect to the location of the ads on media websites, when I open my browser or go to a webpage, it always opens with the same template. It would be possible to position an ad in the upper left-hand corner, which would ensure that an ad was seen every time someone navigated to a web-page. As a web-marketer those are the kind of things I'd be curious to do a test on to see if it had any kind of impact (from the perspective of a media company too it might be worth an experiment in the hope that higher response rates would make it easier to sell the space for higher rates).

I agree with you about the clutter aspect. Looking at a web-page is a different experience from looking at a full-page spread in a glossy magazine. The audience is nowhere near as captive as they are with a TV ad or radio ad (one reason that media sites are probably trying to drive more traffic to web-videos -- in theory these videos should have a higher value for advertisers than an ad on a web-page -- even one with moving elements).

My experience is nowhere near as extensive as yours -- I spent about 4 years with a direct marketer when web advertising and things like e-commerce sites were just becoming the norm. My knowledge of the current state of affairs is already out of date. I know that the tracking capabilities are much more sophisticated than they once were, but I don't know what the norm is for sales generated via web ads versus other forms of media. That kind of information would probably inform research into new methods.

Posted by: JPRS | May 14, 2010 6:18 PM | Report abuse


I'm sure web advertisers and web agencies have analyzed placement, motion, size, etc. and have tons of statistics, just like the traditional media does.

But the one stat that outranks all others is sales — of product for the advertiser, and revenue from ad sales for the medium. Neither is terribly impressive (around 5% of income for newspapers, for example, which is why they are looking at subscription models).

Some products are naturals for the Web, just as they are for infomercials. And the Web is a great place to buy and sell. Witness Ebay and Yahoo and Craigslist as they replace the newspapers' old cash cow of classified advertising.

But I fear there is no magic that will ever make the Web as potent a selling medium as TV was and still is. Mass marketing is still requires mass audience, which TV delivers. The Web is too fractured in subject and interest to provide that kind of mass viewing.

Facebook and Myspace may claim huge users bases, but that not the same as the millions watching Dancing with the Stars or Lost. That's why media buyers still recommend national TV buys backed by local spot buys.


Posted by: tomcammarata | May 14, 2010 7:27 PM | Report abuse


Thanks for the insights. That helps to clarify the situation a bit.

The main thing with micro-targeting via the web would seem to be the cost to advertisers. Whether it will ever be as lucrative for the big ad buyers as TV seems doubtful for the reasons that you've highlighted. Although clearly there is some cost for advertising on the web that companies can justify if the ad rates are right (i.e. sufficiently low enough).

In terms of consumer research, in the case of Facebook, I suspect that their data would likely be valuable for most direct marketers. The appeals may not be done over the web, but certainly in terms of compiling mailing lists, it would seem to reduce a lot of the guess work that goes into creating direct mail pitches based on much less granulated data. That doesn't really resolve the issue though of news media websites. I suspect some form of a subscription service is probably the way most will ultimately go.

Posted by: JPRS | May 14, 2010 9:15 PM | Report abuse


You make two key points: "micro-targeting via the web" and "reducing the guess work."

They have always been the holy grail of marketing and every form of advertising, from TV to direct mail. Who are the prime targets? Where are they? How best to reach them?

The individual nature of the web, from email addresses to web addresses, has long promised greater ease and less cost in reaching desirable consumers. But the fractured nature of the web has also made it difficult to reach them in the kind of great masses that TV or a national ad campaign delivers, whether in magazines, outdoor, or by DM.

Of course, as more and more of Generation Y gravitates to the web and away from traditional advertising media, these become less effective without the Web advertising necessarily compensating for their decline in terms of eyeballs reached.

That's been the paradox of the web that's frustrated marketers for the last two decades.

Posted by: tomcammarata | May 15, 2010 3:39 AM | Report abuse

Online advertising as it has existed offers very little to the individual consumer. That's why dynamic display advertising - which intuitively learns about each consumer, tailoring ads to the user - will be so important. It reverses the trend of consumers "not caring," as Klein writes about here.

Alex Parker

Posted by: AlexParker | May 15, 2010 6:19 PM | Report abuse

The advertising that was making the easy money in the newspapers was the classifieds. Ripping off the little guy. You hear Murdoch blaming Google, but it was craigslist that ate his afternoon tea.

As far as effectiveness goes, I once placed an ad selling my Honda Civic, listed at $2000 obo in the Washington Post for three days, starting on a Saturday. I started getting a call every 3 minutes at 5AM. By 9AM I had sold the car for $3000 and had to unplug my phone for the next two days. I plugged it in on Saturday night and it immediately started ringing.

The random nonsense ads for Macy's and what not in the rest of the paper are annoying noise. I mentally filter them out.

Posted by: staticvars | May 15, 2010 9:20 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company