September 21, 2010

A Defense for Digital Advertising

NOTE: This post is old, and is probably on different subject matter than my current writing. It is possible the information is outdated or my opinions have changed. -- Josh Klein, May 28, 2012

The Top 10 Double-Secret Unknown Facts About Advertising! That’s the title of a provocative post from The Ad Contrarian, in which the author, one Bob Hoffman, tries to rescue the marketing and advertising industries, which he believeshave been hijacked by web-addled digi-maniacs who don’t know a fact from a fart.

Highjacked? Web-addled? Digi-maniacs?!

Bob has decided to throw out the baby with the bathwater. He claims to debunk an “age of hysteria”, but has in turn become a hysterical fanatic for the other side.

Don’t worry, I’m not going ad hominem here; I like Bob and what he usually writes. But his post about web folks ignoring facts to sell digital advertising got me riled up. It’s the folks in TV and print advertising that get away with ignoring reality more often than not.

I’ve opted to mount a defense for digital advertising by refuting Bob, point-by-point. I’ve read all of his sources and come to dramatically different conclusions from his post.

See, Bob manipulated the facts to suit his headline. I never said he wasn’t a talented marketer.

1) 99.9% of people who are served an online display ad do not click on it. (Source: DoubleClick, Benchmark Report, 2009)

Low click-through rate is the reason digital ad folks prefer rich media that doesn’t necessarily require click-through because you can interact without leaving the current page. The Doubleclick Benchmark Report also reports a 2.54% engagement rate, which is fairly dramatic considering the volume of online display ads.

But perhaps more telling is that 100% of people who are served a TV ad do not click on it.

Since the average display time for an online ad is 34.6 seconds, would that mean online display beats the 30-second spot?

2) TV viewership is now at its highest point ever. (Source: Nielsen Three Screen Report, Q1 2010)

The same source reports that 57% of those with access multitask on the Internet while they watch TV.

Still, I wouldn’t dream of disputing that Americans watch lots of TV. That’s why TV advertising is still a great way to generate broad awareness for things like Consumer Packaged Goods.

For the layman: CPGs are those fast selling, inexpensive products that people don’t think too hard about. Think toothpaste, detergent, soap, etc.

Not all products work like CPG. We’ll get to why this matters in a few moments…

3) 96% of all retail activity is done in a store. 4% is done on line. (Source: U.S. Department of Commerce, Q2 2010; Nielsen Three Screen Report, Q1 2010)

In its 2010 research paper, US Online Retail Forecast, 2009 To 2014, Forrester says online retail is more like 7%, but does the percentage matter as much as the fact that it’s a $172 billion opportunity?

In 2009, 154 million Americans (about half) bought something online.

Most importantly, Forrester’s research report on Web-Influenced Retail Sales Forecast says that in 2009, the web accounted for or influenced 42% of all retail sales.

Even when the purchase is made offline, the decision for what to buy often starts online. This is called “ROBO”, for “research online, buy offline.”

Think about your own experience buying anything non-trivial. What influences your purchase behavior? Do you start by searching Google like I do?

4) DVR owners watch live TV 95% of the time. 5% of the time they watch recorded material. (Source: Duke University, Do DVRs Influence Sales?)

Reading the study from Duke, I didn’t find anything related to the numbers quoted. Here’s what I did find:

By the second quarter of 2006, 18 months after the issuance of DVRs, the treated households spent 15% of their viewing time watching recorded content. In contrast, the national TiVo sample spent 22% of their viewing time watching recorded content. We observed a similar evolution in the use of the fast-forward function. In the field study TiVo sample, the use of fast forwards increased from 8.9 per day in Q1 of 2005 to 11.3 per day in Q2 of 2006. Over those same periods, the national TiVo panel’s use of fast forwards was 15.3 per day and 14.6 per day respectively.

So, even 4-years ago, DVR users were watching recorded content (which the study says does not include “near live” content where the show is paused, then fast-forwarded) 22% of the time, not 5%. And the average user fast forwarded around 15 times per day.

The study was designed specifically to measure whether DVR use reduced actual purchase behavior for advertised products. They found sufficient evidence to say “DVR use has no effect.”

However, they admit several faults of their study. Firstly, they only tracked Consumer Packaged Goods advertising. To be more specific, they only tracked 7 brands. Actually, let’s just be 100% honest: they only tracked household cleaning and personal grooming product advertising.

So, maybe it’s true: digital advertising isn’t a great way to market detergent.

But here’s the kicker…

The study also holds an interesting assumption; that “advertising stimulates demand,” therefore, “DVRs would … reduce demand for advertised goods ceteris paribus.”

First takeaway: “ceteris paribus” is Latin for “with other things the same,” which I didn’t know. Cool.

Second takeaway: Maybe the reason DVRs don’t alter purchasing behavior is because TV ads don’t alter purchasing behavior. That’s a scary thought.

5) 99% percent of all video viewing is done on a television. 1% is done online. (Source: Nielsen Three Screen Report, Q1 2010)

True! However, the person watching TV is probably a twelve-year-old who hit the “channel up” button before they got to your car ad. The person in the computer chair is probably someone looking for a car who just Googled “what’s the safest midsize SUV?”.

Not all advertising has to blanket entire demographic segments with high frequency interruptions just to find a tiny percentage of consumers that might be interested.

TV has a particular mindset that one does not need to apply to the Internet. To follow the underpants gnomes model:

Phase 1: Awareness
Phase 2: ?
Phase 3 Profit

With the Internet, you can actually fill in details for phase 2. Internet ads don’t have the reach of TV ads, but TV ads don’t have the targeting of Internet ads.

Be smart and use both.

6) The difference in purchasing behavior between people who use DVRs to skip ads and those who don’t: None. (Source: Duke University, Do DVRs Influence Sales?)

See #4, we refuted this already.

7) Since the 1990s, click-through rates for banner ads have dropped 97.5%. (Source: Li, Hairong; Leckenby, John D. (2004). “Internet Advertising Formats and Effectiveness”. Center for Interactive)

You’re doing it wrong. People who measure “clicks” as a success metric are the same people who measure “impressions” for their traditional ads; they don’t know about anything better, so they do the best they can.

Good digital advertising measures “CPA”, or cost per action. If you’re selling something online, you measure the conversion from impressions to clicks to visits to sales. If the transaction doesn’t end online, you do the best you can; measure contact form submissions, downloads of directions to your store, prints of technical questions to ask a salesman, and so on.

Good digital advertising demonstrates the quantifiable return on investment of the money you spend. If you want to do things like raise awareness with a compelling emotional story, measured by a qualitative survey of a focus group’s intent to consider purchase, you’re perfectly justified in turning to TV.

Also: 97.5% drop in click-throughs since the 1990s? Almost no Internet consumer behavior from back then is relevant today; no Facebook, no Google, no Youtube. And from a study done in 2004? I’d read the report, but it’s 5+ years too old for me to learn anything relevant to my work.

8) Since the introduction of TiVo, real time TV viewing has increased over 20%. (Source: NielsenWire, Nov. 10, 2009)

We’ve really addressed this one by now, but let me share some personal thoughts on DVR.

I think the main attraction of DVR is not that it skips ads; it’s that DVR makes it possible for me to set my own schedule for media consumption.

And as TV manufacturers make their products more like computer monitors (internet connectivity built in, on-demand Netflix, etc.), calling it “TV” and not “Internet video” is just semantics. I think we’ll someday see the convergence of TV and the Internet. Why? Because it’s the best customer experience.

9) Baby boomers dominate 94% of all consumer packaged goods categories. 5% of advertising is aimed at them. (Source: Marketing Daily, July 22, 2010)

Tsk, tsk, this is manipulating numbers. Here’s the paragraph where we get the 94%:

“Nielsen’s research says Boomers dominate 1,023 out of 1,083 consumer packaged goods categories…”

But here’s the paragraph where we get 5%:

While Boomers spend 38.5% of CPG dollars, Nielsen estimates that only 5% of advertising dollars are currently targeted toward adults 35-64 years old.

So Boomers “dominate” 94% of CPG categories (whatever that means), but only 38.5% of the dollars spent. And, again, we’re talking about only CPG.

But let’s ignore this error, and get to the core of my problem with this super secret stat: calling your target consumer “Baby Boomers” is about as useful as a platitude as saying “our marketing objective is to increase sales!”

Demographics are the best TV people could do, but that doesn’t mean it holds true for digital ad folks. Age and household income are useful for targeting only in the absence of the knowledge that 450,000 people Googled “best laptop” last month (Source: Google).

Maybe it’s better to try selling laptops to people interested in laptops, not “baby boomers”.

10) TV viewers are no more likely to leave the room during a commercial break than they are before or after the break. (Source: Council for Research Excellence, May 10, 2010)

Yes, but they’re also no less likely. See, according to the source, it turns out that 28% of people switch the channel or leave the room in the 4 minutes prior to watching a commercial break, 31% during, and 32% in the 4 minutes after the break.

Furthermore:

Multi-tasking was found to accompany about 45% of all media use. Concurrent activities are led by “care of another,” at 12% in the two minutes prior to and during commercial breaks; and “meal preparation,” at 8% in both cases.

So the takeaway from this source might also be…

Even if TV commercials worked, and people didn’t use DVRs to skip them, are they even watching their TVs when they’re on?

Update: You can read The Ad Contrarian’s angry response and my followup to that.

Update #2: And lest you think I hate TV, read my article: TV is awesome, too!

---

Comments are closed.