You are currently browsing the archives for September, 2004.

Does Online Brand Advertising Work?

Earlier this month I attended i20 Events’ conference, Marketing in a Digital World. The discussion was further evidence that behavioral targeting

is stepping up to give paid search a run for its money as the darling of internet ad spending. Wouldn’t it be great, attendees and panelists alike asked, if marketers could deliver their messages only to prospects who have signaled they are ready to buy? Think of the efficiency!

I am in full agreement that paid search and behavioral targeting have revolutionized direct marketing. The poor firms who are trying to eke a business out of list rental and physical direct mail face a grave threat. But what is it going to take to convince marketers that demand creation — not just demand fulfillment — is just as important in today’s digital world as it was in yesterday’s analog world?

In yesterday’s world, brand-building advertising was even more important than direct response advertising. According to Thomas Weisel Partners, advertisers spent $156 billion creating demand with advertising on TV, radio and print in 2003 versus $77 billion to fulfill on demand with direct mail, classified ads and yellow pages listings (US Market Data, 2003). Paid search and behavioral targeting don’t eliminate the need for vendors to create demand among prospective customers, the need to reach customers who haven’t yet begun to consider a new product.

Of course, this discussion must start with acknowledgement that brand advertisers, just like their direct-response counterparts, expect more accountability than ever before. Response will be part of the equation going forward. Period. So let’s figure out how to measure activities — beyond the click activities — that suggest brand-building at work.

Last spring, Adidas ran their “Impossible is Nothing” video spots on Yahoo. According to Steve Hall’s post on MarketingVox, the campaign not only improved ad recall, it also drove a 125% spike in searches for the keyword “Adidas” on Yahoo’s site (MarketingVox). Increased brand awareness, then, can drive meaningful activity not captured in a click-through report.

I looked at a similar phenomenon among the top enterprise server vendors. CNET’s BT Trax, which monitors users’ consumption of content on CNET’s business technology sites, shows that we published more server-related news coverage about IBM than Sun, HP or Dell — and IBM led in news coverage every quarter for the last year. For two of the last four quarters, IBM also led the pack in terms of readership of those stories: Of total reader-requests (RFIs) on server-related stories by visitors to, ZDNet and TechRepublic, IBM ranked #1. But in the most recent quarters (Q2 and the first 2 months of Q3 2004), Sun — still ranked #2 in news coverage — took over the #1 spot in terms of RFIs.

Part of this readership spike for Sun may be the new story Sun is telling. According to Ted Smith, CNET’s resident research guru, “Sun is center-stage because of their announcements and discussions about chucking the lock-box on Java and Solaris code, and for updating their Linux OS desk-tops. The other players simply didn’t shake it up as much as Sun during Q2″ (ZDNet)

But it appears that Sun’s advertising also played a role. To support this new story, Sun launched a brand advertising campaign that included several CNET sites. NNR’s AdRelevance, an imprecise source on dollar amounts but generally thought to be a fair gauge of directional activity, provides a quarterly estimate of ad investment by IBM and Sun on, ZDNet and TechRepublic over the past year.

Sun’s PR pitch may be skewing toward higher-impact stories, but their share of server coverage — quantitatively — has been consistent over the past 12 months: 13% of server coverage every quarter, just behind IBM. In Q1, following a 4th quarter (2003) with no ad investment against CNET’s business tech audience, Sun’s RFIs represented 12% of servers readership. Then Sun stepped up advertising in Q1 and Q2 2004. In the second quarter, while Sun’s news coverage was quantitatively the same as Q1, their RFIs shot up 92% from the prior quarter — bumping IBM out of the #1 position. Sun maintained the top share of RFIs into Q3.

IBM’s advertising, according to AdRelevance, was heaviest in Q4 2003 and Q1 2004, with a sharp decline in Q2 and Q3. Reader interest in IBM news coverage follows the same path — first up, then down — as advertising investment in the prior quarter. Like Sun, IBM’s share of news coverage in the server category was solid and consistent. But IBM’s share of reader RFIs increased in Q1 (following substantial ad investment in Q4 2003), went up further in Q2 (following a Q1 ad investment that was larger than Q4), and then declined in Q3 (following a precipitous drop-off in Q2 ad spending).

Correlating ad spending to audience interest was, not long ago, an art form guided primarily by gut instincts — light-heartedly expressed in the old “I just don’t know which half of my advertising worked” line. We are certainly not yet ready to abandon artistry from the practice of brand advertising, but real-time data mining is adding some science to the mix. Crossing AdRelevance spending reports with BT Trax data on reader interest doesn’t take into account qualitative factors such as the specific content of the news stories, or the degree to which one headline is more eye-catching than another, or even the relative effectiveness of ad creative. But when you put more (and more) of these case studies alongside one another, a picture is emerging. Good brand advertising drives audience response in ways other than click-through, and we’re gaining the ability to measure it.