Fulgoni: The Bottom Line Is Online Ads Work For Branding And Sales
- by Laurie Sullivan @lauriesullivan, July 31, 2009
Get rid of the click as the
de facto standard to measure the success of an online campaign. It's outdated and doesn't represent real success. So says Gian Fulgoni, chairman and co-founder of comScore, at the MediaPost OMMA
Metrics & Measurement conference in San Francisco, Calif., Friday. In the keynote, Fulgoni told attendees the Internet is far more successful in increasing sales. And it may be the most measurable medium, but not everything measurable matters.
So, what now? Fulgoni says advertisers and marketers need to forget the click, focus on the sales impact on campaigns and conduct post-buy analysis. They also need to realize that display ads help search advertising succeed and vice versa. Don't forget the power of creative display ads. Online branding campaigns can be effective. Internet advertising has had an impact on retail that is on par with television.
Advertisers and marketers just want some type of metrics that show the online campaign reaches the demographics and promised target segments. Fulgoni says that in any new medium, it's easy to make promises that exceed the ability of the technology. And to some degree, Internet advertising has done that. "One problem is, it's too easy to exaggerate the promises and claims that can be delivered," he says.
Fulgoni believes the industry has failed to eliminate the click as a metric because it once worked well, and search has had great success as a measurement tool. While the click rate for display ads has dropped to 0.1%, this does not mean that online ads don't work. It just means the metrics that the industry uses are wrong.
Still, far too many CMOs think the click matters. "If we want to move those dollars from traditional advertising to the Internet, we can sit here and complain about the percentage of time spent online and not getting our fair share until we're blue in the face, but I don't think that will influence the CMOs of big companies," Fulgoni says. "The CMOs need to know before they change the way they spend money that they will not have degradation in their ROI. Because God forbid it happens -- all hell will break loose."
There are major problems with using the click to measure the success of online ads. A comScore study of Yahoo and DoubleClick cookies that was conducted several years ago revealed that 30% of Internet users delete cookies in a month. A recent study suggests that rates are up 10%. The original metrics did not take into account that people use multiple computers. Nor do they take into consideration that people now delete five different cookies for the same site in a month on one computer.
Cookie deletion creates major problems -- such as 2.5 times overstatement of unique visitors in a server log, depending on frequency of site visitation, as well as up to 2.5 times overstatement of reach and a similar understatement in frequency in ad server logs.
Traditional brand marketers still want metrics on creative, reach and frequency. The cost-effectiveness of online advertising will speak for itself if the industry can determine and better understand the impact that campaigns have and use the correct metrics.
So, comScore combined traditional metrics with a type of ad-serving technology from Microsoft Atlas to create metrics that looks at ad placement. It combines Microsoft's ad-serving data with demographic information from comScore's panel. The Reach and Frequency Planner, or RF Planner, aims to help advertisers determine and predict how consumers will respond to their digital ads.
Richard Huff, senior analyst at Atlas Institute, who leads the project for Microsoft, told attendees in a later panel discussion that it's critical to have an accurate audience measurement tool, and it's important that the industry finally realizes it's time to make a change.
Several comScore studies have confirmed that online campaigns drive offline sales, according to Fulgoni. In the first study, comScore took four categories, 53 brands and 200 of the most trafficked sites. The company looked at people exposed to display advertising and what they did in the month following. Findings reveal that 18% searched on the brand advertised and 29% went to the advertisers' sites. Consumers who were exposed to the display advertising spent 55% more time than the average visitors to these sites the next month. The rise in time spent is matched by a similar increase in page views -- about 51%.
Rather than use the click as a metric, comScore looked separately at the impact of how search and display advertising might change consumer behavior, as well as search and display working together. During the four-week test, the research firm analyzed the increase traffic display ads give to brand sites. The patterns were consistent across automotive, finance, CPG, retail and apparel, media and entertainment, electronics and software, and travel.
Then, comScore analyzed the impact that online campaigns have on retail sales by matching the name and the address of consumers to retail loyalty card databases. The supermarket Kroger, for example, has issued about 60 million loyalty cards, which provide a massive data set to understand the degree that online search and display campaigns drive retail sales. The findings suggest a lift that is five times stronger when people are exposed to search ads alone, compared with display. Search alone produces an 82% lift, compared with display at 16%, and 119% when search and display are combined. About 82% of online ad campaigns measured by comScore have generated an average lift of 22% in CPG brand sales in retail stores.
Sure would be nice if Gian had something new to say.
very interesting piece. Is there similar data for b2b advertisers and buyers?
thedigitalhobo:
LOL. Sorry if you've heard my speech before.
But you need to know that a Forrester survey just found that 35% of marketers are still using the click as their primary metric for evaluating the effectiveness of their branding ad campaigns. So, clearly my message hasn't sunk in with everybody. I have more work to do. More reach and more frequency of this particular message is in order.
Thank you for this article. It is spot on the money and I cannot emphasize enough how important it is for advertisers and interactive publishers alike that we continue to bring attention to the fact that we are not properly measuring the effectiveness of branding campaigns by focusing on clicks and other related metrics.
@thedigitalhobo. If Gian has to keep "banging the gong" because way too few in the online word are listening, then I say keep banging it Gian - only bang it harder!!! Someone much wiser than me once said (some attribute it to Einstein) "not everything that should be measured can be, and not everything that can be measured should be". With the plethora of inaccurate or meaningless metrics that abound in the online world, that certainly applies!
I do not believe clicks should be removed from the lexicon of today's marketer as the metric has true value in the proper context. However, the problem is that click measurement compresses the entire branding and sales process of: 1. Awareness, 2. Attitudes, 3. Favorability, 4. Intention, 5. Action, and perhaps even 6. Preference into a single event.
It’s unreasonable to think consumers go from awareness to action (or Step 1 to Step 5) in a single sitting, which is how many uninformed marketers apply the click metric today. In reality, it takes time, multiple messages, multiple exposures, and even multiple forms of media to drive a consumer to take any action.
I sense the key here is to remind marketers of the branding and sales process, and provide a clear map of the metrics that are most effective based upon the objective at hand. Clicks are a good metric to measure Step 4 (Intention) to Step 5 (Action). So, let’s not kill the metric as it is doing its job.
Let’s remind marketers what they already know and apply for television, radio, OOH, and other offline media: pick the right brand metric for the task at hand. Ironically, a "Right Metric for the Job" campaign targeting our own marketers may pay big dividends here!
I couldn't agree more - it's time the industry realizes that display ads are not Direct Response vehicles. And consequently, they should not be measured as such. While CTRs are useful for measuring trends, they are not good indicators of the impact the ads are having.
Case study: we recently ran a 1-month campaign (June 2009) for a regional retailer in which we isolated one market for heavy display advertising. The results were overwhelmingly validating: test market showed a 14% lift in sales vs. comparable markets, and based on in-store surveys, the Internet ranked higher than TV as the medium that influenced their visit to the store. We also found that 90% said they had been to the retailer's web site before coming to the store. Display CTRs were nothing to write home about but site traffic for the test market more than doubled. We also found the average person who hit the local store info page had seen 4-5 display ads.
IMHO the ultimate metrics for measuring the impact of display advertising are: 1) revenue and 2) awareness / interest (which is a predecessor of revenue). CTRs are interesting stats, but not much more.
http://twitter.com/stevelatham
Its good to see Gian championing this issue and as with anything it needs repetition to be driven home. At Fuor Digital we are seeking to not go entirely beyond the click as it is a good metric to be included in the overall analytics of a campaign, but it definitely should not be the only measure of the success of online media. As an example, in a recent campaign in the financial services industry we found, with path-to-conversion reporting, that 39% of our funded accounts came from over 10 view throughs and not one single click. To have optimized out of those placements based on click data would have cost our client much revenue.
Let's not dismiss the click altogether. Clearly a click is an expression of interest. Not many people accidentally click on display ads! And many if not most studies today indicate even exposure to a banner ad may have some affect on brand lift. And more and more studies confirm that attribution needs to go beyond the last click (whether search or display). The lift obtained by the combination of search and display has been documented as far back, or maybe further, as the Yahoo "Closing the Loop" Study of years ago. Gian's point (at least to me) is that too many people still ONLY use the click as their metric. We all (should) know by now that there are dozens of other metrics that can and should be considered in the ultimate success of a campaign.
I agree with Gian. CMOs need to look beyond CTR if they want to look at the true ROI from online.
If you think about the purchase funnel, it starts from brand awareness to favorability to intent to action. Display advertising, the main form of advertising on social networks, is very powerful in creating brand awareness and moving the consumer to a favorable stage.
Click or Search advertising plays a big role in converting those leads into intent and action.
We will be shooting ourselves in the foot if we don't look at cross channel synergies and impacts.
Cheers
Nitin
http://unscramblesm.blogspot.com/
Here's something new to say, and do: move beyond the focus on impressions and clicks to a new and rigorous focus on discovers and shares.
You can do it. You can measure it. You can optimize it. You can make your clients' "ad" dollars go a lot farther.
Just to make it more complex we need to keep in mind that the way you present your brand to get those clicks has impact on the prospects' attitude, intention, and buying decision, both now and into the future. You can get lots of clicks (through use of certain words and visuals that some marketers sink to) while at the same time destroying the perceived value and image of a brand.
I have to take exception to Gian's proposition. Proposing that measuring the click should be altogether eliminated simply perpetuates the errant notion that marketing efforts need to either be "branding" OR "direct response," but not both. This is old-school thinking. In reality, all programs have both components to them, and marketers and their service providers do them a disservice by only focusing on one set of metrics, as opposed to looking at the entire scope. The "click" in online advertising represents the dividing line between branding and direct response, and eliminating it leaves a gaping hole in any attempt to understand the full spectrum of the impact of a marketing effort. This is why the most sophisticated DMers analyze and model results based on a cross-correlation of response (or click) and conversion. To simply focus on conversion, you sacrifice volume - to simply focus on response, you sacrifice ROI. Matrixing the two together provides the balance and tradeoff analysis that marketers need to make decisions. Then adding in post-campaign analysis that focuses on traditional "branding" measures makes the picture complete.
Absolutely agreed- Time is the new currency, you must give people something to spend to time with. Branded communities and live online interactive entertainment that has no pre, mid or end roll but incorporates the product into the show is the key. http://www.pookyamsterdam.com/2009/07/5-reasons-to-use-second-life-as-media.html Carpe Diem
Great article. One question, I read the research deck and for the findings that suggested that of those exposed to online advertising 18% searched the brand advertised and 29% went to the advertiser's site, what were the numbers for the unexposed? I am trying to understand the differential.