Display Ad and Search relationship Research

ComScore and iProspect studies on search and display ad relationships.

Key take aways:
– Though it depends on the industry, there is potentially an additional activity of 60% versus the direct clicks (31% click, 21% say they type in the URL – iProspect).
– Approximately 1/3 of users have clicked on a display ad in the last 30 days(ComScore) to 6 months (iProspect)
– Conversely, 2/3 of users do not click on any display ads
– about 16% of users make up 80% of the clicks from display ads (ComScore)
– CTRs are in general decline, at about 0.1% (ComScore)
– Display ad value is quantifiable beyond the CTR based on search and direct URL entry following ad exposure.

Display ad cost should be measured against the incremental value based on KPI lift factors. On average display exposure increases site visits from 4.5% (control group not exposed to display ads) up to 6.6% (test users exposed to display ads). In other words, the incremental lift is 2.1%. (ComScore). View-through, repeat visit tracking are important parts of media metrics. There is no way to properly assign value with out them.

Both studies are worth reading.
ComScore: How Online Advertising Works: Wither the Click?http://www.comscore.com/Press_Events/Presentations_Whitepapers/2008/How_Online_Advertising_Works_Whither_The_Click
iProspect:

Narrative:
iProspect released “Search Engine Marketing and Online Display Advertising Integration Study” this month. Though it is very thought provoking, it needs to be interpreted from the right perspective. Primarily, this is not a study about ads, it is a study about users. This distinction is important because if the presentation the numbers is not interpreted properly, it can lead to some erroneous conclusions.

To illustrate my point:

“The key message from this study is that online display advertising is far from dead — its 31% direct response rate confirms that,” said Robert Murray, CEO, iProspect.

When we look at response rates, we look at how many times our ads are clicked versus how many times they are shown, or the click through rate. What Robert Murray is referring to is that 31% of the people surveyed said that the had clicked on a display ad at some point over the past 6 months.

As I looked at the iProspect study, I recalled the ComScore study released in December 2008. It reviewed integration from an ad perspective and the user perspective rather than just the user perspective alone. I think this is important for several reasons:

1) Cost basis: most display advertising is still sold on a CPM. The value of a “user” has to be relative to the cost of the communication.
2) Industry: behavior varies greatly by industry.
3) Exposure: are users cognizant of how many exposures they receive before they react. This goes to cost basis.

When we look at each part of these two studies, we some commonalities:

1) Both studies found that roughly 1/3 of users clicked on an add (ComScore in the last month, iProspect in the last 6 months).
2) There are strong synergies between search and display advertising. ComScore showed a 38% lift in advertiser’s branded search after display ad exposure, while the iProspect study simply showed that, of those who said they saw an ad, the response of 27% was to conduct a search on the brand, product or category.

The divergence
As a person with roots in media, on and off-line, every time someone suggests buying more ads, or bigger ads, I ask several questions. Key among them is: What is the incremental value of spending the money?

This is where the ComScore study is more helpful. It measures the lift in KPIs, such as site visits, competitive searches, TM / Brand searches and incremental sales. Contrast this with the iProspect study, which is survey based, and depends on users recollections over a six month period, with no control group against which to compare the test subjects. If you want to know the real value of additional advertising, it has to be measured not in absolute terms, but relative to the outcome of not increasing the advertising. In other words, what was the incremental affect received from spending more money.

One of the interesting findings on the ComScore study is that there is a 45.7% lift in site visits as a result of exposure to display ads over a 4 week period. Of those not exposed to the advertising, 4.5% eventually reach the test advertisers site, while 6.6% of those who were exposed reached the site, either by clicking, using search or navigating to the site directly. Another way to read this is that 68% of the people who reached the display advertisers’ sites would have done so with or with out the advertising. So while the total visit was 6.6% of the users who saw display ads, these ads contributed 2.1% of the users’ visits.

What is important is that the results vary greatly by industry. From a low of 21% lift in the travel industry to a high of 114% in the auto (though with a very low base % of visits to start).

On the flip side, there is also an increase in competitors’ sites visits following exposures to display ads. Over a 4 week period, the lift is 23.4% (13.5% vs 16.6% of users).

In essence, what display advertising does is spark shopping activity in general.

In addition to the number of people who eventually reach the advertisers’ sites, the way they get there is important; it directly affects the core measurement of CTR. Every one I speak with about the impact of display advertising acknowledges that the click is only one way to measure the influence of display advertising, but they are usually lost when trying to measure non-click activity. The iProspect study shines a light on the other behaviors as reported by users. 21% said they typed in the advertiser’s URL, while 27% did a search on the product, brand or company. Combining this insight with the lift that the ComScore study shows, and you can get some idea of a factorization you can apply to the CTR to estimate

Leave a Comment