An analysis of multichannel marketing campaign performance

Your ad here, there and everywhere

Editor’s note: Molly Elmore is vice president, market research at InsightExpress, a Stamford, Conn., research firm.

Over the past 20 years media consumption has continued to fragment, creating a challenging and hostile media environment for marketers. And, as more advertisers buy integrated campaigns across multiple media channels, the need for holistic measurement becomes greater than it has ever been. There’s been no shortage of metrics to evaluate holistic campaign performance; however, data on the efficacy of a campaign across and within channels has been much more elusive. As a result, developing an understanding of how the various channels affect consumer perceptions, as well as how the channels can work together to further enhance the way people feel about brands, has become critical to savvy, data-focused marketers.

This article is meant to help close the knowledge gap by using our firm’s normative data to explore the specific effects of frequency on commonly-applied awareness and intent metrics for various advertising channels. All of the data was derived from the InsightNorms Cross Media normative database, which houses ad effectiveness data from 75 mixed-media campaigns that included advertising running in television, print, radio, out-of-home and the Internet.

Worthy of continued exploration

It is no secret that advertising works, but exactly how today’s complex media plans influence an audience is a question worthy of continued exploration. Often this question is addressed using cross-media research - an analysis that explores how the various media channels work in combination to build consumer awareness and preference toward an advertised brand.

As the foundation, cross-media measurement uses a design of experiments and an online survey to gauge consumer response to advertising. Using both behavioral tracking and questions about offline media consumption, these studies allow for analysis of media response segmented by respondents who were exposed to different parts of the campaign. More specifically, this methodology uses opportunity-to-see (OTS) metrics, which provide a way to classify respondents into exposure groups. A critical component of a cross-media study is the identification of a control or unexposed group, which provides a measurement baseline of people who are in the “footprint” for at least one of the media plans but who were not exposed to any of the measured advertising. Using this method we are able to isolate unique subgroups of the sample based on media exposure (e.g., didn’t see any ads, saw the online ads, saw the TV ads, saw both online and TV ads). By comparing the brand metric scores for people in our baseline sample (i.e., the control cell) to respondents exposed to various parts of the campaign (i.e., the test cell) we can measure the effect of exposure on shifting attitudes.

In isolation, these studies provide a diagnostic measure of campaign performance and diagnose the required media and creative changes needed to ensure campaign success. Yet it is an aggregate analysis across several campaigns that really starts to contribute knowledge and insight into the best practices and the true nature of how media channels function in a multimedia environment.

Most cross-media campaign measurements use brand funnel metrics to diagnose the impact of the campaign on consumer perceptions of the brand. Figure 1 illustrates a typical cross-media output for a campaign that included television, print and online. And while InsightExpress has measured the effect of many other media, for simplicity’s sake this analysis will focus on these three media and how they work together. The control group on the left of each pair of bars (in light blue) represents the brand response for our baseline group (control cell) and it is provided for comparison. The exposed group on the right of each pair of bars (in dark blue) shows people exposed to all three media channels, with a delta box above the x-axis showing the difference between the control and exposed groups. A green shaded delta box indicates a significant change in that metric at the 90+ percent significance level.

The chart reveals that, for every metric measured, a combination of media had a significant impact over no exposure to the advertising. People become more familiar with the name of the brand and are better able to recall having seen advertising for that brand. More importantly to some advertisers, exposure also enhances favorability toward the brand and makes people more likely to purchase the products or services being marketed. Many brands we work with are solely focused on driving intent since this measure is the closest proxy for increased sales.

To illustrate this point, the increase of 7.3 percentage points in purchase intent shown in Figure 1 can translate into many potential buyers. If one million people see the multimedia campaign, the exposure to the advertising would translate into 70,000 potential customers with an interest in purchasing.

Numerous questions arise

While it is clear that multichannel marketing efforts are effective at driving changes in brand perceptions, numerous questions arise around how each of the channels drive the changes observed and how many times someone must be exposed to each channel to see those changes. The first step in uncovering the effect of channels is to isolate the frequency component and look at a series of campaigns where people were only exposed to one impression for a given media channel. For example, they saw one television ad, one online ad, one print ad, etc. Since the campaigns in each media channel group vary, each group has its own control group average. This analysis relies on a comparison of the campaign deltas.

Our data (Figure 2) illustrates that after only one exposure, the print channel is most effective. Changes in advertising awareness, brand favorability and purchase intent show that only one exposure to a print advertisement can draw the attention of a consumer and persuade them to change their opinion towards a brand.

Two of the key benefits of print (particularly with magazines, which this data is mostly based on) are 1) there is often no other advertising on the page at the time the ad is viewed, and 2) the advertised brand is visible 100 percent of the time. Online advertising often competes with other ads on the page, and both TV and online utilize multiple frames or animation which means that many of the ads do not have the advertiser’s logo present all of the time.

Often very effective

While one online impression may not yield the same brand metric changes seen for other media, it has been well established that multiple exposures to an online campaign are often very effective at driving changes. Since a typical online impression is much less expensive than an offline print advertisement or TV spot, it may be more cost-effective to reach people online multiple times versus offline only once. Figure 3 illustrates the changes in awareness metrics after people are exposed to an online campaign a varying number of times.

For all of the awareness measures, there is a big jump in online campaign efficacy after four or more exposures. The data also indicates that the “sweet spot” in building awareness occurs between four and seven exposures, which is consistent with previously-released research in this area. The data shows that there may be diminishing returns after eight or more exposures. The media planning implication is that an online campaign must be shown to people four to seven times to maximize the likelihood of driving awareness measure increases. There may be exceptions to this recommendation when utilizing newer online formats such as video, which will be explored later in this analysis.

May take more impressions

The persuasion measures (Figure 4), which are often much more difficult to change, tell a consistent story with the data shown above, however, it may take more impressions to maximize a measure like purchase intent. This is logical as it is often more difficult to get someone to want to commit their financial resources to a brand versus simply educating them about the name or benefits of a brand.

The data indicates, however, that while intent is maximized after eight or more exposures, the results will still be strong after only four exposures. When combined with the awareness data, the takeaway is that a media plan should strive to expose site visitors to an online campaign at least four times. If the budget is available to increase that frequency, maximized increases may occur after eight exposures.

On the other hand, the television data by frequency (Figure 5) shows that many more exposures are needed to maximize the increases in awareness measures. After people have had an opportunity to see a campaign 16 or more times, the changes in awareness peaked. One thing to note is that the data collected on TV exposure has a greater standard error than the online exposure data so it is possible that the research is understating the frequency results for this channel. However, it remains clear that people must see that ad multiple times before the increases occur.

The persuasion measures (Figure 6) are again consistent with the awareness data for television frequency. A campaign is most likely to drive a maximized increase in persuasion after at least 16 exposures.

Print-channel frequency data is more similar to online than to television. After four exposures to a print campaign, the awareness measures appear to be maximized. The methodology for determination of exposure for print measurement, as with online, tends to be very accurate relative to television which may play a role in why those two media channels require fewer exposures than television to drive increases.

Print results by frequency show that the awareness measures of a campaign take a large jump between two to three and four to five exposures (Figure 7), which is consistent with the online data. Persuasion metrics (Figure 8), however, show a steady increase after each exposure with the largest increases occurring after at least six exposures.

Across all channels, the frequency data shows that to drive the greatest changes in both awareness and persuasion metrics, people must be exposed to a campaign multiple times. This transcends all three of the media channels analyzed. The goals of a given campaign should dictate the ideal frequency level, as well as the desired media channels for communication.

Channels have converged

In a way, the online and television channels have converged with the introduction of full-screen online video advertisements. Online video is primarily shown in conjunction with television programming made available after the original air date. Video campaigns typically outperform traditional online display due to their larger size, engaging content and the lack of clutter during viewing.

The normative data for this media tactic show that it does outperform the online channel, especially in advertising awareness (a measure of whether people recalled seeing the campaign) and intent.

The biggest difference between the online display norms and the online video norms is that there appear to be diminishing returns for video after seven exposures (Figure 10). It is possible that, because the ads are much more engaging, it is a case of too much of a good thing. This is data across nearly 100 campaigns so there will be exceptions, but it seems that, in general, managing to six or seven video exposures is ideal.

Very complex

Analyzing the cost-effectiveness of the various media channels can be very complex because CPMs (cost to purchase 1,000 impressions) vary tremendously within each media. In an effort to explore this arena, this analysis used average costs published by Morgan Stanley to compare media performance relative to the investment required.

InsightNorms television data does not break out types of television buys, so an average CPM was calculated. The following inputs were used in the cost-effectiveness analysis. There is a large gap between the costs associated with television and print relative to online costs. The analysis that follows explores if the additional costs are justified.

For purposes of simplicity, each campaign will assume the same number of impressions so that the calculations can control for reach.

Calculation for ROI:

Delta * 1,000,000/Avg. Freq = Number of people influenced (1,000,000 is arbitrary and allows removal of campaign size as a contributing variable)

Total Cost (CPM * 1,000,000/1000) / Number of people influenced = ROI

Example: delta of 5.0% with CPM of $10

ROI = $0.20, which means it cost $0.20 to influence one person for the given metric. The lower the cost, the more successful the marketing effort.

In terms of building awareness, the online channel is very cost-effective, including at higher exposure levels (see Figure 11). Since the cost of Internet advertising is low relative to offline marketing, it is a cost-effective media channel provided the creative is well designed. One benefit to television at this stage is that TV does have very high reach so a campaign using this channel can put an advertiser’s message in front of a large number of people.

The results are fairly consistent for purchase intent (Figure 12). The online channel remains a very cost-effective way to persuade consumers to buy products and services. The interesting change is in the print channel data, which suggest it is much more cost-effective to drive intent changes via the print channel than to build awareness.

Strategically effective way

This analysis underscores the value of multichannel marketing efforts as a strategically effective way for marketers to communicate with their target consumers. Reaching people via various media is proven to be more effective than single-channel tactics; however, it is clear that this is a function of frequency in addition to the exposures to various forms of advertising.

Across all channels, campaigns benefit considerably from repeatedly exposing consumers. InsightExpress consistently recommends that media planners aim to reach people multiple times, as repeated evidence proves that this approach can have a significant impact on how well a campaign communicates with and persuades people.

This analysis also illustrates that while offline channels may perform better on an impression-by-impression basis relative to online marketing, when cost information is incorporated, the Internet can be an extremely cost-effective way to communicate with a target audience.