For decades, brand marketers have relied on survey-based methodologies to measure the effectiveness of their campaigns. Survey-based measurement can certainly be executed correctly within a margin of error, but the results are heavily dependent upon sampling methods that pose a number of challenges:
Sample size: The survey’s sample size may be very small, or it may skew toward some segments that aren’t reflective of your target audience. Making planning decisions based on this type of data can lead your marketing astray from achieving your objective with the right audiences.
Statistical significance: Before you can draw insights, it’s critical to make sure that the data has reached statistical significance. But achieving statistical significance at a granular level is extremely difficult without spending an enormous amount of your budget to recruit large sample sizes for each dimension of your marketing. You may be left with a partial story of marketing effectiveness if data at the publisher, placement, or creative levels doesn’t have enough statistical significance to be meaningful.
Timing: Since it can take weeks or months to develop the ideal sample size and statistical significance, you may find yourself waiting for meaningful data while the campaign continues to exhaust your budget without any opportunity to make optimizations mid-flight. By the time statistically significant data becomes available, it could be too late to reallocate budget, or the campaign could be over.
Even if a brand marketer receives meaningful survey results for one campaign, there’s no guarantee that the results will be available with the same sample size, statistical significance, and timing for the next campaign, which makes it difficult to maintain a closed-loop brand marketing strategy, which I wrote about in an earlier post.
Today’s CMOs and brand marketers are looking for data-driven brand measurement that they can hang their hats on. They want an approach that takes in all of the data points from their marketing plan, as opposed to a sampling of survey responders. And they want to use that data much faster to inform optimizations throughout a campaign.
The answer lies in building a single brand engagement currency to measure impact across multiple channels, and optimize at a tactical level. Using a currency that is based on actual user interactions with your brand provides the level of data needed to analyze impact at a placement, keyword or creative level, and to make frequent optimizations that drive lift in brand engagement. It also enables highly accurate scenario planning so you can launch new campaigns that are designed to achieve specific quantitative goals.
Although digital brand marketing is growing, TV still reigns king for many brand marketers. So it’s important to use the same currency to measure the branding impact of both TV and digital (rather than using GRPs to measure TV advertising, online metrics to measure digital advertising, and surveys to measure branding impact). Only then can you most effectively optimize your total budget across TV, search, display, mobile, etc.
CMOs and brand marketers should be excited about the future ahead of
them. The data they need is already available and waiting to be put to work, but their
marketing technology needs to add value to make it work. When the tools and methodologies align
with the data, the full potential of brand marketing is finally unlocked.
Anto, while I can't disagree with what you have posted, one sentence stood out to me.
"He wants to prove that his brand marketing dollars are having an impact, but believes that traditional tools and methodologies lack the ability to do so with the accountability his company demands.". In that sentence one phrase stood out "with the accounability his company demands".
In my decades of working in this area, I am yet to meet the marketer or marketing company that is prepared to pay the costs that such levels of granular accountability demands.
Statistical methods are improving every year. We have more data than ever and more granular data becomes available constantly. We have more computing power at our disposal than anytime in history. Our microeconomic and macroeconomic modelling techniques (and more importantly our verification and validation methods) are extremely advanced. We're missing the financial backing.
John, I have to agree with you. Marketing directors and higher ups at almost every agency client I have worked with ask really good questions but expect someone else to fund not only the research but also the staffing to get the answers. This is amply evident in many of the ROI studies I have had the misfortune to encounter. The models have the capacity to evaluate many factors but are fed mainly ad dollars and sales, because these are easily obtained. The research departments haven't got the resources to obtain the missing---mopre relevant---information. Thus little is learned ----buy "they" keep asking the same "demanding" questions and keep getting the same unresponsive answers, namely that advertising works...sort of.