| ||||||||||||
PRINT
COMMENT
SUBSCRIBE
RSSMeasuring the effectiveness of marketing and advertising is tricky business. There are a slew of complex issues including cross media, viewthrough, sequence of exposure and so on. And while a handful of companies can afford the luxury of such sophisticated analysis on their campaigns, most are still trying to get the basics of measurement right. And it is the very basics of measurement on which sound media and marketing plans are built.
The first step towards meaningful measurement starts with identifying the right metric or the measure of success. This sounds really simplistic - of course we need a success metric. But it is not uncommon for campaigns to be launched without a clear definition of success metric or for us to pick the wrong metric.
The first (and obvious one) question to ask is: What is the campaign objective? Your metric should be derived from what you want your campaign to achieve - a campaign geared towards increasing an e-mail database with new sign-ups cannot be deemed successful if it achieved a low cost per click. The objective was to drive e-mail sign-ups, so the measure of success could be cost per e-mail sign-up. My rule of thumb is that if you cannot explain the campaign objective in a sentence or less, you are complicating matters. However, sometimes the campaign objective doesn't lend itself to direct, short-term measurement, as in the case of a call center geared towards increasing customer satisfaction. In such situations, the chosen metric should be correlated to success even though it isn't a direct measure of success. In the above example, first call resolution might be strongly correlated to customer satisfaction and hence could be a good metric to pick.
The next question to ask is: Is it stable? You don't want to pick something that is so volatile that it doesn't lend itself well to trending, analysis or forecasting. With a zillion metrics and thousands of data sources, it is important to pick a metric with a reliable data source. Of equal importance (unless you want to drive all your analysts away) is ease of data pulls and manipulations. If you expect to use this metric extensively, be sure you take into account the user interface for getting to the desired data.
Can it be benchmarked? The process of benchmarking is as important (if not more) as the process of choosing the metric. Once we say that cost per e-mail sign-up is the desired metric, we need to put a number against it - a number that will differentiate success from failure. Is the cost per sign-up of $100 a success, or not? This determination can be made in a number of ways. The metric can be compared over time, regions, across similar brands or sometimes even industries. There is third party data available across industries and also across ad types that can be utilized for this purpose. However, my experience with such information hasn't been the best since there are so many factors that such aggregated data hides, i.e., in the case of display advertising, how targeted (or not) was the media? Did the creative promote a special offer, thus increasing click-throughs? And so on. It is, therefore, best to consider benchmarks you are familiar with or those that you understand the context for.
Lastly, ask the question - Can I influence it? If the metric is the measure of your campaign's success, you need to control the levers that influence it. Consider a common situation where one group is responsible for generating site traffic while the other is responsible for converting that traffic via e-commerce sales. If I am the former, my measure of success is site traffic or site traffic per dollar spent. The latter group is measured on conversion rate or revenue per site visit. Of course, there is interaction between the two - quality of site traffic impacts conversions. These can be managed within min./max. parameters as long as overall metrics accountability lies with those that control it.
Good measurement lies at the heart of continued success. The above questions, though focused on metrics, will force you to consider multiple facets of your campaign and crystallize your marketing and media strategy.
1 person recommends this article.
RSS FEEDS
While there is a risk of not coming up with enough proof of success with the less complex approach, I'd rather take that risk on a small investment in measurement than risk spending a lot more on a complex measurement solution that might still leave me unsure of how much success a specific marketing activity achieved. At some point, the law of diminishing marginal returns kicks in and we need to be concerned about the ROI of our measurement spending.
Good success metrics will tend to be indicators that tell you the money you spent appears to be having the desired impact on a customer behavior that is known to lead to good business outcomes. It can be as simple as a conversion rate at any step in a selling process. It can also be helpful to monitor any of the key drivers or performance indicators that help to explain why a business is or isn't profitable. Examples might include, average order size, or customer purchase frequency.
The last time I went for my annual medical examination, my doctor measured my waist. Apparently, the benchmark for waist circumference the medical community has established for men is 38". Anything above that is supposed to be a leading indicator for a whole array of medical problems. We don't need a battery of complex and costly tests and procedures that might give us more precision, but would ultimately lead to the same learnings and the same decisions as measuring our waist. Through both approaches, we'd decide to eat healthily and get exercise, and we'd monitor our progress through our waist measurement.
In the end, we measure so we can learn something that enables us to make better decisions that lead to better outcomes. There will be cases where it makes sense to apply a costly and complex measurement process. In the meantime, anyone not willing or able to spend on a complex approach would do well to follow the approach you've proposed.