Measure and Maximize the Impact of Online Activity on Offline Sales
By Doug Miller, VP of Analytics, RocketFuel Inc.
As Seen In: MarketingProfs
Measurability is the battle cry of every digital marketer; and with the drive toward measurability, entire industries have been born. First ad servers, then web-analytics packages, then attribution vendors. And if you have seven figures to spend on technical expertise and a PhD in statistics, digital media-mix modeling is another way to measure your digital-marketing investments.
The result is a proliferation of metrics and measures for every campaign. So the challenge today is how to decide exactly which measurements make sense based on the objective of the campaign. The ultimate challenge, of course, is not how to measure your campaign but how to optimize in real time to maximize the impact on brand-marketing or direct-marketing objectives.
We work with top consumer packaged goods and auto brands. Those brands have been advertising via traditional channels for over a century, and they now have a very specific challenge: Even in this, the "Internet Age," very, very few people buy their products online. So, all the measurement in the world with online metrics won't ever tell those brands whether they are getting ROI with their digital investments, much less exactly how much.
Those clients ask for digital campaigns to be optimized to a click-through rate of 0.1% or to see a "cost per new vehicle built" on their website at, say, $5.50. The problem is that those optimization metrics on clicks and onsite activities don't efficiently drive offline sales.
First Things First
To measure offline sales impact within these two industries, we use the same approach:
1. We measure offline sales impact via third-party trusted research vendors. For CPG, they include Nielsen, comScore, and Datalogix. For autos, they include Korrelate and Datalogix. Doing so gives our clients independent third-party verification of the impact of their digital marketing dollars.
2. We then optimize digital media to drive purchase intent lift among in-market consumers. With CPG, for instance, that might mean increasing purchase intent for a specific frozen dinner among frequent grocery shoppers.
The core challenge is this: How do you distill anonymous online behaviors into measurable results that will generate more product sales at local stores, or more vehicles sold and driven off local car lots?
Five simple steps can make offline sales measurable and allow digital marketers to optimize their inflight campaigns to capture more of consumers' wallets.
1. Measure offline sales
Use Nielsen, comScore, Datalogix, or Korrelate to measure offline sales with a traditional offline sales study, or onboard your CRM database via a data-management platform (DMP) to measure offline sales directly.
By measuring the overall impact of offline sales, you aren't relying on coupon downloads, engagement with the website, building a vehicle online, or other proxies to measure the performance of your digital investments.
2. Focus on in-market shoppers first, not necessarily brand loyalists
Focus your media on consumers who are in-market for your product. For CPG, that means people who shop frequently and at least share equally in the shopping for their household. For autos, that means people who plan to purchase a new car. Sophisticated techniques for identifying these consumers include bespoke data modeling across millions of available data features—including first-party and third-party data segmentation and CRM data.
But you should also use short-form in-banner surveys to confirm those segments and find others that fit your relevancy criteria. For autos, ask, "When do you plan to purchase a new car?" For CPG, ask, "How often have you grocery-shopped in the last two weeks?" Use the survey data to optimize your campaign toward the best segments. And if you have advanced modeling and artificial intelligence, feed the survey responses to the model in high volume to speed and scale optimization.
3. Drive increased consideration or purchase intent, or both
When you seek in-market shoppers, you don't generally want loyalists who are already buying your brand week in and week out. If you want more consumers to purchase your brand of frozen dinners, you want to find people who might be in-market for a competitive product: They are people who, after media exposure, will try your brand. The same goes for autos: Find people who are considering competitive brands; they may well consider your brand after media exposure.
Though you can use third-party data for identifying the competitive sets, you should also use short-form surveys to measure the lift in purchase intent. For CPG, ask, "Which products do you plan to purchase on your next shopping trip? (select all that apply)" with an answer set that includes your product and competitive products. For autos, ask, "Which cars would you consider purchasing?" and provide the same type of answer set.
4. Ignore clicks, coupons, leads, recipe views, and everything else you track in your ad server
Many times we run A/B tests to optimize one part of a media campaign to coupons or new-car leads, and the other portion to drive offline purchase intent. The segments optimized to driving purchase intent have 200-300% the offline sales impact as do coupon- or lead-optimized segments. We have seen that same behavior with luxury vehicles, sub-compacts, SUVs, frozen dinners, brownies, and many other products.
5. Be bold! Don't settle for a few hundred survey responses
To drive offline sales, you need thousands to tens of thousands of survey responses. Confidence for any given segment takes at least 800 survey responses. With 10 segments, that quickly translates to at least 8,000 survey responses. Once you have those survey responses, you can accurately optimize the campaign to drive offline sales. Being bold with your survey-sampling drives offline sales value.
If you measure and optimize to the right real-time metrics in your digital campaigns, you can drive 2-3 times the return in offline sales for products from breakfast cereal to high-performance sedans.
First you need to measure the offline sales impact, then determine what optimization metric aligns best to offline sales. Once you have that metric, be bold and decisive, ignore the deluge of other metrics (clicks, coupons, time on site), and focus on the metrics that matter to drive real results and ROI for your brand.