Counting Offline Conversions from Online Advertising, Made Easy

Sorry, you have just fallen prey to “Article Link Bait”; there is no easy way at counting online advertising’s impact on offline conversions. Simply put, it’s hard work.

Why is Getting Attribution So Important?

I am always astounded when I talk to an e-business manager/director and they state, “We do not care how many sales we generate in the brick-and-mortar locations because we do not get credit for it.” Wow! With the concept of ROBO (Research Online, Buy Offline) now a well-established and corroborated theory, I would want to get credit for my hard work, and perhaps a few more dollars in my budget.

So solving the attribution problem is a valid task. Add to this the fact that, according to Forrester Research, mobile commerce will represent $31 Billion in 2016. $31 Billion sounds like a lot, but based on total annual retail sales of $ 4.3 Trillion dollars in 2012, mobile represents less than .07 percent of retail sales. Imagine if, along with your digital advertising efforts, your mobile efforts got credit for all of the sales where a mobile device was used during the purchase process.

Last week I was attending SES San Francisco and got to sit in on a couple of sessions where the topic of attribution was discussed. One of the speakers asked what the various methods the audience was using to measure digital and mobile advertising effectiveness were. “Last Click” attribution was the method denoted as the most popular by a show of hands. Last Click attribution has pretty much been discounted by the majority of marketers as a viable means of valuing the integrated value impact of a multi-touch sales process. By now we all know that consumers use a multitude of sources as they engage with various media and content types on their path to purchase. For more ammunition on why Last Click is not the way to go, here is an article by Adam Singer on the topic.

Microchips Embedded in Their Forehead

Online has cookies, tags and invisible pixels to help track online advertising utilization leading to sales conversion, but until we are able to implant microchips in people, we have distinct challenges related to tracking offline conversion of online advertising. Author’s Note: I do not condone large scale implanting of microchips in humans, but if you would like to participate in a small “designed experiment”, drop me a line (humor).

So for our purposes, we need to blend as many data sources as possible to help create a model that allows for projecting sales with a reasonable rate of error (+ or -). Large brands have undertaken the problem by utilizing vector auto-regression analyses incorporating point of purchase scanner data, loyalty program data, combined advertising and promotional variables to develop weighted media mix models to predict various advertising elements’ impact on sales. For the remainder of us that cannot afford a six to seven digit research and analysis project, I offer the following methodology:

Sales Promotion as a Tracking Vehicle

Of course the most basic way to track online advertising to offline conversation is through the use of coupons, daily deals and any other ad type where you ask the consumer to take in and show the ad or ad code to provide the consumer with some benefit, dollars off, free something’s, etc. In my article: Local Promotions – Connecting Online/Mobile Search to Offline Conversion I overview just one way with which to incorporate online couponing as a tracking vehicle.

Retail Tracking – Less is More

For retailers, my firm has successfully measured online advertising impact by first defining a benchmark daily sales amount based on the absence of any digital advertising (seasonally adjusted). Second, once the benchmark is established, we begin flighting advertising elements in discrete combinations for a fixed and tightly defined set of SKUs. Through a process of comparison of advertising “heavy up” and “heavy down” expenditure moves at the local level, we are able to develop a correlation regarding both single advertising source (e.g., Search Engine PPC) as well as multi sources (e.g., PPC with Display). The important part of any project and analysis like this is to keep it simple and isolate measurements using a very small set of SKUs. The end result is a “cause and effect” measurement ‘trap’ that you can impress variables (different ad combinations) to understand ROI and impact.

To measure the impact of mobile on the sales process, there are some exciting new options being developed in the marketplace. As I was writing this article, I was listening to a pitch from LocAid, who enables companies to track customers’ movement via geo-fences, with opt-in of course. Geo-fencing according to, “is a feature in a software program that uses the global positioning system (GPS) or radio frequency identification (RFID) to define geographical boundaries. A geo-fence is a virtual barrier.” By combining online advertising and promotion with geo-fencing and the retail scanner data analysis described above, we can develop ‘traps’ to measure consumer action from the ad stimulus.

If It Were Easy Everybody Would Be Doing It

In summary, measuring online advertising’s impact on offline conversion is hard work, but as I have seen, not impossible. Once the exercise is completed, we end up with reliable media effectiveness projection models that can be used to guide ad expenditure and media planning. The key is to isolate and limit what you study, count and modify. I still believe: If it can be measured it can be optimized.

Related reading

A white chalk drawing of a bird holding a hashtag, set against a green chalkboard backdrop.
An image of a pie chart styled as a medal, with stats about video content underneath.