Former Philips Consumer Electronics global marketer shares tactics for judging how well online drives retail sales.

How can we measure the effect of online promotion on sales? Linking online marketing to offline sales has been difficult because most of us sell via retailers, and we lose track of the consumer on their way to the store. In general, retailers are also not highly motivated to track the reasons why consumers came in through their door -- or at least not to share their findings.

But there's hope. The Internet is not just a great promotion vehicle, it's also the tracking source that enables us to close the loop and see what happened after the visitor left the Web site and went shopping.

It's far more cost-efficient to follow up the purchase online with an array of post-purchase tools -- and therein lies the precious ROI information we all seek.

So here are a few tactics to consider when trying to make the connection between online information-gathering and offline purchase:

• Pre-purchase intent surveys
• Post-purchase surveys
• Online coupon redemption
• Online rebating
• Online warranty registration linked to prior online registration
• In-store or exit surveys in conjunction with retailers
• Panel-based data like Nielsen Home Scan.

Surveys

Let's start with a relatively simple online survey example that illustrates the technique. This one was cheap and quick to produce, but it yielded a ton of useful information.

Case History: Online Survey to Determine TV Buying Influencers

Cars, consumer electronics and other technical/consideration purchases are heavily researched online, but not often purchased there. AOL conducted a survey in the actual places where the online promotion had happened, such as online channels dealing with entertainment.

A total of 1,525 online interviews were conducted: 1,004 people who had purchased TVs within the last six months, and 521 people who intended to buy a TV within the next six months. A rigorous screening and qualification process ensured a high level of veracity.

Results of the surveys revealed that in-store displays (58 percent) and past experience/previous ownership (48 percent) are the most important sources of TV purchase decision-making, while retailer flyers (13 percent) and online (12 percent) are ranked by TV purchasers as the most important media sources for new TV information.

This kind of survey provides invaluable direction as to the effect of online promotion, but also points to the necessity of linking online promotion with in-store promotion. The disadvantage of surveys is that they are simply samples of the universe, so it's hard to provide credible real dollar information generated from the program.

1st/2nd/3rd (Combined) Most Important Information Source for New TV Purchase

Here are some other ways of getting at the ROI data (or, maybe more accurately ROMS -- Return on Marketing Spend):

Online Coupon Redemption

• Promote a coupon, get printout taken to store
• Redeem at register and ask redemption house to break out offline vs. online coupon redemption
• Various alternatives, such as require registration to get coupon/bump up against warranty information
• Gives raw ROMS/ROI

Online Rebate/Gift with Purchase

• Cookie visitors who visit your sites
• Offer manufacturer rebate or gift with purchase
• Require online registration with identifiers as proof of purchase.

Example: Enter barcode, code printed inside Pepsi top etc.

• Track rebate progress online or
• Deliver gift online
• Match up cookies and pre-and post-purchase registrants to determine how many visits lead to purchase.

Voila! You start to have a picture of how many dollars were actually generated by the promotion.

For example, if you only promote the program online, and only redeem the offer online, you start to have a fairly pure ROMS picture, even when the purchases are all made offline.

Of course, you will not capture all purchases because of "breakage", where people buy without ever redeeming the offer, but you can infer a total once you understand how much breakage you are experiencing.

Warranty Registration

• Simply ask which was the most influential medium when people register
• Track number of visits and time spent pre and then post via cookie matching
• Get partners like Tacoda or Revenue Science or DoubleClick or Atlas involved to help figure out where people spend their time.

In-Store Surveys

This method requires close cooperation with retailers -- especially the offline ones. At a recent iMedia Summit, it became very apparent that retailers experience the same silo separation between offline and online as exists in large companies. 

Doing exit or cash register line surveys can be very helpful.

• Example: BestBuy surveys show 19 percent of visitors would not be there but for online promotion
• Need very specific questions to get meaningful ROI data.

Cautionary note: Offline buyers/category managers need to buy into the methodology, and be willing to keep an open mind as to the effect of online.

Panel-Based Data

Nielsen Homescan and others may give a good behavioral picture for mass product consumption based on online behavior.

Again, this method is directional rather than absolute in terms of real ROMS.

What Other Things Can Be Done?

The auto industry can compare test drive takers to online registrants, and capture email addresses at dealerships. CPG companies can measure scanner data lift based on online roadblocks vs. GRPs. It's all about practical data.

So, in sum:

• Get your databases working together
• Ask your retailers to cooperate
• Link POS and Online: Use "As Seen on the Internet" as a benefit
• Continually give reasons for consumers to come back online
• Use pure online promotion plays to determine ROI effect.

There is no perfect way yet to tie online promotion directly to offline sales. You need to do the work of aggregating all the different sources into a simple ROMS/ROI calculation. Don't be surprised when humble tools like POS materials show up as more important in converting dollars than online. They also usually beat the heck out of TV.

Derek Hewitt is President of iMediaLearning, Inc. Prior to joining iMedia, Hewitt ran global interactive marketing for Philips Consumer Electronics. In that capacity he was responsible for CRM, global database development and original content online. He was also Philips interface with various movie and music entities and developed other content relationships. He is a 20-year brand marketer with Madison Avenue agencies who has worked on campaigns for such clients as UPS, Four Seasons, Sony, Pan Am, Waterford Crystal and Club Med. He also owned an ad agency in New York for six years, whereafter (tiring of incessant compensation discussions) he formed his own interactive media company, took it public and sold it.