Did I say that online and offline marketers need to fuse together into a whole marketer? Did I say that online marketers have all the data but the offline direct marketers know the better questions to ask? Well, here are a few great examples to that point from The New Direct Marketing by David Shepard Associates
I wolfed down the book last night after borrowing it from my colleague at Unica, Jay Henderson. Jay is another industry veteran that I get to work with. Web analysts know him well as he has been one of the key people behind the NetGenesis web analytics solution. Later he lead the NetGenesis division after SPSS acquired the firm. Years later now, Jay is at Unica. Unica just gets the best and brightest to work with us.
If you are a direct marketer then you probably know this book well. It is a well rounded text on direct marketing but the key new theme that it introduced (back in its first edition in 1990) is an approach based on statistical modeling. Rather than just relying on more traditional direct marketing techniques such as RFM for segmenting prospects and customers it suggests to use regression and other modeling techniques to better predict which prospects/customers will maximize your returns when targeted. Based on the statistical model, each prospect/customer is scored and ranked for targeting. This is exactly what Unica customers are doing today when they use Affinium Model to mine their customer database and use Affinium Campaign to treat each target cell with the most relevant offer.
So, what can web analyst contribute here in order to make the model much, much stronger? More data than the offline direct marketer could ever dream of! Every online prospect's every click can be fed for example from Unica's web analytics solution Affinium NetInsight into Affinium Model for building even better predictions. Companies have used this approach for example to predict whether an anonymous web site visitor is male or female based on their click behavior. The visitor is then subjected to targeted ads based on the prediction which is something that can be accomplished for example with real time personalization solutions such as Affinium Interact. But it first requires the online and offline marketer to come together and talk.
The book doesn't stop there. What else can web analysts learn?
- Web analysts today buy internet ads or keywords based on success in conversions or revenue. Web analytics solutions by now can capture delayed conversions in cases where the purchase is not made during the initial visit but during a subsequent one (doh!). But already in 1990 The New Direct Marketing suggested not to ignore what it called "Back-end performance management". Namely, to calculate the life time purchases that result from such acquired customers and not just the initial purchase. (doh again!). Otherwise, how could you really tell whether you eventually broke even on the, say, 5 dollars that you paid for a visitor clicking through from a paid keyword?
- We web analysts can jump to conclusions too quickly. For example, a keyword attracting existing customers to repeat purchases may yield higher margins than keywords used by prospects. Our algorithms (in our heads or in our bid management solutions) would suggest to prioritize the keyword that yields higher return. Yet a catalog business knows that you still have to invest into new prospects nonetheless otherwise your customer database will eventually become stale and your business is at risk.
Books, feed your mind!