Thursday, October 18, 2007

The opportunity cost of your ad spend


Freakonomics points out that cycling to work may get you killed, but the risk of crashing (12 times more for cyclists vs. auto drivers) is offset by the heart-healthy benefits (people who don't cycle to work have a 39% higher mortality rate). All of which reminds us of your advertising budget.

Many clients allocate advertising at the beginning of the fiscal year, then call it a day. Budgets are carved up among television, radio, print, outdoor, direct mail, Yellow Pages, and sales collateral. The problem with this is that your media plan may be based on brilliant impression forecasts -- GRPs, CPM, reaching the target at the lowest cost in the highest performing media -- but without ongoing shifts in your advertising pattern based on measurement, you'll never optimize results and eliminate risk.

It's like bicycling to work every day, even if a hurricane is blowing rain in your face. You can't pick one approach and then not change it based on real-world events.

Why is changing a plan important? Most marketers understand that different products have different margins, and that different customers have different lifetime value. The same goes with media outlets. One print pub could give you a cost per call of $50, vs. another pub -- running exactly the same ad -- with a $500 cost per call. 20% of your print probably drives 80% of your total print results. You cannot predict this in advance, no matter how good your CPM forecasting, because the real world of media consumption will always surprise you.

How do you change? Measure. If you track the impact of each specific ad -- such as with unique 800 phone numbers -- you can gradually cull out underperforming outlets and redeploy funds into the media vehicles with the highest yield. Once the media "flat tires" are identified, we pull them off, and put new "wheels" on the advertising machine.

Stay tuned for some ad-optimization forecasting tools coming to our new mediassociates.com web site when it relaunches in a few weeks.

0 comments: