When setting efficiency targets for a marketing program it's critically important to step back from the corporate P & L statement and think instead about incremental business.
Let's say the corporate P&L looks like this:
I have had discussions with CEOs with this type of P & L who've argued: "Look, we only make $3 in profit for every $100 order, so we just can't afford to spend money on paid search, we can't gain any traction at that efficiency".
Obviously, this is wrong-headed. Net profit percentage has nothing to do with what you can afford to spend on marketing programs.
Others will look at this P & L and say: "I'm currently spending 15% of my revenue on marketing and still making a profit. I guess I could spend 15% of my paid search sales on media and still make money."
Better, but that's not right either. Nor would it be any more accurate to say: "15% + 3% profit means the most I could possibly afford is an 18% cost-to-sales ratio. Otherwise I'm losing money."
These perspectives all ignore the fact that we're not talking about what the company can spend on the average order, we're talking about what the company can spend on the next order.
When we start thinking about the paid search program driving incremental orders, the numbers change dramatically, and here's why: many of the expenses on the P & L are unaffected by the costs of attracting and fulfilling the next order. Specifically: rent, salaries and other marketing expenses don't move a bit as a result of spending to drive incremental business, and only some operational expenses scale with additional order handling.
With this in mind, this company obviously does need to cover the cost of merchandise, the cost of fulfilling an additional order, and the cost of the marketing to drive that order. So, on a $100 order, $57 goes to cost of goods, perhaps $9 to fulfillment, leaving $34 that could be spent on marketing, without cutting into the bottom line.
Let's see what adding $2 million in marketing expense does to the P & L if it drives incremental sales volume at a 34% advertising/sales ratio:
We show this as a web marketing program, driving an additional $5 million in Web sales and ~$0.9 million in phone sales. Obviously the same logic would apply to any type of direct marketing effort.
Notice that the increases in operational costs -- tied to additional picking/packing, cardboard, returns processing and credit card fees -- scales with the additional sales activity, but does not bite in to the overall bottom line because we factored it into the cost to sales target for the additional advertising.
Indeed, while this more aggressive marketing push reduced the profitability percentage it did not reduce the profit dollars, and as the saying goes: you don't put percentages in the bank.
Stated one way, this marketing initiative came at an expense ratio (or A/S) of 34%. Stated another way, it came at an average Cost to Margin ratio of 0.34/(1-0.57) = 79%. This notion of cost to margin allows for more targeted marketing efforts exploiting margin differentials all the way down to the order level.
This marketing target isn't the be all and end-all target, but it is a place to start the discussion. More aggressive than this means investing (losing) money in the short term on the promise of lifetime value, friend referral business and brand awareness. Less aggressive targets may yield immediate term profits, but some lost opportunity for long-term growth.
It bears repeating that, for this company, the total marketing budget can't come in at 34% of sales revenue. We're talking about incremental marketing expenditures for somewhat mature businesses. Thankfully, for most businesses, a good chunk of revenue comes from repeat customers requiring no additional marketing dollars to secure their order. It is these "lay-down" orders that cover the overhead and profits. Getting to that point where repeat customers carry the load may require either tighter incremental marketing efficiency targets to cover overhead and profit, or greater investment and risk tolerance to get to the promised land.