Thirty years ago, during the roaring 90s at Lands' End, there was a philosophical debate about investing in catalog mailings to housefile customers.
One camp said you should mail to a 10% variable profit level, because after accounting for fixed costs, you were approximating a break-even approach to mailing. This meant that a housefile mailing might be sent to 4,500,000 housefile customers.
Another camp viewed the world differently. They said you mailed to a -10% variable profit level, because after you converted customers to a purchase, those customers had higher long-term value, and consequently you would make money on the transaction on an annual basis.
Neither side had an actual Marketing Budget Experiment to prove or disprove the other side. It was an argument that could never be won, and as we've learned from politics, those are the arguments that our modern world is optimized for.
Today we have Marketing Budget Experiments.
Here, we invest more in catalog housefile marketing and we lose money in the process. The top table shows what happens if we do it for just one year. The bottom table shows what happens if we do it every year for five years.
In the top table, we see that the investment makes sense. We lose money in year one, then we make money each year thereafter, causing us to make money overall.
In the bottom table, we see that if we employ this strategy as a "best practice" (doing it every year), we lose money for four consecutive years before finally turning a profit. Overall, the brand loses nearly $1.5 million over five years.
The endless argument continues ...
You get to decide which choice makes the most sense.
But you should make that decision with full knowledge of what your p&l looks like (on a Variable Profit Level) with/without the tactic, right?