Last December, I started a small venture with my friend, Danilo Malanczyj. We’re (hopefully/probably/potentially) a few months away from spending a good chunk of change on advertising.
Question is, how much? I/We have no idea – yet, and there are some big picture strategy questions which are still a function of our patenting success…but…
I do know enough, to proclaim the preliminary guidance online, is bunk.
“5 to 7 percent of sales” is a ”rule of thumb” – I say B.S.
“11% of your gross sales, then subtract rent” – common, really, it’s a harder question than that.
“Set an ROI goal for your Marketing” – was getting closer, to something I want to do, but still in my view inadequate analytics.
I’ve spent, about 4 hours so far thinking about this problem. I intend to spend quite a bit more time. At this point, I’m fairly positive, answering the question “How much should a widget manufacturer spend on advertising?” involves quantifying the impact via a sensitivity analysis:
- Estimating/Modeling the per unit cost sensitivity to each marginal unit sold.
- Estimating the additional number of units sold, per marketing dollar spent.
- Adjusting pricing strategy, as a function of marketing campaign size.
- Quantifying the volatility of revenue, as a function of marketing dollars spent
- Planning for out-of-stock and negative cashflow due to higher than expected growth risk.
- Quantifying economies and/or diseconomies of scale, as a function of additional units sold.
- The weighted cost of capital and/or the ROI demanded by partners and investors.
I conjecture, that some level, a marginal dollar of advertising becomes a) more risky than desired, b) less effective than estimated or worse c) a complete waste of money.
There is no way, the function of ROI on marketing dollars spent, is linear. Therefore, given the aforementioned, a finite optimum does exist.
I’m going to keep thinking about this…



