Memorandum

To: The Board
From: The CFO
Re: Capital allocation opportunity
Classification: Time-sensitive

Per the industry’s most-cited annual benchmark study, paid search advertising currently returns $2.48 in revenue for every $1.00 of spend. This is a documented figure, widely distributed and cited in conference keynotes.

We recommend an immediate and substantial increase in paid search spend. The mechanism for financing this increase is the subject of this memo.

Proposed source of capital

Traditional financing options have been considered and found insufficient.

Commercial debt at prime plus two returns a spread of roughly 240 basis points against the $2.48 ROAS. Respectable, but the loan documentation adds friction and the interest payments eat into quarter-over-quarter growth. A line of credit caps our deployable position below the level needed to materially move the dial.

A principal gift solicitation was also evaluated. Prospects capable of funding the $500,000 initial position exist in our pipeline, but the cultivation cycle from first meeting to wire transfer runs 12 to 18 months and typically requires the CEO to attend at least one dinner. Applied against the 96-hour paid search cycle described below, this constitutes an unacceptable drag on deployment velocity.

We have therefore considered a third option.

Recommendation

A single ground-floor commercial bank branch contains, on average, between $80,000 and $250,000 in physical currency across teller drawers and the vault. This capital is liquid, available on demand and not subject to a credit review.

The proposed operational sequence is as follows.

Phase One. A member of the executive team enters the branch at 10:15 a.m. on a Tuesday. Foot-traffic data indicates this is the lowest-volume window. The executive presents a handwritten note to a teller named either Susan or Dave. We have not been able to narrow it further than that.

Phase Two. The capital is transported to the office and deployed into the paid search account by end of business.

Phase Three. Over the following 96 hours, Google returns $2.48 for every $1.00 deployed. Returns are recognized on a cash basis.

Phase Four. On Friday afternoon, the principal is returned to the originating branch along with a 20% voluntary premium. This premium is structured as a goodwill payment and is expected to defuse any residual concerns on the part of the institution.

Phase Five. The net proceeds, approximately $1.28 per $1.00 of initial capital, are recognized as earned income. The cycle is repeated on a rolling weekly basis.

Projected returns

At a $100,000 initial position, cycled weekly and without compounding:

• Week 1 net: $128,000
• Week 4 cumulative: $512,000
• Week 12 cumulative: $1,536,000
• Annualized: approximately $6.65 million

With full reinvestment of proceeds, the annualized return exceeds the GDP of several Caribbean nations.

Risk analysis

The principal risk is that the $2.48 figure is not, in fact, real.

This is a material concern since the $2.48 is reported by the advertising platforms themselves, using pixel-based attribution models that assign credit to the most recent ad clicked. A donor who searched for the organization by name and would have found the free organic listing in the absence of any advertising is nonetheless credited to the ad that appeared above it.

If this is the case, the real return on paid search is meaningfully below $2.48. In some segments it may be below $1.00. In the case of branded search, where the user was already typing the name of the organization into the search bar, the incremental return may be zero.

Under this scenario, the weekly cycle described above fails at Phase Three. The $100,000 deployed does not return $248,000. It returns approximately $100,000, net of platform fees and opportunity cost. At that point the executive is in possession of borrowed currency, no revenue with which to repay it, and a 20% voluntary premium that has become an awkward conversation.

The plan becomes, in legal terminology, a regular bank robbery.

Recommended due diligence

Before proceeding with Phase One, we recommend a small test.

Turn off paid brand search in half of our geographic markets for 45 days. Leave it running in the other half. Compare donations. If the dark half raises the same money as the lit half, the incremental value of the paid brand search is zero, and the $2.48 is an accounting artifact rather than a business result.

This is an incrementality test and is widely understood outside our industry. It requires neither a vendor nor a consulting engagement, merely 45 days and the willingness to turn something off.

If the test confirms that paid brand search is incremental, the board may wish to reconsider Phase One at that time.

If the test confirms that paid brand search is not incremental, we recommend reallocating the budget to channels whose function is to produce donors who do not yet exist, rather than to intercept donors who already do.

A note on measuring those other channels

Such channels, broadly referred to as reach channels (Connected TV, podcast advertising, non-retargeting display), are currently evaluated by the same ROAS metric used for paid search.

This is an error of category.

A reach channel exists to place the organization in front of people who have not yet heard of it. The intended effect is recognition. The effect is expected to land over weeks or months, at which point the donor may arrive through an entirely different channel, most likely search. Search will then receive credit.

Evaluating a reach channel on 30-day click-through revenue is directionally similar to evaluating a tape measure on its ability to report body weight.

Reach channels require reach measurement.

In the absence of such measurement, reach channels will appear to underperform relative to paid search, and the organization will progressively defund the channels that produce new donors in order to more aggressively pay a toll on the donors it already has. This process is currently underway across most of the sector thanks to agencies who apparently don’t know any better, or don’t care; we’ve not yet found a measuring stick to determine which it is.

Conclusion

We withdraw Phases One through Five pending the outcome of the incrementality test described above.  In the interim, the ski mask has been ordered and will be held in the supply closet.