Full-Funnel Plan Built on One Arithmetic Sheet

The deliverable is one spreadsheet and a document of about fourteen pages, and the spreadsheet is the argument. Channel spend, expected volume, the capacity of the people who have to handle that volume, and the month cash actually leaves and arrives, in a single model that has to balance. Funnel plans fail their first quarter because those four things were written by four people and never made to agree with each other. Capacity is the first constraint, not demand: An industrial manufacturer whose applications engineers can produce nine technical quotations a week does not have a top-of-funnel problem at forty enquiries a month. It has a quotation problem being treated with media budget. An admissions team of four cannot absorb the enrolment peak the plan is buying for. I size the channel plan underneath the constraint, which is why plans from this practice frequently cut spend somewhere and then say where that money should go instead. A calendar, not a year: No annual plan. Four quarters, with dates taken from a real calendar and a decision attached to each date. In education that means exam results and the tercih window, a fixed number of weeks in which enrolment intent exists and outside which spending is a donation. For manufacturers it is the fair schedule and the buyer's fiscal year, which is frequently not yours. For foundations it is Ramazan and the last ten days of December. Where a client budgets in lira against media priced in dollars, the model carries a re-plan trigger: a stated percentage move that forces a scheduled review, instead of a renegotiation every month. Measurement, stated plainly: Below roughly 150,000 euro a month in media I will not propose mixed media modelling. The data cannot carry it and the output will be confident and wrong, which is worse than no output. What goes in instead: geo holdouts where the market allows one, a single conversion definition per funnel stage, and platform-reported alongside CRM-reported numbers with the gap named rather than reconciled into one figure that somebody will quote in a board meeting. Remarketing audiences enter the plan only with Consent Mode v2 and a KVKK consent record behind them. An audience built without that is a liability being carried as an asset. The Ad Grant, since foundations always ask: Ten thousand dollars a month of restricted, low-intent inventory, governed by an account-level click-through floor and rules that punish exactly the keywords you want to buy. It belongs in the plan as a discovery channel with its own conversion definition. It is not a substitute for an acquisition budget, and a fundraising model that treats it as one has a hole in it that shows up in the second quarter. What this does not include: No creative concepts, no brand positioning, no naming, no messaging framework. That is a different discipline and I am not it. No media buying, no account management, no commission on spend. I do not run what I plan, which is exactly what lets me write a plan that spends less. No website or landing page work. And no plan you can present unchanged in month seven; if it survives that long unamended, it was never specific enough to be wrong. Who this is not for: Companies before product-market fit, where the model would be fiction with decimal places. Businesses whose real constraint is the product or the price, where a funnel plan postpones the conversation they need to have. And anyone who needs a board deck rather than a plan a team will work from. The fourteen pages are written to be argued with internally, not presented. How it ends: On a project, a ninety-minute session eight weeks after delivery to check which numbers held and which assumption broke first. On a retainer, that check is the standing quarterly meeting, and I re-cut the model in front of you rather than emailing a new version with the changes hidden in it.