Every marketing budget meeting contains the same argument. One person wants more for the channel that feels like it works, another for the one that is cheap per lead, and a third points out that the last quarter's best month came from something no one funded. Without a rule, the biggest voice or the most recent anecdote wins.
We suggest a rule that a finance team can audit and a marketing team can live with. Fund each channel by what a qualified opportunity costs, measured over that channel's own lag, and move money on a fixed schedule within a floor and a ceiling.
The rule needs one input, and we described how to build it in the one attribution table the CFO will trust: for each channel, the spend and the qualified opportunities it produced, read from the CRM and the books and not from the channel's own report.
If you do not have the table, build it before allocating anything. An allocation made without it is a guess, even if it is a considered one.
The usual alternatives each measure something nearby and miss.
Cost per lead rewards the channel that produces the most contact forms, whatever they are worth. A channel can have a low cost per lead and a cost per opportunity that is twice another channel's, because most of its leads never answer a call.
Last-touch credit gives the money to whichever channel happened to be last before the form. Search tends to win, because people search for you after hearing about you somewhere else. Funding search at the expense of whatever created the demand eventually starves search of demand.
Cost per qualified opportunity asks a harder question, and the answer sits close to money: what did it cost to put a real, sales-accepted prospect in the pipeline? It is not revenue, which arrives months later and is shared among many touches, but it is far enough along the funnel to mean something and early enough to act on.
Judging every channel on the same window. A paid search campaign shows opportunities in weeks. A content series, an event or a partner program may take a quarter or two to show anything. Measure each over its own lag, set when the channel is funded, and do not cut a slow channel for being slow before its window has closed.
Ignoring that cost per opportunity rises with spend. The first dollars in a channel reach the people most likely to respond, and later dollars reach people less likely to. A channel with a low average cost can have a high marginal cost, and the allocation is a marginal decision. Treat the best channel's current figure as a ceiling on what the next dollar will do, not as a promise.
Leaving a bucket that nobody measures. Brand, events and sponsorships are often exempt because they are hard to measure. An unmeasured bucket grows. Give it a fixed share, a named owner and a review date, and judge it by the same table at the end of its window, even if the answer is that the measurement is poor.
Three guardrails keep the rule from swinging too hard.
A floor for each channel that has earned a place, so that one bad quarter does not end a channel with a long lag. A common choice is half of the previous quarter's spend.
A ceiling, so that no channel takes more than a set share of the total, because the marginal cost is rising and the data on a big move is thin. A third is a sensible start.
A cadence. Money moves once a quarter, in steps of ten to twenty percent of a channel's budget. Weekly shuffling reacts to noise, and annual allocation leaves a bad bet running for a year.
Within those limits, the rule is simple: move from the channel with the highest cost per qualified opportunity to the one with the lowest, in a step, and watch the next window.
Imagine a company spending thirty thousand dollars a month across five channels. Over the last two quarters the table shows these figures per month: paid search, ten thousand dollars for five qualified opportunities, so two thousand each; a partner program, six thousand for three, so two thousand; outbound sequences, six thousand for two, so three thousand; an events line, five thousand for one, so five thousand; and content, three thousand for one, so three thousand, with a longer lag than the rest.
The debate would normally be about events, which feel valuable, and search, which is the biggest. The rule says something simpler. Events cost the most per opportunity. They are not cut to zero, because the floor holds half of their budget. A step of fifteen percent moves seven hundred and fifty dollars a month to the partner program, which ties search for the lowest cost and, unlike search, is well below the ceiling. Content is left alone, because its window has not closed.
The next quarter's table then decides whether the partner program's cost held at the larger budget. If it rose to three thousand, the move has found the ceiling and the money goes back or sideways. Either way the finance team can see the rule, the input and the result, and the conversation is about the table.
Some influence never shows in the CRM. A conference talk makes someone remember you, and they reach out six months later through search. Do not pretend this away. Add a question to the first conversation, "how did you first hear about us", and record the answer as a field. It is imperfect and self-reported, and it fills a gap the system cannot see. Use it to sanity check the table, not to replace it.
Put your last two quarters on one page: spend and qualified opportunities by channel, with the lag you expect for each. Compute the cost per opportunity and mark the highest and the lowest. If you cannot produce the page, you have found the first thing to fix. If you can, you have the first allocation move, and a reason to give the CFO. Agency reports that hide the number are a common cause of the gap, as we describe in the report that hides the number the CFO wants.
We build the table and run the rule as part of marketing run as a governed system, with a human at every spend gate. We start with a no-risk discovery: we assemble your last two quarters from the CRM and the books, show the cost per opportunity by channel and propose the first move, and you pay for the discovery only if you go on. A build comes with a guaranteed estimate, so if we estimate low, we absorb the difference.
A budget is easier to defend when each move can be traced to a table.
Watch the 6-minute guided demo, then book a walkthrough: we read your CRM, ad accounts and content and show what a governed system would do differently. You pay only if you proceed. Or see how we approach it.
More on the same problem:
"Something isn't working." "We're outgrowing our tools." "We're afraid to make the wrong move." No-Risk Discovery is a short, practical conversation that gets you clarity before you commit to anything big. We'll tell you if we're a fit. If we're not, we'll tell you that too.
Say who you are and ask. A person replies within one business day.
