The acquisition journey runs for months before it produces a result anyone can point to, and for all of that time there is nothing readable to judge the marketing engine by. The number that eventually gets judged can move for reasons that have nothing to do with the campaigns, a market scare or a shift in confidence that suppresses conversion regardless of what marketing is doing, with no way yet to separate that from actual performance. Leadership questions specific investments at the table, and right now nothing sharper than time is available as an answer.
Short-cycle proxies further up the funnel, tuned to the length of the funnel's own lag, give a read on efficiency long before revenue matures. What the market did separates cleanly from what the campaigns did. An answer for the table arrives before months have passed, not after.
A campaign can look completely dead for weeks, nothing moving, before it produces a spike of results a couple of months later, and every instrument in place reports back after the decision window on that campaign has already closed. Months of spend get steered on no readable signal at all, because the eventual read arrives too late to have informed anything. When the results finally land, someone is expected to say whether they were caused by the campaign or would have happened anyway, at a table where specific investments already get questioned. No answer either way is currently available, which is its own exposure.
Proxies further up the funnel move inside the window where a decision still matters, instead of waiting for the late outcome to confirm or deny it. A position on a live campaign becomes possible to hold instead of waiting it out. An answer sits ready before the next round of questioning, not after.
Part of the funnel converts in days through a self-serve motion, and part runs through a long sales cycle where a lead has to move through pipeline stages for months before it counts. Both are supposed to report inside one shared number, yet the internal argument about what that metric should even be is still unresolved, and each side reads it differently depending on which motion it favors. Whatever single figure gets published, the business unit it disadvantages that quarter has grounds to dismiss it as not describing their funnel. The position between units with different cycles carries no mandate to settle the definition for either of them.
Short-cycle and long-cycle activity gets modeled as what it is, two funnels with different lags, rather than forced into one blended number. A pipeline read stays legible to the sales-led side without flattening the self-serve side into noise. And a shared reporting language emerges that both sides can actually recognize as their own.
A short-cycle proxy is an upstream signal, such as a visit, a sign-up, or a qualified lead, that moves early enough to be useful when the final outcome, like a funded account or an activated merchant, will not be known for months. It substitutes for waiting on the lagged number by giving a read on efficiency while decisions can still change.
The proxy is tuned against the funnel's own history, matching its timing and carryover to how long the lag actually runs before the final outcome matures. That calibration is what makes an early signal trustworthy rather than a guess dressed up as data, and it needs enough history in the funnel to be established properly.
The same proxy logic applies, with both motions modeled separately: the self-serve side on its own short cycle, and the sales-led side on pipeline stages that can take months to close. Reporting them as two funnels with different lags, rather than one blended number, is what makes the combined pipeline read legible to both sides of the business.
When the funnel does not yet have enough volume or history to link an early signal to the eventual outcome with any confidence. In that case a proxy would be a guess wearing the shape of evidence, and the honest step is to keep watching until the link can be established.
Budget decisions stop waiting on results that will not mature for months and start responding to a read that moves inside the same window as the decision itself. Marketing gets judged on what it can currently show, not on a number that arrives after the quarter it was meant to inform is already over.