Don't spend until you have to

Two rules underpin every action Yuzu takes, and they are worth stating in their original form, because the original is sharper than any rephrasing we could produce. Don't spend until you have to. When you spend, spend at the point of maximum impact.

These are not poetic statements, they are operational ones. Every email a seller sends, every meeting a seller books, every minute of attention a seller allocates to a particular deal is a spend, and the unit of spend that matters is not the message itself but the seller's time. The default in most sales operations is to spend continuously and uniformly, every Monday a check-in cycle, every deal a rough equivalent of touches per week, every relationship maintained by some standard cadence regardless of what is actually happening in it. The result is a low-yield use of the most expensive input in the system, which is the seller's attention.

The discipline of not spending until you have to is harder than it sounds, because it requires letting deals sit when no behavioral signal is calling for action, and most sellers feel uncomfortable doing that. It means not sending the just-checking-in email when the buyer has not yet given you a reason to send anything. It means trusting the rhythm of how deals close at your company and intervening only when the rhythm breaks. Most sales tools push the opposite default, the every-seven-days follow-up cadence, the post-meeting nudge at the fourteen-day mark, the value-add touchpoint at the thirty-day mark, all of which are time-based rules that fire on every deal regardless of state. Those defaults are why so much of inbound and outbound activity feels performative to the buyer, because performative is what it is.

The discipline of spending at the point of maximum impact is the second half, and it depends on the first half being respected. When something does call for action, the action should be the right one for the moment, informed by the full deal context, drafted with the specific stakeholder in mind, sized to the actual situation. A short, specific, behaviorally-triggered message at the right moment outperforms ten generic check-ins by an order of magnitude, and the reason it does is that it is the one message in the buyer's inbox that day that is grounded in something real rather than in a tool's idea of what week it is.

How this works in practice, inside Yuzu, is straightforward. Behavioral state is watched continuously and silently. No notifications fire on time-based rules, no sequences are sent on calendar dates, no weekly check-ins are pre-scheduled. The default state of the system is silence, and that silence is intentional, because a sales tool that fires constantly trains both seller and buyer to ignore it.

Trip wires fire only on behavioral thresholds. When a champion's reply latency exceeds their personal baseline, when a deal's recency drops past a cohort hurdle, when a competitor name shows up in a transcript, when a stakeholder's role changes at the buyer, these are the conditions that produce action. Time alone never does. Behavior is the trigger, and behavior is what makes the trigger meaningful when it fires.

Each fired wire produces a single specific action, not a sequence or a five-step play. One move, designed for this moment in this deal with this person. A drafted message in the right channel for the right person, a briefing for the seller before the next call, a scheduling proposal at the calendar moment that historically converts. The leverage of the system is in having the right move ready at the right time, in saving the seller the cognitive load of figuring out what to do, not in flooding the seller with options.

The seller has the final say. Yuzu does not autonomously send messages on the seller's behalf without explicit configuration, and even with configuration the boundaries are tight. The drafted move is presented, the seller approves or edits or rejects, and the seller is the one whose name is on whatever gets shipped. The leverage is in the preparation, not in the removal of the human from the loop, because the human is the thing the buyer is actually buying.

This is the inverse of how most sales-AI tools work, and the inversion is intentional. The mainstream pattern is generate as much outreach as possible, schedule it across as many cadences as possible, and let volume do the work of converting the small percentage that responds. That strategy worked when the cost of generating outreach was high enough that even mass-produced outreach had some effort behind it. It does not work now that the cost of generation has approached zero, because the floor of quality has dropped to the floor of a language model and buyers have learned to recognize it instantly.

When generation is cheap, restraint is the differentiator. The team that sends ten thoughtful, behaviorally-timed messages a week beats the team that sends a thousand generic ones, every time, in every category we have looked at. We build for the first team.

This principle has a corollary that surprises people the first time they hear it. Don't try to save deals that are not worth saving. If a deal is in the Don't Spend quadrant, the right move is no move at all. Some of those deals will close themselves, some will not, and spending hours trying to revive a dead one comes at the cost of the live ones in your other quadrants. The discipline is letting the cold ones go cold, on purpose, so you can pour effort into the warm ones.

The two rules together produce a sales operation that looks strange from the outside. Quieter than expected, with less activity and more impact, with a pipeline where the seller has more time to think, more room to take real meetings seriously, and fewer cycles burned on motion that does not move deals. That is the operation we are building Yuzu to enable.