The puck and the ball: why demand generation is a timing problem
Written by Robin Caller on May 29, 2026
Two pieces of advice have been circling my head since a conversation about our new website. The first you already know: skate to where the puck is going, not to where it has been. It is Wayne Gretzky’s line, taught to him by his father Walter, and by now it is the most overquoted maxim in technology marketing. The second I owe to a Swedish friend, who told me there is a saying from football that points the other way: kick the ball where it is. Deal with what is actually in front of you, now.
Both are right. They are not the same advice, and they are not even the same sport. And here is the quiet joke of our trade: we quote the first and we live by the second.
Everyone says do both. That is not the point.
So let me clear the obvious out of the way. Of course you need both. Anybody who has run demand for more than a quarter knows that anticipation without live capture leaves money on the table, and live capture without anticipation means you only ever meet buyers late. The good anticipation platforms already blend signals. Nobody serious is choosing one philosophy and ignoring the other.
The failure is not that the tools cannot blend signals. It is that teams hold those signals too statically. A signal that was true on Tuesday gets frozen into a score, and the score then runs the routing, the follow-up and the budget for weeks, long after the thing it described has moved.
That is why I keep coming back to the same sentence. Demand generation is a timing problem before it is a data problem. The data question, what do we know about this account, is the one everyone optimises. The timing question, what does this account need from us right now, is the one that quietly decides whether the data was worth collecting.
Which leaves the question this piece is actually about. Not whether to react or anticipate, but how you choose which one a given account needs today.
Audiences move in and out of buying
Start with the thing the Ehrenberg-Bass school has been saying for years, and which the LinkedIn B2B Institute popularised as the 95-5 rule: at any one time, the large majority of your market is not buying. Buyers move in and out of the market on their own clock, not yours.
And the clock is set by things you do not control. A contract enters its renewal window. A vendor deprecates an integration the buyer depended on. A reorganisation moves the live project to a different subsidiary. None of these show up as a tidy lead, and all of them change whether anticipation or capture is the right move this week.
So a signal today and silence tomorrow is not a contradiction. It is the normal shape of demand over time.
Most of us already know one kind of score going stale: the lead that fades because the contact stopped engaging. Hold that thought, because the decay I want to talk about is stranger. The score stays high, and it is still wrong.
Role decay: when the score is high and wrong
Here is what I mean. A person’s role on a deal changes over the life of that deal. The contact who looked like the decision-maker when your system first scored them becomes a supporter once procurement takes over, or steps aside entirely when the budget moves. The number the score shows stayed high. The role premise underneath it expired. I have started calling this role decay, to keep it separate from the ordinary kind. The lead did not go quiet. The assumption the score encoded simply stopped being true.
None of this is new as an observation. Buying-group theory has described the shifting cast of a purchase decision for decades, and the same person playing different parts is the oldest finding in it. What gets said less often is what that does to a score, which is built once and then trusted as though the role behind it were fixed.
Picture the version every revenue team has lived. An account lights up because three contacts researched a category last quarter, so the score climbs and the account is handed to sales as hot. Except the live project sits in a different subsidiary. It is a renewal, not a new purchase. The real motion runs through a procurement loop with its own technical evaluator, and that evaluator barely registers because their activity looks small next to last quarter’s research. Sales takes the meeting the score pointed at, finds the wrong division with no live project, and the actual buyer is never called. That dead meeting is the cost of trusting a number whose role assumption had already expired. You do not need subsidiaries for this to bite: the same thing happens when a champion who drove the early interest becomes a bystander the moment procurement engages.
There is a second axis to this, and it is worth naming on its own. At any one moment the same person is the decision-maker on one deal and a supporter on another, in the same company, in the same week. The score does not know which. So it is not only wrong over time, it is wrong across your portfolio at a single point in time.
The fix is not simply more data. It is to score role-in-context against the motion, not the person. Before the lead routes, the useful question is not only who this person is, but which buying motion they belong to right now, and whether the signal you are holding belongs to that motion at all. If you want to build that habit properly, it starts with how you define the buying group in the first place.
How you actually choose
This is where the two metaphors stop being decoration and start being a decision.
A lead score quietly answers two different questions at the same time. How interesting is this account, and should we pursue it now. The better teams already split fit from interest, which helps. But both of those are still measures of how interesting an account is. Neither one tells you whether to move on it today, and when a high score triggers pursuit automatically, that second question never gets asked. A high score is a reason to pay attention, not a reason to act today.
So read the two apart. The score tells you the account may matter. A separate, smaller read tells you which game to play. Before you route, run one check on the signal in front of you:
- Is it live, or is it last quarter’s?
- Is it role-relevant to this motion, or from someone playing a different part?
- Is it on an active motion at all?
Three yeses, and you kick the ball where it is: act now. Anything less, trajectory or the wrong role, and you skate to where the puck is going instead: invest in position, stay useful, and do not pursue it as though it were in-market. For any given signal the honest options are narrow. Act on it, watch it, or ignore it for this motion.
Treat the absence of a signal with the same care, because silence can be information rather than emptiness. An account that has gone quiet may be in a budget freeze, mid-reorganisation, or simply out of market for now, and each of those calls for a different move. Reading the signals themselves well is a deeper subject, and there is more on it in our work on signal architecture.
Knowing which game you are playing
The discipline here is not having both philosophies. Almost everyone has both. The discipline is knowing, account by account and week by week, which one this moment calls for, and being willing to change your answer when the role and the timing change.
Demand generation is a timing problem before it is a data problem. Get the timing read wrong and the cost is concrete: budget burned chasing a trajectory that never arrives, handoffs poisoned by scores that are high and wrong, and live demand missed because, on the day, it looked too small to matter. Get it right and you waste almost nothing.
Kick the ball where it is. Skate to where the puck is going. Just know, every time, which game you are in.
