Signals don't produce pipeline without routing

Signal-based selling is the current orthodoxy. The thesis is sound: instead of spraying cold email at strangers, you watch for intent, a page visit, a funding round, a job change, a hiring spike, and you act on it. The tools are good. Common Room, UserGems, RB2B, the rest. Teams are buying them fast.

Almost none of them have thought past the purchase.

Here is the gap. A signal fires. Then what? Who sees it? Who decides it’s worth acting on? Who picks the account up, and how fast? What’s the rule for when two signals fire on the same account on the same day? What happens to a signal that fires on a Friday afternoon, or on an account that’s already in an active deal? A signal without answers to these questions is not a pipeline input. It’s a notification, and notifications get ignored.

The pattern is everywhere. A team buys a signal tool, wires it to a Slack channel, and declares signal-based selling live. The channel pings all day. For a week, people click. By month two, the pings are noise. The signal that would have produced a deal is buried between three low-value pings about an intern’s page view and a job change at an account that’s been a customer for a year. The tool is working. The routing isn’t there, so the tool’s output goes nowhere.

A signal is the first third of a loop. The full loop is signal, routing, action. Routing is the part nobody builds, because it’s unglamorous engineering, and because the tool vendor’s demo ended at the ping.

Routing means several specific things. It means prioritization: not every signal is equal, and a scoring rule has to rank them so the top of the list is actually the one worth acting on. It means assignment: the signal goes to a specific human owner, not to a channel full of humans, because shared ownership is no ownership. It means an SLA: a signal that isn’t acted on within a defined window escalates or expires, because a stale signal is worse than no signal, it manufactures false confidence. It means suppression: a signal on an account that’s already being worked is rerouted, not duplicated.

The cost of an unrouted signal stack is two failures stacked on each other. The first is the wasted tool spend: you bought a sensor with no actuator. The second is worse. An unrouted signal stack produces a feeling of coverage. Leadership believes the team is acting on intent because the dashboard shows signals firing. The dashboard is measuring the wrong thing. It’s measuring detection, not action. A signal that fired and went nowhere is not pipeline. It’s a log entry.

The fix is to build the routing layer before you buy the second signal tool. Decide the priority order. Assign owners. Set the SLA. Wire the suppression. Make the loop close, end to end, before you add more inputs to the front of it. One signal source with a closed loop beats five signal sources feeding an open one.

One company I looked at is hiring specifically to build this. Their job description is almost entirely about standing up a signal stack. They’ve bought the inputs. What they’re hiring for is the connective tissue, the routing and prioritization and action layer that turns those inputs into pipeline. They’re the rare team that diagnosed the gap correctly. Most teams just buy another signal tool and wonder why the dashboard keeps getting fuller while the pipeline doesn’t.

A signal is not a strategy. It’s a raw input. What you do with it, how fast, and through what rules, is the system. Build that part, or accept that your signal tools are very expensive notifications.

A signal stack without routing is an expensive notification. I build the loop that closes. Worth a 30-minute look at your stack? → Book a call.