Home / Glossary / Signal Decay
Signals · the meta-discipline · deep dive

Signal decay. Acting on a stale signal is worse than no signal at all — it signals you weren't paying attention.

Every buying signal has a half-life — the period after which its predictive value falls below baseline cold outreach. Funding signals half-decay in ~30 days; exec moves in ~90; tech changes in ~60; hiring signals in ~120; job changes in ~90. The implication is brutal: a signal that fires Monday and triggers outreach 5 weeks later is not just a wasted opportunity — it's actively negative, because the recipient pattern-matches "I missed your moment" and reads the outreach as evidence the sender wasn't paying attention. Most outbound stacks systematically operate on signals that are 4-8 weeks too old, because the detection latency + queueing + SDR-batching pipeline introduces multiple multi-week delays before any signal reaches the prospect. This essay covers the decay curves for every major signal type with empirical half-lives, the operational discipline of signal-age scoring, the reply-rate math that explains why fast-and-imperfect beats slow-and-perfect, the comparison between latency-aware and latency-blind stacks, and the 8-step playbook for compressing the detect-to-send window.

Category: Signals · Read time: 10 min · Updated: 2026-05-24 · DECAY-1.0
TL;DR
Signal decay is the rate at which a buying signal loses predictive value over time. Every signal has a half-life — the days after which its expected conversion rate is half of its peak. Funding signals: ~30 days. Exec moves: ~90 days. Tech changes: ~60 days. Hiring spikes: ~120 days. Job changes: ~90 days. Past the half-life, conversion falls off cliff-like; past 2× half-life (60 days for funding, etc.), signals often perform worse than cold outreach because the recipient pattern-matches "I missed your moment." The cruel irony is that most outbound stacks systematically operate on signals 4-8 weeks too old — detection latency (Crunchbase: 12-72hr; ATS scraping: 24-72hr) plus internal queueing (SDR weekly batching adds 3-7 days) plus enrichment cycles (data appending adds another 2-5 days) compounds into a multi-week lag before any outreach happens. The teams that win on signals build latency-aware stacks: real-time detection, hourly briefing generation, daily SDR send queues, and signal-age decay scores baked into priority ranking. The teams that lose treat signals as monthly campaigns. The honest summary: signal freshness matters more than signal accuracy. A 90% accurate signal acted on in week 1 outperforms a 95% accurate signal acted on in week 5 by 2-3×. Compress the pipeline; the math of decay rewards speed disproportionately.

01What signal decay is

Signal decay is the rate at which a buying signal loses predictive value over time. Every signal that triggers outbound action has a finite window during which it correlates with elevated conversion probability; past that window, the signal's predictive value falls back toward baseline (or below it).

The mechanics of decay are structural, not arbitrary. Each signal type implies a downstream operational window — a window during which the buying behavior the signal predicts is actually happening. Funding rounds imply a board-pressure-to-spend window of roughly 30-60 days. Exec moves imply a first-90-days stack-audit window. Tech changes imply a 60-day adjacency-evaluation window. Once those operational windows close, the signal no longer predicts the behavior it was meant to predict.

Three properties of decay matter for outbound operations:

1. The half-life. The number of days after which conversion probability falls to half its peak. Different for each signal type. The half-life is roughly the operational design point — outreach within half-life-days captures the bulk of the value.

2. The post-half-life cliff. Most signal types don't decay linearly — they decay sharply past the half-life. A funding signal that's 45 days old isn't 50% as valuable as a fresh one; it's more like 20-30% as valuable. Practical operational implication: missing the half-life window costs more than the math suggests.

3. The negative-value tail. Past 2× half-life, many signal types produce worse outcomes than cold outreach. The reason: the recipient recognizes the signal as stale and infers the sender wasn't actually paying attention. "Saw your Series C from 4 months ago" reads as automation-noise, not as relevance.

The reframe
Signal age is itself a signal — about the sender. When a prospect receives outreach anchored on a 6-month-old event, they don't just discount the relevance — they infer that the sender's detection infrastructure is months behind, which they take as evidence the sender isn't really paying attention to them specifically. That inference is correct and is precisely why stale-signal outreach often underperforms cold outreach: the cold prospect doesn't have a fresh data point that the sender ignored.

02Decay curves by signal type

Empirical half-lives for the major signal types, drawn from aggregated outbound data across categories:

Signal half-lives · days after firing
Funding signal
Half-life: 30 days · dead zone: 90+ days
Day 03060Day 90
Tech change signal
Half-life: 60 days · dead zone: 180+ days
Day 060120Day 180
Exec move signal
Half-life: 90 days · dead zone: 180+ days
Day 090135Day 180
Job change signal
Half-life: 90 days · dead zone: 180+ days
Day 090135Day 180
Hiring signal
Half-life: 120 days · dead zone: 240+ days
Day 0120180Day 240
Product launch signal
Half-life: 21 days · dead zone: 60+ days
Day 02142Day 60

The pattern: signals with operational consequences spanning quarters (hiring, exec moves) decay slowly; signals with sharp time-boxed urgency (funding, product launches) decay fast. The right operational discipline reads the urgency from the signal type and matches outreach velocity to it. Funding signals require same-week sends; hiring signals tolerate same-month sends. One-size-fits-all weekly cadences over-invest in fast-decaying signals and under-invest in slow ones.

03Reply-rate as a function of signal age

Concrete reply-rate data for funding signals, broken down by days-since-announcement at send time. The pattern is similar (with different absolute numbers) for other signal types:

Reply rate to funding-anchored outbound, by signal age at send
Days 0-7
14-22%
Days 8-30
10-15%
Days 31-60
5-8%
Days 61-90
2-4%
Days 90+
1-2% (=cold)

Two things to notice. The gap between week 1 and week 4 is the difference between 18% and 12% reply rate — a 33% drop just from 21 days of delay. The gap between week 4 and week 12 is the difference between 12% and 1.5% — an 87% drop. The damage accelerates dramatically once past the half-life. The teams that compress their pipeline to send in week 1 are capturing 2-3× the reply rate of teams sending in week 4-6, which is the typical "we send funding outreach in our weekly cadence" pattern.

04Where the 4-8 week lag comes from

Most outbound stacks systematically introduce multi-week lag between signal firing and outreach sending. The lag stacks across four operational stages:

Stage 1: Source latency (2-7 days). The signal source itself has latency. Crunchbase: 12-72 hours behind the announcement. PitchBook: 24-96 hours. ATS scrapers: 6-24 hours. LinkedIn data partners: 4-24 hours. Even fast sources rarely fire within an hour of the underlying event.

Stage 2: Enrichment / data appending (1-5 days). The raw signal arrives without contact data, firmographic context, or scoring metadata. The enrichment pipeline that adds these usually runs in nightly or weekly batches — adding 1-5 days before the signal is queue-ready.

Stage 3: Internal queueing (3-14 days). The biggest lag source. Most sales teams batch signal briefs into weekly or monthly outreach campaigns. A signal that arrives Tuesday may not be in front of an SDR until next Monday's queue; the SDR may not work that queue until Wednesday; the outreach lands a full week+ after the signal fired.

Stage 4: Sender / template prep (1-5 days). SDRs typically draft personalized outreach in batches, then schedule sending over several days to avoid burst-sending detection. Adds another 1-5 days to the pipeline.

Total: 7-31 days from signal-fire to outreach-send, with a typical mid-point of 18-25 days. For funding signals (30-day half-life), that means most outbound sends are happening past the half-life. For product-launch signals (21-day half-life), most outbound sends are past the dead zone.

Watch for
The "fresh signal source" trap. Many teams invest heavily in signal-source freshness — paying for premium real-time feeds — while leaving the internal queueing and SDR-prep lag unchanged. The result: signals arrive in the queue in 2 days but sit there for 14 before going out. The internal lag is usually 5-10× the source lag; optimizing the wrong half of the pipeline wastes the source investment.

05Latency-aware vs latency-blind stacks

The operational difference between an outbound stack that respects signal decay and one that doesn't:

Dimension
Latency-blind stack
Latency-aware stack
Source latency
3-7 days · single source, batched ingestion
<6 hours · multi-source, real-time ingestion
Enrichment cycle
Nightly batch · 1-day lag
On-demand · <1 hour
Signal-age scoring
None · all signals queued same
Built-in · fresh signals jump queue
SDR queue cadence
Weekly batch · 3-7 day queue lag
Daily queue refresh · <1 day
Send timing
Monthly campaigns · all signals lumped
Send-on-detect · within 24-72 hours of fire
Decayed-signal handling
Sent regardless · "we already have the brief"
Auto-suppressed · stale signals filtered before send
Resulting median age at send
~22 days · past half-life for most signals
~4 days · safely within peak window
Resulting reply-rate uplift
1× baseline · signal advantage lost to staleness
2-4× baseline · full signal advantage captured

The difference between the two stacks is roughly 2-4× in effective reply rate — meaning a latency-aware stack produces the same pipeline as a latency-blind stack at 25-50% of the send volume. The operational ROI of compressing the pipeline is the largest single lever in signal-anchored outbound, and the least-pursued.

06Operationalizing signal-age scoring

The minimum-viable signal-age scoring system tags every signal with three pieces of metadata at fire time:

  • Signal type — funding / hiring / exec / tech-change / job-change / product-launch. Determines the relevant half-life.
  • Fire timestamp — when the underlying event occurred (not when you detected it). For funding rounds, this is the announcement date; for hiring, this is the posting date.
  • Decay score at present — a 0-100 score representing how much of the signal's predictive value remains. Computed as roughly 100 × 2^(-age/half-life). A funding signal at day 30 scores 50; at day 60 scores 25; at day 90 scores 12.

The score then drives three operational behaviors:

Priority routing. Fresh signals jump the queue. A funding signal at decay-score 90 outranks a funding signal at decay-score 50 even if the latter is at a higher-fit account. The math justifies it: 2× the conversion probability on the fresh signal vs the stale one.

Pre-send filtering. Signals below a minimum decay score (e.g., 20) are auto-suppressed from outbound. The marginal value at that age is below baseline cold, so the send actively damages reply-rate metrics.

SDR scheduling. SDRs see only signals fresh enough to be worth their time. Stale signals are removed from queues rather than sitting in them waiting to expire.

The teams that implement this scoring see their effective reply rates lift 30-60% within a quarter — purely from removing the stale-signal sends that were dragging the average down.

07Compressing detect-to-send

The 8-step playbook for compressing the operational pipeline from week-scale to day-scale:

  1. Audit current end-to-end latency. Measure median time from signal-event to outreach-send across each signal type. Most teams discover their median is 18-25 days; the audit makes the problem visible.
  2. Real-time signal ingestion across multi-source. Replace single-source batch ingestion with multi-source real-time pipelines. SEC EDGAR + Crunchbase + PR wires for funding; Greenhouse + Lever + LinkedIn for hiring; etc. Target: <6 hour source latency.
  3. On-demand enrichment instead of nightly batch. Trigger enrichment at signal-fire time rather than waiting for nightly runs. Add another <1 hour to the pipeline rather than another 24 hours.
  4. Build signal-age scoring into the queue. Score every signal at queue-time. Route fresh signals to top of queue; auto-suppress stale ones.
  5. Daily SDR queue refresh, not weekly batches. SDRs work their queue every day, not once a week. Fresh signals reach SDRs within 1 day of firing rather than 7.
  6. Template-first sends with rapid personalization. Pre-built signal-anchored templates that SDRs personalize in 90 seconds rather than write from scratch. Time-from-queue-to-send drops from 3-5 days to under 1.
  7. Tie SDR comp to send-latency, not just send-volume. Most SDR scorecards measure activity (sends per week). Adding "median signal age at send" as a measured KPI changes behavior dramatically.
  8. Track reply rate by signal age cohort. Maintain dashboards showing reply rate broken down by days-since-signal-fire. The data itself drives behavior — once SDRs see the 14% / 10% / 5% gradient, they prioritize differently without further intervention.

08Common mistakes

Mistake 1
Treating signal age as a nice-to-have rather than a primary scoring dimension. Most outbound stacks score signals on ICP fit + signal strength but ignore age. A fresh weak-fit signal often outperforms a stale strong-fit one. Build age into the score directly.
Mistake 2
Investing in source freshness while ignoring internal lag. Premium real-time data feeds are wasted if the internal SDR queue cadence is weekly. The internal lag is usually 5-10× the source lag; optimize that first.
Mistake 3
Sending stale signals "because we already have the brief." The sunk-cost fallacy applied to signal pipelines. A 90-day-old funding brief is worse than no brief — it actively damages the relationship. Suppress, don't send.
Mistake 4
Applying the same outreach cadence to all signal types. Product launches have 21-day half-lives; hiring signals have 120-day half-lives. Treating them on the same weekly cadence systematically misses the launch window and over-rushes the hiring window. Match cadence to decay rate.
Mistake 5
Measuring SDR performance by send-volume alone. Volume-only metrics incentivize SDRs to clear their queue in any order — usually starting with the easy or comfortable signals, not the freshest. Add "median signal-age at send" as a measured KPI to fix the incentive.
Mistake 6
Not communicating signal age to the prospect. When you do reach out fresh, say so. "Saw your funding announcement this morning" or "noticed your hiring spike over the last 2 weeks" demonstrates the freshness of your awareness, which is itself a relevance signal. Implicit freshness goes unnoticed; explicit freshness reinforces the message.
Try Mama free

The signal fired 24 hours ago. The clock is already running.

Mama detects, enriches, scores, and queues every signal with its age + half-life baked in. Fresh signals jump the queue; stale ones auto-suppress. The median signal age at send drops from 18-22 days (industry typical) to under 4. Same signals, 2-4× the reply rate, just from compressing the pipeline.