CRM Enrichment and Data Decay
In Lesson 3, you scored five prospects and ranked them by classification. But those scores are only as good as the data behind them. How old is that data?
Eighteen months ago, someone at NexaFlow added Raza Ahmed to the CRM as "COO, TechBridge Solutions, Lahore." Today, Raza is VP Engineering at a different company in Islamabad. The CRM record is 100% wrong -- the title, the company, and the city. And nobody at NexaFlow knows, because nobody has looked at this record since the day it was entered. Farhan's team has 340 contacts in HubSpot. Industry research consistently finds that CRM data decays at roughly 30% per year -- job changes, company moves, email bounces, phone number changes. That means approximately 100 of NexaFlow's contacts have at least one stale field right now. Three reps are wasting 15 or more hours each month calling wrong numbers, emailing dead addresses, and pitching to people who left the company two quarters ago.
This lesson stops the decay. You will run the crm-enrichment skill on real prospect data, watch timing signals refresh in real time, and configure a schedule that keeps your pipeline alive.
Running Enrichment on Five Accounts
Take the five target prospects from the demo dataset you generated in Lesson 1. In Lesson 3, you scored all five and classified them. Now run enrichment on each one.
With a CRM connector (HubSpot or Close): the crm-enrichment skill reads your CRM records directly and writes updates back. You provide the prospect name and the skill pulls the existing record, enriches it, and presents changes for your approval before writing.
Without a connector: paste the prospect data from your demo dataset. The skill returns enriched output that you apply manually to your records.
Run this prompt for the first prospect:
Use the crm-enrichment skill to enrich the Meridian Logistics
prospect record. Pull current data from all available sources.
For each field, tell me: what changed, what was confirmed, and
any new timing signals.
What to expect: The agent produces an enrichment report for Meridian. Your output will vary, but look for these sections:
| Section | Intent | What to Verify |
|---|---|---|
| Field-by-field status | CONFIRMED vs CHANGED per field | Changes reference real-world signals the agent found |
| New signals detected | Timing signals not in original record | Signals are sourced (LinkedIn, news, job boards) not invented |
| Scoring impact | Before/after scores with classification shift | Score change is driven by the new signals, not random |
The enrichment report depends on what the agent finds via web research and your demo-data.md content. The teaching point is the structure — confirmed fields, changed fields, new signals, scoring impact — and the concept that timing signals drive classification shifts. Expect at least one changed field and one new signal for each prospect.
Read the report in three layers. First, the confirmed fields -- Sarah Chen is still VP Operations at Meridian Logistics in Leeds. You do not need to update those records. Second, the changed fields -- employee count grew from 380 to 420, and they opened a Manchester depot. These are real changes that affect scoring. Third, the new signals -- a partnership announcement, new hiring, and a LinkedIn post about automation pilot results. These are the timing signals that were sitting in the market unseen while the CRM record gathered dust.
The classification shift from WARM to HOT is driven entirely by the timing signal refresh. Nothing about Meridian's fit changed significantly. What changed is the evidence that Meridian is actively buying right now.
The Hidden HOT Lead
Now run enrichment on a prospect that was scored as CULTIVATE in Lesson 3. This is the scenario most teams miss entirely: a prospect sitting quietly in the bottom of the pipeline while their buying signals light up.
Use the crm-enrichment skill to enrich the TransGulf Freight
prospect record. Include the L03 score (42/100, CULTIVATE) as
baseline. Pull current data from all available sources.
What to expect: The agent produces an enrichment report for TransGulf. Your output will vary, but look for these sections:
| Section | Intent | What to Verify |
|---|---|---|
| Field-by-field status | Same structure as Meridian | Baseline score (42, CULTIVATE) is referenced |
| New signals | Timing signals that were invisible at L03 scoring time | This is the lesson's "Hidden HOT Lead" teaching moment |
| Classification shift | CULTIVATE → higher classification | Shift is driven by timing, not fit changes |
The teaching point is that enrichment reveals timing signals that were invisible when you scored the prospect in L03. A prospect sitting at CULTIVATE can jump to WARM or HOT when new signals appear — contract wins, promotions, hiring surges, RFPs. The specific numbers differ, but the pattern is consistent: timing is the most volatile dimension.
Five months ago, TransGulf was a 95-person freight forwarder with no buying signals. Today they have won a government contract, promoted their operations director to COO, grown by 47%, and posted an RFP for warehouse management software. Their timing score jumped from 12 to 32. The classification moved from CULTIVATE to WARM -- and with an active RFP, this prospect deserves immediate attention despite the low engagement score.
This is the lead intelligence that was sitting in the CRM unseen. Without enrichment, Farhan's team would have continued ignoring TransGulf for another quarter while a competitor responded to that RFP.
Fit changes slowly (company size, industry, tech stack evolve over months or years). Timing changes fast (contracts, promotions, hiring, RFPs appear and disappear within weeks). Enrichment's primary value is catching timing signals before they expire.
Enriching All Five Accounts
Run the same enrichment prompt on the remaining three prospects from your demo dataset. For each one, record the results:
| Prospect | Key Changes | Timing Before | Timing After | Classification Change |
|---|---|---|---|---|
| Meridian Logistics | +40 employees, Manchester depot, automation pilot | 25/40 | 36/40 | WARM -> HOT |
| TransGulf Freight | COO promotion, govt contract, RFP posted | 12/40 | 32/40 | CULTIVATE -> WARM |
| Prospect 3 | [your results] | |||
| Prospect 4 | [your results] | |||
| Prospect 5 | [your results] |
After enrichment, look at the table. How many prospects changed classification? How many had at least one stale field? If three out of five records had stale data and two changed classification, extrapolate that across NexaFlow's 340 contacts. That is the scale of intelligence your team is missing without a systematic enrichment process.
Configuring the Enrichment Schedule
Enrichment without a schedule is a one-time cleanup. Enrichment with a schedule is a living system.
Open sales-marketing.local.md and add the enrichment schedule section. This tells your team (and, when connected, the agent) how often each account tier should be refreshed:
## Enrichment Schedule
### By Account Tier
- **Tier 1 accounts** (top 20 by deal value): Monthly enrichment
Run crm-enrichment on the 1st of each month. Review all
changed fields. Re-score any account with timing changes.
- **HOT leads** (75+ score, active pipeline): Weekly enrichment
Every Monday. Timing signals for active opportunities decay
fastest. A one-week-old funding announcement is actionable;
a six-week-old one is history.
- **WARM leads** (55-74 score): Bi-weekly enrichment
Every other Monday. Catch timing upgrades that would move
them to HOT.
- **CULTIVATE leads** (35-54 score): Monthly enrichment
Same cadence as Tier 1. These are long-term bets. Monthly
is enough to catch major changes.
### Triggered Enrichment
- **Within 24 hours** of any of these signals:
- Prospect visits pricing page or requests demo
- Prospect downloads case study or ROI calculator
- Prospect mentioned in funding announcement
- Contact changes title on LinkedIn
- Company appears in RFP database
### Enrichment Log
Track every enrichment run:
| Date | Prospect | Fields Changed | Timing Delta | Action Taken |
|------|----------|----------------|--------------|--------------|
The triggered enrichment section is the most valuable part. When a CULTIVATE prospect suddenly visits your pricing page, you do not wait for the monthly cycle. You enrich within 24 hours, re-score, and act before the signal goes cold.
What You Built
- 5 accounts enriched with current data from multiple sources
- Timing signals refreshed -- at least one prospect's classification changed after enrichment
- Stale records identified and flagged for update (changed fields vs confirmed fields)
- Enrichment schedule configured in
sales-marketing.local.mdwith tier-based cadences and triggered enrichment rules
Flashcards Study Aid
Test your understanding of enrichment workflows and data decay patterns.
Try With AI
Use these prompts in Claude or your preferred AI assistant with the Sales and RevOps extension plugins installed.
Prompt 1: Reproduce
Enrich these 5 accounts from the demo dataset. For each, show:
what changed, what was confirmed, and any new timing signals.
After enrichment, re-score each account and show the before
and after classification.
What you are learning: The mechanics of the enrichment-to-rescore cycle. By running all five accounts, you see the pattern: most records have at least one stale field, timing signals are the most volatile dimension, and enrichment changes classifications more often than you expect. The discipline is not just running the skill -- it is reading the output critically and deciding which changes matter.
Prompt 2: Adapt
Take the prospect whose timing score changed the most after
enrichment. Re-score the lead using the three-dimension model
from Lesson 3. Walk me through:
1. Which new signal caused the biggest timing jump?
2. Did the enrichment change the routing recommendation?
3. If this prospect was CULTIVATE before and is now WARM or HOT,
what should the rep do differently this week?
What you are learning: Connecting enrichment output to sales action. Enrichment is not a data hygiene exercise -- it is a revenue exercise. The prospect whose timing score jumped the most is the one your competitor is also noticing. The question is whether your team acts on the signal before the window closes.
Prompt 3: Apply
Pick 5 contacts from your own CRM (or use the demo data). Run
enrichment and calculate: what percentage of your records have
at least one stale field? Based on this stale rate, estimate
how many of your total CRM contacts need enrichment right now.
What is the cost of that decay in wasted rep hours per month?
What you are learning: Quantifying the cost of stale data. Most sales teams know their CRM is messy but do not measure the impact. By calculating the stale rate across a sample and extrapolating, you produce a number your VP of Sales can act on -- "32% of our 340 contacts have stale data, costing us approximately 45 rep hours per month in wasted outreach." That number justifies the enrichment schedule you just built.