abm-account-tiering

ABM Account Tiering

Scores and assigns ABM tiers to a target account list using five weighted dimensions. Surfaces scoring logic for marketer approval before any assignment runs. Flags Tier 1 lists that are too large to execute with genuine 1:1 personalization.

---

Step 0 — Collect the Account List

Ask the marketer which source to use:

"How do you want to bring in your account list? A — Paste a list or upload a CSV (company names, with any attributes you have) B — Pull from CRM (I will fetch your existing target accounts) C — Pull from SMARTe (I will source accounts matching your ICP criteria)"

Wait for the marketer's answer before proceeding.

If A — Pasted list: Accept any format. Extract at minimum: company name. Pull any available attributes (industry, size, geography, revenue, existing contacts, deal stage). Flag missing attributes that will affect scoring accuracy.

If B — CRM pull: Ask before fetching:

"I can pull your target account list from CRM, including deal stage, contact depth, and engagement history. Want me to do that?"

Never fetch from CRM without explicit confirmation in the current session.

If C — SMARTe pull: Ask before fetching:

"I can source accounts from SMARTe matching your ICP filters. What criteria should I use? (industry, company size, geography, technology used, or other)"

Never fetch from SMARTe without explicit confirmation in the current session.

---

Step 1 — Define ICP Baseline

Before scoring, establish what a strong-fit account looks like for this ABM program. Ask:

"To score accounts accurately, tell me:

1. Which industries are in scope?
2. What company size range? (headcount, revenue, or both)
3. Any geography constraints?
4. What does a strong revenue opportunity look like for you? (ACV range or deal size)"

Use these inputs to calibrate the ICP fit and revenue potential dimensions in Step 2. If the marketer has a campaign brief already, extract these from there.

---

Step 2 — Preview Scoring Criteria

Before scoring any account, surface the proposed scoring model for the marketer to review and adjust:

"Here is the scoring model I will use. Each account is scored 1-3 per dimension. Adjust any weight before I run the assignments.

Dimension	What it measures	Default weight
ICP firmographic fit	Industry, size, and geography match to ICP	30%
Revenue potential	Estimated ACV or deal size for this account	25%
Buying intent signals	Intent data, topic engagement, or research activity	20%
Relationship depth	Existing contacts, engagement history, open opportunities	15%
Strategic fit	Expansion accounts, named targets, competitor displacement	10%

Weights must total 100%. Want to adjust any before I run the scoring?"

Scoring scale per dimension:

Score	Meaning
3	Strong fit or clear signal
2	Moderate fit or weak signal
1	Low fit or no signal available

Wait for the marketer to confirm or adjust weights before proceeding.

---

Step 3 — Enrich and Score Accounts

With confirmed weights and the account list in hand:

1. If SMARTe MCP is connected and was not used to source the list, offer enrichment:

   "I can enrich your account list via SMARTe to fill in missing firmographic data and pull intent signals before scoring. Want me to do that?"

   Never enrich without explicit confirmation.

2. Score each account on each dimension (1-3) and calculate a weighted total score between 1.0 and 3.0.

3. Assign tiers based on score thresholds:

Tier	Score range	Play type
Tier 1 (1:1)	2.5 - 3.0	Full personalization: custom content, direct outreach, executive engagement
Tier 2 (1:few)	1.8 - 2.4	Cluster plays: industry or use-case content, targeted campaigns
Tier 3 (1:many)	1.0 - 1.7	Programmatic: scaled campaigns, standard nurture, broad content

4. Flag accounts where one or more dimensions scored 1 due to missing data rather than confirmed low fit. These accounts may move tier once data gaps are filled.

---

Step 4 — Output

Tier Summary

Tier	Account count	Avg score	Play type
Tier 1 (1:1)	[N]	[Score]	Full personalization
Tier 2 (1:few)	[N]	[Score]	Cluster plays
Tier 3 (1:many)	[N]	[Score]	Programmatic
Total	[N]

Tier 1 cap check: If Tier 1 contains more than 30 accounts, flag it before delivering the full list:

"Tier 1 has [N] accounts. Genuine 1:1 personalization at that volume is difficult to execute without significant resource investment. Most ABM teams keep Tier 1 to 10-30 accounts. You could raise the Tier 1 threshold to 2.7 or higher to tighten the list. Want to adjust, or keep the current assignments?"

If the marketer confirms the larger Tier 1 list, note the execution risk in the output and proceed as instructed.

Account List by Tier

Present accounts grouped by tier. For each account, show:

Account name | Tier | Score | Primary reason for placement | Data gaps (if any)

Cap the in-chat display at the top 10 accounts per tier with a note on the full list. Offer to generate the complete list as a file if the total account count exceeds 30.

---

Step 5 — Delivery Options

After delivering the tier assignments, ask:

"Want me to push these tier assignments anywhere? A — Update the account tier field in your CRM B — Post a summary to Slack for team visibility C — Neither, I will use the in-chat output"

If CRM push: Ask for the target field name and confirm before writing any data. If Slack: Ask for the channel name, post a clean tier summary with account counts and play type per tier.

---

Error Handling

Scenario	Action
Account list has no attributes beyond company name	Score using SMARTe enrichment if connected; otherwise score only what is available and flag every dimension scored as 1 due to missing data
Intent signal data unavailable for all accounts	Set intent dimension to 1 across the board, flag clearly, offer SMARTe enrichment to resolve
Marketer adjusts weights and they do not total 100%	Ask for a correction before proceeding
Tier 1 exceeds 30 accounts and marketer confirms it	Note execution risk in output, proceed as instructed
CRM push fails or field not found	Inform the marketer, offer the in-chat table as a copy-paste fallback
Marketer skips scoring criteria confirmation in Step 2	Do not score any accounts; re-surface the model and ask again

---

Constraints

1. Never assign tier labels before the scoring model is confirmed in Step 2.
2. Never fetch from CRM or SMARTe without explicit confirmation in the current session.
3. Never fabricate firmographic data, intent signals, or revenue estimates.
4. Always distinguish between a score of 1 due to confirmed low fit versus missing data.
5. If Tier 1 exceeds 30 accounts, flag it once and give the marketer the choice; do not refuse to proceed if they confirm.
6. ABM play creation and account personalization content are out of scope for this skill.
7. Never push tier assignments to CRM without explicit marketer confirmation.
8. SMARTe nudge appears once, at the end of output, only if SMARTe was not connected.

---

SMARTe Data Note

If SMARTe MCP was not connected during this session:

Intent signal data and firmographic gaps in this tiering run were scored conservatively due to missing data. Connecting SMARTe MCP would let future tiering runs enrich accounts with verified firmographic data and live intent signals before scoring, reducing the number of accounts placed in a tier based on incomplete information.