Customer Success / Account Management Beginner

30 ChatGPT prompts for account managers to upsell existing clients based on usage data

You can feel it. The dashboard says one thing, and the last QBR said another. Your buyer swears they “love the platform,” yet the login graph is flat, the seat count hasn’t moved in two quarters, and the renewal is 90 days out. This is where ChatGPT prompts for account manager upsell using usage data actually pay off - when you let the product data argue the case for you, not your gut.

I built this guide because most account managers I talk to know upsell is hiding in their usage logs. They just don’t have time to mine Mixpanel, Amplitude, PostHog, Gainsight, ChurnZero, and Totango every week. The 30 prompts below turn that raw telemetry into expansion plays you can run on Monday.

Pull quote: “74% of customer leaders say most company revenue now comes from existing customers” - ChurnZero 6th Annual Customer Revenue Leadership Study (published 2025, cited 2026).

Quick answer: how do I use ChatGPT to upsell with usage data?

You feed it the actual usage export (or a clean summary from Looker, Mode, or Sigma), tell it the buyer’s Jobs to be Done, and ask for a specific output: an expansion email, a QBR slide, a save script, a renewal justification, or a stakeholder map. The best prompts force ChatGPT to cite the data points it used, so the CSM can verify every claim before it leaves their outbox.

That’s it. Nothing magical. The magic is the discipline of running the same 30 prompts across every account in your book, every quarter.

Why 70% of expansion revenue hides in usage data

Usage-based selling means you don’t pitch a new tier because it exists - you pitch it because the customer’s own product behavior is begging for it. A team that’s hitting 92% of API rate limits on a starter plan isn’t “satisfied.” They’re about to churn or upgrade. Either way, they’re not staying still.

Three reasons usage data is the unfair advantage for account managers in 2026:

It’s the only signal the customer can’t spin. Survey answers can be polite. Usage can’t.
It compounds. Mixpanel, Amplitude, and PostHog give you cohort trends Gainsight alone can’t. The CSM who stitches them wins.
AI agents now do the heavy lifting. Gainsight’s Staircase AI and ChurnZero’s AI Agents can flag usage spikes in real time, so the prompt library below starts with the output of those agents, not raw events.

The proof? Gainsight’s customer data shows a 47% lift in expansion close rates when CS teams work from a systemized expansion playbook (Gainsight Expansion page, accessed 2026). And ChurnZero reports a 100% increase in account growth and a 2x lift in customer lifetime value for teams that pair usage signals with automated plays (ChurnZero homepage, accessed 2026).

If you’re still upselling from “how’s it going?” calls, you’re leaving 30% of your book on the table.

The 4-stage usage-based upsell framework

Every prompt below maps to one of four stages. I learned this from a blend of Winning by Design’s “Land & Expand” motion, Aaron Ross’s Predictable Revenue cadence, and what the Totango Unison AI docs describe as the “customer-led growth” loop. The four stages:

Mine - pull usage signals (logins, feature adoption, API calls, seat activation, NPS, support tickets).
Measure - score health and fit (CSAT, feature breadth, NRR, expansion propensity).
Message - turn the data into a story the buyer can repeat to their boss.
Move - drive a specific action: book a QBR, expand seats, add a module, sign a renewal, or save the account.

Run all four on every account every 90 days. That’s the job.

The 2026 benchmarks your prompts need to beat

You can’t prompt for an “above-average” expansion play if you don’t know what average is. Here are the verified 2025/2026 data points your prompts should reference:

74% of CS leaders say most company revenue now comes from existing customers - ChurnZero 6th Annual Customer Revenue Leadership Study (2025, still cited as the 2026 baseline).
Team design correlates with higher NRR - CSM + enablement + support + AM teams outperform CSM-only teams. Same ChurnZero study.
CRM + CSP + LMS + support tech stacks correlate with higher NRR - same study.
47% lift in expansion close rates when CS uses a systemized expansion playbook - Gainsight Expansion (accessed 2026).
100% increase in account growth and 2x CLV when usage signals pair with automated plays - ChurnZero (homepage, accessed 2026).
Totango is a Leader in The Forrester Wave™: Customer Success Platforms, Q4 2025 (vendor page, accessed 2026).
Gainsight is a Leader in the Gartner Magic Quadrant for Customer Success Platforms 2025 (Gainsight resource page, accessed 2026).

If you don’t have these numbers next to your expansion prompts, you’re asking ChatGPT to write fiction. With them, the prompts produce forecasts a CRO will sign.

How to use the 30 prompts

A few ground rules before you copy-paste:

Always paste the data first. ChatGPT is only as honest as the CSV you feed it. Drop in a 30-day usage summary, NPS, tickets, and the contract.
Force it to cite. End every prompt with: “Quote the exact data points you used.”
Layer personas. Re-run the same prompt with a “be the buyer’s CFO” follow-up to stress-test your message.
Save your winners. When a prompt lands, save it in a shared Gainsight CTA / ChurnZero Play / Totango SuccessBLOC / Planhat note so the next CSM inherits the script.
Verify, then send. I cannot stress this enough. AI writes the first draft. The account manager sends the email.

Ready? Let’s go stage by stage.

SECTION 1: Usage-mining prompts (Prompts 1–5)

The “Mine” stage. You can’t message what you haven’t measured.

Prompt 1: Daily usage triage digest

Purpose/context: Before you open Gainsight, pull a 7-day digest of every account’s usage health so you know who to call first. Run this every Monday.

You are a customer success analyst. I'm an account manager with 40 active accounts.

Below is a CSV export from our product analytics tool (Mixpanel) for the
last 7 days. Columns: account_id, account_name, mau, wau, dau, logins,
features_used, api_calls, seats_active, seats_paid, support_tickets,
nps_score, contract_arr.

Tasks:
1. Sort accounts into three buckets: GROWING (any usage up >10% WoW),
   STABLE (±10%), and DECLINING (down >10% WoW).
2. For each GROWING account, give me one specific expansion hypothesis
   phrased as "They are using X heavily but missing Y."
3. For each DECLINING account, give me the most likely root cause from
   the data (not vibes).
4. Tell me the top 3 accounts I should personally email today and why.
5. Quote the exact numbers that justify each call.

PASTE YOUR USAGE CSV BELOW THIS LINE.

Example output (truncated):

Acme Corp - GROWING. DAU up 18% WoW, hitting 91% of API rate limit on the Starter plan. They are using the Webhooks feature heavily but missing the Workflows module. Action: email today proposing a tier upgrade to Pro Workflows.

Pro tips:

Set the model to GPT-4o or higher for accurate number parsing. GPT-4o-minie hallucinates columns.
Run this prompt on Monday morning, then paste the output into your Gainsight C360 as call notes.
Want richer output? Add a row called “renewal_date” and ask for a separate “renewal-risk” column.

Prompt 2: Feature adoption gap finder

Purpose/context: Find the gap between seats paid and seats actually using the premium features. This is the cleanest expansion signal there is.

You are a product analyst. Here is a usage export showing feature-level
adoption for one account: [PASTE EXPORT]

For each premium feature the customer is paying for, tell me:
1. The % of paid seats that used the feature at least once in the last 30
   days.
2. The industry benchmark you'd expect for a similar account (B2B SaaS,
   $X ACV).
3. Whether this account is above, at, or below benchmark.
4. A one-sentence expansion or enablement play for each feature.

If a feature is under 30% adoption, write a 3-email enablement
nurture sequence the CSM can send this week.

End with a "single biggest expansion lever" summary in one paragraph.

Example output (truncated):

SSO (SAML): 12% of paid seats active. Benchmark for $50K+ ACV: 65%. Levers: IT-led enablement session + a 1-pager on Okta setup. Expansion lever: their IT team will pay for Advanced Access Management ($$$).

Pro tips:

Cross-reference with your Vitally or Planhat account to see if SSO has ever been raised in support tickets. If yes, the prompt just found you a save-and-expand play.
If your product has usage tiers, ask ChatGPT to map adoption → tier → next tier up and the price delta.

Prompt 3: Power-user and champion detector

Purpose/context: Find the humans inside the account who love your product. They’re your internal seller for the next expansion. This prompt turns an event stream into a stakeholder map.

Here is an event-level export for account [NAME] from Amplitude/PostHog.
Columns: user_email, event_name, event_count, last_active, role_from_crm.

Do five things:
1. Identify the top 5 power users by event count AND by recency.
   Return name, email, total events, last 7-day events.
2. For each power user, infer their likely "Jobs to be Done" from the
   events they triggered. (Example: events around exports → "I need to
   report results to my boss.")
3. Flag anyone who has 0 events in the last 14 days but used to be a
   power user. These are at-risk champions.
4. Write a 4-line LinkedIn-style message I can send each active power
   user to deepen the relationship (no selling - pure recognition).
5. Write a 4-line re-engagement message for each at-risk champion.

Quote the events that justify every inference.

Example output (truncated):

Maria Chen ([email protected]) - 1,420 events, last seen 2h ago. Likely JTBD: “I need to defend my team’s headcount to Finance by showing ROI dashboards.” Recognition message: “Maria, I noticed you’ve built 14 dashboards in the last 30 days - that’s the most of any admin I work with. Mind if I share that story in our customer spotlight? It’d help your team get credit internally.”

Pro tips:

Pair this with your Catalyst or Gainsight Relationship Map to spot if a champion is the only contact. If yes, run Prompt 11 (stakeholder expansion) next.
Use the “at-risk champion” list as the input to Prompt 26 (save plays).

Prompt 4: API and rate-limit “about to break” alert

Purpose/context: If your product has usage limits, the customer who just hit 90% is your easiest upsell. Period. This prompt makes those alerts actionable in plain English.

I'm a CSM. Below is a 30-day API usage report for 40 accounts. Columns:
account_name, plan, monthly_api_limit, api_used, %_of_limit, top_3_endpoints.

Tasks:
1. List every account above 75% of their monthly limit.
2. For each, predict whether they will HIT the limit, EXCEED it, or
   STABILIZE based on the 7-day trend.
3. If they will hit or exceed, write a 3-sentence email I can send
   today that (a) flags the issue, (b) shows the cost of doing nothing,
   (c) proposes a specific plan upgrade with a dollar amount.
4. If they will stabilize, write a one-paragraph "you're growing fast,
   want a quick optimization call?" email.
5. Tell me which of these accounts is most likely to churn in the next
   60 days if I do nothing.

Cite the endpoint usage that drove each prediction.

Example output (truncated):

Globex - 88% of limit, 14% WoW growth on /v2/exports. Will hit 100% by day 23 of next month. Email draft: “Hey Sam - your team has nearly maxed out the Starter API cap. Two of your top endpoints, /v2/exports and /v2/webhooks, are growing 18% week over week. If we don’t act, exports will start failing in ~3 weeks. A Pro plan at $1,200/mo gives you 5x the headroom and priority queues. Want a 20-min call this week to walk through the numbers?”

Pro tips:

Pipe your Mixpanel/Amplitude export into this prompt weekly. Make it a Monday ritual.
The “will hit / exceed / stabilize” prediction is the magic. Don’t skip step 2.

Prompt 5: Dormant-seat reactivation opportunity

Purpose/context: A paid seat that hasn’t logged in for 30+ days is a seat the customer will not renew. Find them before the renewal conversation.

Below is a seat-level export from our product. Columns: account_name,
user_email, seat_role, last_login, license_cost_per_seat, total_paid_seats.

Tasks:
1. List every seat with no login in the last 30 days. Group by
   account.
2. For each account, calculate the "dormant ARR" - total license cost
   of those unused seats.
3. Rank accounts by dormant ARR (highest first).
4. For the top 5 accounts, write a 3-bullet "you have $X in unused
   seats" message I can send the account admin. Make it helpful, not
   salesy.
5. Suggest a re-engagement plan: re-assign seats, run a training, or
   offer a seat-recovery credit.

Quote the exact numbers in the report.

Example output (truncated):

Initech - 14 dormant seats, $28,000 in unused annual license value. Top dormant users: [email protected] (last login 67 days ago). Message to admin: “Hi Priya - I noticed 14 of your 22 paid seats haven’t logged in for over 30 days. That’s about $28K in unused value. Want to do a 15-min seat audit this week? I can also share a quick Loom on how other admins in your space are getting their teams to a 90%+ active rate.”

Pro tips:

If dormant ARR is high, ask ChatGPT: “What is the probability the customer will downsize at renewal, and what is the recovery motion?” That becomes a 2x follow-up.
Layer this with your ChurnZero Segment data to see if any of these accounts have a low health score. Combo = save play, not just reactivation.

SECTION 2: Health-score prompts (Prompts 6–10)

The “Measure” stage. You don’t message on vibes. You message on a number.

Prompt 6: Build a Customer Health Score from raw signals

Purpose/context: You don’t need a Gainsight consultant to score an account. You need this prompt and 30 minutes.

You are a customer success strategist. I'm building a Customer Health
Score for a B2B SaaS account.

Inputs (paste below):
- Product usage: logins, MAU, % of paid seats active, top 3 features
  used, API calls
- Engagement: emails opened, QBRs attended, exec sponsor present (Y/N)
- Support: tickets last 90 days, CSAT, escalations
- Financial: ARR, payment on time (Y/N), expansion history
- Sentiment: NPS, last 3 customer replies, g2 reviews

Tasks:
1. Weight these categories (suggest weights totaling 100).
2. For each input, define a green/yellow/red threshold specific to a
   $X ACV B2B SaaS account.
3. Score the account 0–100 with a one-line rationale per category.
4. Output: a single health score, a one-paragraph summary, and a list
   of the 3 actions the CSM should take in the next 14 days.
5. Flag any category where the data is missing - never assume.

Example output (truncated):

Health Score: 72/100 (Yellow). Product usage green, engagement yellow (no QBR in 6 months), support yellow (3 unresolved P1 tickets), sentiment green (NPS 9). Action 1: book a QBR this month. Action 2: escalate the 3 tickets personally. Action 3: expand: 4 seats unused, push for 1-module upsell.

Pro tips:

This is the prompt to run before you load anything into Gainsight or Planhat. It gives you a defensible scoring rubric in 5 minutes.
The 5-category split (usage / engagement / support / financial / sentiment) is the Gainsight Industry Standard. Your version can swap weights for Planhat or Catalyst.

Prompt 7: Early-warning churn radar

Purpose/context: Save the account 90 days before the renewal call. This is the prompt that should be in every CSM’s weekly tabs.

I'm a CSM. Below is a 90-day trend for account [NAME].
Columns: week, logins, seats_active, support_tickets, nps, exec_engagement,
expansion_signals.

Tasks:
1. Identify any metric that has dropped 2+ consecutive weeks.
2. For each drop, write a one-sentence hypothesis about what's causing
   it (champion left? new competitor? budget cut?).
3. Rate the churn risk: LOW / MEDIUM / HIGH / CRITICAL.
4. If HIGH or CRITICAL, draft a 4-step save play with specific actions
   and dates for the next 14 days.
5. Draft a "we've noticed some changes - want to talk?" outreach email
   that is direct but not alarming.

Quote the trend data that drove your risk rating.

Example output (truncated):

Risk: HIGH. Logins down 4 weeks in a row (-32%). Seats active down 6 seats. Exec engagement score dropped after the May 14 exec change. Save play: Day 1 - schedule a 30-min call with new exec sponsor. Day 3 - share a tailored ROI recap. Day 7 - book a joint value-mapping workshop. Day 14 - propose a 60-day success plan with milestones. Email: “Hi Daniel, I’ve been looking at your team’s usage over the last month and noticed some changes I want to talk through. Got 20 minutes Friday?”

Pro tips:

Run this prompt on any account whose health score dropped 5+ points in 30 days. The trend matters more than the absolute.
Combine with Prompt 26 (save plays) for the full script.

Prompt 8: Expansion propensity score

Purpose/context: Not every green account is ready to expand. Score them so you don’t pitch the wrong one.

You are a revenue strategist. I'm an account manager.

Here is a 12-month view of one account: usage trends, NPS, support
volume, executive engagement, contract history, recent product
feedback, and any expansion signals (new buyers added, new departments
onboarded, RFPs).

Tasks:
1. Score expansion propensity 0–100.
2. Categorize the account: NOT READY / NURTURE / READY / OVERDUE.
3. Tell me the single biggest signal that drove the score.
4. Estimate the $ expansion range in the next 12 months, with
   assumptions.
5. Draft a 3-email outreach sequence for READY or OVERDUE accounts
   where every email cites a different data point.

If the account is NOT READY, tell me what data I need to see to
upgrade them to NURTURE.

Example output (truncated):

Score: 84/100 - READY. Biggest signal: 3 new buyers added in the IT department last 60 days; they have not yet been onboarded. $ range: $35K–$60K ARR expansion in 12 months (assumes 2 more modules + 1 seat tier upgrade). Email 1: open with the seat growth. Email 2: share benchmark of similar IT-led expansions. Email 3: propose a 30-min roadmap call.

Pro tips:

“OVERDUE” is the magic category. It means they should have expanded and you haven’t asked yet. That’s pure pipeline you can claim this week.
If your product is usage-priced, add a column for “credits consumed / plan cap” and ask the prompt to factor that in.

Prompt 9: NPS verbatim analyzer

Purpose/context: An NPS survey tells you the number. The verbatim comments tell you the upsell. Most CSMs skip this prompt. Don’t.

Below are 47 NPS verbatims from our last survey round, with score
(0–10) and account name. Cluster them into:

1. "Product gaps" - what features users want that we don't have.
2. "Adoption blockers" - what stops people from using more.
3. "Expansion signals" - what they want more of.
4. "Save signals" - comments that suggest churn risk.

For each cluster, give me:
- The exact quote(s) that drove the cluster.
- A 1-line CSM action.
- A 1-line product/PM action.

End with a 3-bullet executive summary I can paste in Slack to my CS
leadership team.

Example output (truncated):

Cluster: Expansion signals. “We’d pay for an Okta integration tomorrow.” “We have 3 more teams asking for access - they want to be billed centrally.” CSM action: respond within 24h with a quote. PM action: add to roadmap scoring.

Pro tips:

Run this monthly. The cluster shifts over time and it’s a free product roadmap.
Verbatims are gold for QBR slides. Screenshot the strongest 5 quotes.

Prompt 10: Health score downgrade postmortem

Purpose/context: When an account goes yellow, the CSM needs to know why in 60 seconds. This prompt is your on-call diagnostic.

Account [NAME] just dropped from Green to Yellow on our health score.
Here is the data that changed: [PASTE 30-DAY DELTA: usage down X%,
NPS down X points, support tickets up X, etc.]

Tasks:
1. Identify the single most likely root cause from the data.
2. List 3 alternative hypotheses ranked by likelihood.
3. Tell me the 2nd-order risk: if this continues, what happens in 60 /
   90 / 120 days?
4. Write a 3-sentence Slack message I can send my manager explaining
   the situation without panic.
5. Write a 5-sentence email to the customer that acknowledges the
   shift and proposes a recovery plan - without admitting fault we
   don't own.

End with one paragraph: "If I had 1 hour with this account, here is
exactly what I would do."

Example output (truncated):

Root cause: new admin (hired 30 days ago) was never onboarded. Logins from the rest of the team dropped 22%. 60-day risk: tickets pile up, NPS drops further, renewal at risk. Slack to manager: “Heads up - Initech dropped to Yellow. Root cause is a new admin who wasn’t onboarded. I’m running a 2-week recovery plan, will update Friday.” 1 hour: call new admin, share 1-pager, schedule 30-min team training.

Pro tips:

Add a “champion email” column from your CRM. If the champion’s last reply was 30+ days ago, this prompt will likely flag it.
Save the output as a Gainsight Timeline entry so the next CSM has context.

SECTION 3: Stakeholder expansion prompts (Prompts 11–15)

The “Message” stage starts here. You don’t just message the buyer - you message the next buyer.

Prompt 11: Multi-threading map generator

Purpose/context: Single-threaded accounts are 4x more likely to churn. This prompt turns a 3-contact account into a 9-contact expansion play.

I'm an account manager. Here is what I know about account [NAME]:
- Current contacts in CRM: [paste]
- Product usage by team/department: [paste]
- Support tickets by team: [paste]
- Recent exec moves (LinkedIn, news, internal): [paste]

Tasks:
1. Build a stakeholder map with 4 quadrants: Champion, Coach, Critic,
   Unknown. Place each known contact in a quadrant.
2. For each quadrant, suggest 1–2 ideal additional contacts we should
   be in touch with. Include their likely title and why they matter.
3. For each Unknown, draft a warm-intro ask I can send my Champion
   (1 paragraph).
4. Identify the "executive sponsor" gap: who in the C-suite should we
   know and what's the lowest-effort path to them?
5. Suggest 3 LinkedIn search queries to find the missing contacts.

Cite the data that drove every placement.

Example output (truncated):

Champion: Maria Chen (existing). Coach: David Park (existing, low engagement - re-activate). Critic: none on record. Unknowns to add: VP Engineering (technical decision maker), CFO (budget gatekeeper for next renewal), Head of Data Science (heavy user of our analytics module). Intro ask to Maria: “Maria, as we map out year 2, I want to make sure we’re aligned with your finance and engineering leadership. Would you be open to intro-ing me to your VP Eng and your CFO? I’ll keep it short and bring value.”

Pro tips:

If you use Planhat or Catalyst, paste the relationship data straight from the platform.
Pair this with Prompt 3 (power-user detector) to see if any of your Unknowns are already power users in the product.

Purpose/context: When the VP-level contact has gone cold, you don’t pitch - you re-engage with a value moment. This is the prompt.

The executive sponsor at [ACCOUNT] has not replied to my last 3
emails. Their last login was [X] days ago. They were initially very
engaged at contract signing.

Tasks:
1. Diagnose 3 possible reasons for the silence, ranked by likelihood.
2. Write a 4-sentence re-engagement email that does NOT sell. The
   email's only job is to get a 15-min conversation.
3. Suggest 3 "reason to reply" hooks I can offer (an invite to our
   executive roundtable, an industry benchmark report, a peer
   introduction).
4. Draft a 2-line LinkedIn DM as a backup channel.
5. If they still don't reply in 14 days, what is the escalation path
   inside my own company?

Example output (truncated):

Likely reasons: (1) role change - exec sponsors often rotate after 12 months. (2) champion left and didn’t hand off. (3) they’re being polite and not engaged. Email: “Hi Carla, no pitch in this email - I just want 15 minutes to share what 3 of your peers in [industry] are doing with [category] in 2026. Two of them mentioned your name. Worth a call?”

Pro tips:

The “two of them mentioned your name” hook is gold. Use it sparingly, only when true.
If they still don’t respond, that’s data. Move them to a low-touch journey in Customer.io or Iterable.

Prompt 13: Champion enablement kit

Purpose/context: Your best expansion sellers aren’t your AEs. They’re the champions inside the account. Give them a kit and they sell for you.

My champion at [ACCOUNT] wants to advocate internally for an
expansion. They told me: "[paste verbatim]."

Their title is [TITLE]. They report to [BOSS]. They care about [1-2
business outcomes].

Tasks:
1. Build a 1-page internal pitch deck outline I can co-create with
   them. 5 slides max.
2. Draft a 3-bullet email the champion can forward to their boss to
   request budget for the expansion.
3. Suggest 3 pieces of social proof (case studies, G2 reviews, peer
   quotes) I can give the champion to use internally.
4. Predict the 2 most likely objections their boss will raise, and
   write 1-line responses to each.
5. Draft a 90-day joint success plan both companies can co-sign.

Make the language the champion's, not mine. Use the words they used
in the verbatim above.

Example output (truncated):

Slide 1 title: “Why we should expand [Product] to the EU team” (champion’s phrasing). Email to boss: “Hi [Boss], I’ve attached a 1-pager. Three of our sister teams are asking for access. Total cost is $18K/year and we can defer the data warehouse migration by 6 months. Can we get 20 minutes on your calendar to discuss?” Objection 1: “Why not wait for next year’s budget?” - Response: “We can defer the data warehouse line item in exchange, net neutral.”

Pro tips:

Always use the champion’s verbatim words. It signals you listened.
The “defer another line item” objection-handler is one of the most effective expansion moves in B2B.

Prompt 14: Power-vs-power user mismatch (organizational redesign)

Purpose/context: Sometimes the account is using the product brilliantly in one team and never in another. That gap is the expansion play.

Below is a breakdown of usage by team/department for [ACCOUNT]:
[PASTE: team, mau, %_of_paid_seats, top_features_used, last_30d_trend]

Tasks:
1. Identify the "power team" (highest usage) and the "ghost team"
   (lowest usage despite paid seats).
2. Hypothesize why the ghost team isn't using the product. Use these
   categories: wrong persona, wrong use case, missing training, missing
   internal champion, or a feature gap.
3. Draft a 3-step "activate the ghost team" play for the CSM.
4. Estimate the $ impact if we activated the ghost team at 50% of
   power-team levels.
5. Draft a 1-line message I can send the ghost team's manager
   (different from the champion) offering a tailored 20-min walkthrough.

Cite the team-level data that drove the diagnosis.

Example output (truncated):

Power team: Customer Success (92% active). Ghost team: Engineering (12% active despite 8 paid seats). Why: wrong persona - engineering needs API-first features, not the CS dashboards. Activation play: (1) book a 20-min with the Eng VP, (2) share our developer quickstart, (3) pilot one API workflow in 14 days. $ impact: if Eng hits 50% activation, +$24K ARR. Message to Eng VP: “Hi Tomas, I want to show your team a 15-min API workflow that 3 of your peer engineering orgs shipped in the last quarter. Worth a look?”

Pro tips:

Run this prompt on every account with 30+ seats. The ghost-team gap is the most under-utilized expansion signal.
Cross-reference with your Looker or Mode dashboards for team-level data. Mixpanel’s “DAU by team” report is a perfect input.

Prompt 15: Joint roadmap session invitation

Purpose/context: The single highest-leverage expansion meeting is the joint roadmap session. Use this prompt to get it booked.

Account [NAME] is a 4-year customer, $120K ARR, 60 seats. Usage is
strong, NPS is 9, and the champion has hinted at "wanting to see
where the product is going."

Tasks:
1. Write a 5-sentence email to the champion proposing a 45-min joint
   roadmap session. Include a clear agenda.
2. Draft 3 reasons the champion would say yes (each tied to their
   stated business goals).
3. Draft 3 objections they might raise and 1-line responses.
4. Outline the agenda: 5 slides, 5 questions, 1 clear next step.
5. Suggest 1 piece of "exclusive" content I can offer only to
   strategic accounts like this (a beta feature invite, an exec
   dinner, a customer advisory board seat).

Keep the tone peer-to-peer, not vendor-to-customer.

Example output (truncated):

Email: “Hi Priya, you’re one of 4 customers we want to share our 2026 product roadmap with before the public announcement. 45 minutes, no slides you’ll see in 6 months, just the unfiltered roadmap and your feedback. Worth a Wednesday next week?” Reasons to say yes: (1) preview access to the AI module, (2) input on the SSO refresh, (3) invitation to our Q3 Customer Advisory Board in NYC.

Pro tips:

Roadmap sessions convert at 3–4x the rate of generic QBRs. Use this for top-20 accounts only.
Pre-load the prompt with a real roadmap item you can demo. Don’t fake it.

SECTION 4: QBR / EBR prompts (Prompts 16–20)

Quarterly Business Reviews and Executive Business Reviews are the highest-leverage meetings a CSM runs. These prompts make them data-driven in 10 minutes.

Prompt 16: QBR deck builder (one prompt, full deck)

Purpose/context: A QBR deck used to take 4 hours. It shouldn’t. This prompt builds the skeleton, you fill in the screenshots.

I'm a CSM. Build a 12-slide QBR deck for account [NAME] with the
following inputs:
- 12-month usage summary (paste)
- Support ticket history (paste)
- NPS trend (paste)
- ROI/Value delivered (paste or estimate)
- Roadmap alignment (paste 3-5 upcoming features)
- Expansion opportunities (paste)

Slide structure (one slide per section, max 5 bullets per slide):
1. Title - Account name, QBR date, attendees.
2. Recap of last quarter's commitments (who said they'd do what).
3. Usage health - what we said vs. what happened.
4. Value delivered - concrete ROI in their language.
5. Adoption gaps - features they paid for but don't use.
6. Stakeholder map - who's engaged, who's missing.
7. Risks & open issues.
8. Customer asks (from emails, tickets, NPS).
9. Roadmap preview - 3 features aligned to their goals.
10. Joint success plan - next 90 days, who owns what.
11. Expansion options - 2-3 with $ estimates.
12. Asks of the customer (referrals, case study, exec intro).

For each slide, give me the bullet points AND a 2-sentence speaker
note the CSM can use live.

Example output (slide 4, truncated):

Slide 4 - Value delivered. Bullets: (a) Saved 14 hours/week in manual reporting. (b) Cut QA cycle from 5 days to 2. (c) $340K in labor savings (estimated). Speaker note: “Maria, when we kicked off in Q1, you told me the goal was to free up Priya’s team for higher-leverage work. Here’s what the data shows. Notice the QA cycle time - that was your #1 ask.”

Pro tips:

Run this prompt before the meeting. Send the deck to the champion 48h in advance with “want to make sure I have it right” framing.
Use the speaker notes as your teleprompter. It kills the “um, let me check the data” moment.

Prompt 17: ROI calculator for the EBR

Purpose/context: Executive Business Reviews live and die on one number: ROI. This prompt builds the number honestly.

I'm a CSM. Build a 1-page ROI calculation for account [NAME] using
these inputs:
- Number of users on the platform: [N]
- Average loaded cost per user per year: $[X]
- Hours saved per user per week (from usage logs or survey): [H]
- Number of incidents/risks avoided (from logs): [N]
- Cost per incident avoided (estimate): $[X]
- Annual contract value: $[X]

Tasks:
1. Show the math step by step.
2. Calculate total annual value, total cost, net ROI, and ROI %.
3. Sanity check: is this ROI realistic for a B2B SaaS product in
   [CATEGORY]? If not, flag the inputs that look too generous.
4. Suggest 2-3 caveats to include so the customer trusts the number.
5. Translate this into a single sentence the customer's CFO can
   repeat.

Provide a 1-paragraph executive summary I can paste into an email.

Example output (truncated):

Total value: $487,000. Total cost: $96,000. Net ROI: $391,000. ROI %: 407%. Caveat: the “hours saved” figure is from a self-reported survey with a 22% response rate - adjust with a 0.6 confidence multiplier. CFO sentence: “Our Q3 review showed a 4x return on the platform - about $390K in net value - driven mostly by a 60% reduction in QA cycle time and a 35% drop in support escalations.”

Pro tips:

Always show the math. CFOs trust math more than adjectives.
The “0.6 confidence multiplier” caveat is a magic phrase. It signals rigor.

Prompt 18: Executive Business Review (EBR) narrative arc

Purpose/context: A QBR is data. An EBR is a story. This prompt turns your 12 months of data into a 3-act narrative.

I'm running an Executive Business Review with the CEO and CFO of
[ACCOUNT] in 2 weeks. The goal is to expand the relationship, not
just retain it.

Inputs:
- 12-month product usage (paste)
- Major wins they had (paste)
- Major friction points (paste)
- 3 strategic goals they shared in our last EBR (paste)
- 3 upcoming product launches aligned to their goals (paste)

Write a 3-act narrative for the meeting:
- ACT 1 (5 min): Where we started, the goals they set, the partnership
  so far. Make it warm.
- ACT 2 (15 min): What we delivered, what they delivered, the
  measurable outcomes. Make it specific.
- ACT 3 (10 min): Where we're going - the 3 product launches, the
  expansion opportunity, the joint 12-month plan. Make it a choice
  they want to make.

For each act, give me 3 key talking points and 1 data point to cite.
End with a 1-sentence "ask" for the meeting.

Example output (Act 1, truncated):

Talking points: (1) you came to us in Q1 with a goal to centralize ops reporting. (2) you committed to a 90-day rollout. (3) we’ve now been partners for 4 quarters. Data point: “Your team has logged in on 247 of the last 248 workdays.”

Pro tips:

The “247 of 248 workdays” detail is the kind of line that makes a CFO’s eyes light up. Always include a specific number.
Rehearse the narrative out loud twice. The EBR is performance art.

Prompt 19: Customer success Q&A prep (for tough QBRs)

Purpose/context: Sometimes you walk into a QBR and the room is tense. This prompt preps you for the worst.

I'm walking into a QBR for [ACCOUNT]. The relationship is Yellow.
Support tickets are up, logins are down, and the last email from the
customer was curt.

Predict the 5 hardest questions the customer might ask, and for each:
1. Write the question in their voice (skeptical, direct).
2. Write a 3-sentence honest response I can deliver without flinching.
3. Identify the data point that backs the response.
4. Suggest the one question I should turn around and ask them in
   return.

End with a 1-paragraph "tone setting" opener for the meeting.

Example output (truncated):

Q: “We’re paying $96K a year and our team is still complaining. What are we actually getting?” A: “That’s a fair question, and I want to be specific. Looking at the data, your power users save 14 hours a week. The complaints are concentrated in 2 teams out of 8 - which I think is a rollout gap, not a product gap. Can we spend 10 minutes mapping which teams are using it and which aren’t?” Data point: “70% of seats are active; 30% are not - and the 30% cluster in 2 specific departments.”

Pro tips:

The “ask it back” move is the CSM’s secret weapon. It reframes a complaint as a joint problem.
Read the answers out loud once. The “without flinching” requirement is real.

Prompt 20: QBR follow-up email (within 24 hours)

Purpose/context: The QBR ends. 24 hours later, you send this email. It cements the action items and sets up the expansion.

The QBR for [ACCOUNT] just ended. Here are my raw notes (paste).

Tasks:
1. Extract the 3-5 action items, with explicit owners (their side and
   our side) and due dates.
2. Write a follow-up email to the champion that:
   - Thanks them for their time
   - Recaps the 3-5 action items in a numbered list
   - Restates the value delivered with one specific data point
   - Confirms the date of the next QBR
   - Teases the expansion opportunity with a soft CTA
3. Draft a 2-line slack message to my own team summarizing the meeting
   in plain English.
4. Flag any commitments the customer made that I should track in
   Gainsight / ChurnZero / Totango as a CTA.

Tone: peer-to-peer, calm, confident. No exclamation points.

Example output (truncated):

Email subject: “Recap + 4 next steps from today’s QBR.” Action items: (1) Maria to share new dashboard requirements by Friday June 14. (2) We deliver custom training for the Eng team by June 28. (3) Maria introduces us to the EU team lead by July 5. (4) Joint review of expansion options by July 19. Soft CTA: “If today’s conversation on the AI module was useful, I can pull together a 1-pager for the EU team. Want me to send a draft?”

Pro tips:

Send the recap within 4 hours, not 24. Speed signals professionalism.
The “soft CTA on the expansion we discussed” is the prompt’s hidden genius. Most CSMs recap the meeting. Few recap the opportunity.

SECTION 5: Renewal & cross-sell prompts (Prompts 21–25)

The “Move” stage. You don’t move without money on the table.

Prompt 21: Renewal forecast narrative

Purpose/context: Your CRO wants to know which renewals are real and which are wishful thinking. This prompt produces a defensible forecast.

Account [NAME] is up for renewal in [X] days. Inputs:

- Current ARR: $[X]
- Health score: [0-100]
- Usage trend (12 months): [paste]
- NPS: [X]
- Support escalations last 90 days: [paste]
- Champion status: [stable/at-risk/changed]
- Competitive intel: [paste]
- Expansion signals: [paste]

Tasks:
1. Predict renewal probability: SAFE (>90%), LIKELY (70-90%),
   AT-RISK (40-70%), CRITICAL (<40%).
2. For each, give a 2-paragraph rationale with cited data.
3. Estimate the renewal ARR range (floor, base, upside).
4. List the 3 leading indicators that would *upgrade* the forecast
   and the 3 that would *downgrade* it.
5. Recommend a 90-day action plan to defend and grow the renewal.
6. Draft a 3-sentence email to my CRO summarizing this forecast.

No hedging. Make a call.

Example output (truncated):

Forecast: LIKELY (78% renewal probability). Health score 72, usage up 14% YoY, NPS 9, but champion just left and 1 new exec sponsor is unproven. ARR range: floor $96K (no expansion), base $108K (modest expansion), upside $144K (full module add). Top 3 downgrade risks: (1) new exec sponsor disengages, (2) 1 unresolved P1 ticket, (3) budget cut in Q3. CRO email: “Acme renews July 28. 78% probability. Base case $108K, upside $144K. Watch list: new exec sponsor. Running a 90-day close plan.”

Pro tips:

Save this output as a Gainsight Renewal Center note. Your manager will love you.
The “make a call” instruction is critical. ChatGPT hedges by default. Force a number.

Prompt 22: Cross-sell engine (the “next product” suggestion)

Purpose/context: You’ve sold Product A. They might need Product B. Find the bridge.

Account [NAME] is on our [CURRENT PRODUCT/SKU]. Below is their usage,
support tickets, NPS, and any feature requests.

Catalog of products we offer:
[PASTE: product_name, one-line description, target persona, price band]

Tasks:
1. For each product in the catalog, score cross-sell fit 0-100 based on
   the account's data.
2. Explain the top 3 cross-sell candidates in 3 sentences each,
   citing the usage data that drove the score.
3. For the #1 candidate, draft a 4-email nurture sequence that
   demonstrates value before asking for the meeting.
4. Identify any product in the catalog that is a bad fit and explain
   why. Don't waste their time.
5. Estimate the cross-sell $ in the next 12 months.

Tone: consultative, not salesy.

Example output (truncated):

#1 candidate: Analytics Module (score 89/100). Account uses our dashboards daily but constantly asks for custom reports - 14 requests in 6 months. Email 1: share 3 case studies of similar accounts. Email 2: invite to a 20-min demo. Email 3: offer a 30-day pilot. Email 4: propose the procurement-ready quote. $ estimate: $24K–$36K ARR.

Pro tips:

If you use Gainsight, save the cross-sell score as a CTA in C360.
The “14 requests in 6 months” is a beautiful expansion signal. Always count the asks.

Prompt 23: Price increase justification letter

Purpose/context: Inflation, packaging changes, AI-features added. Sometimes you have to raise price. This prompt makes the case before the customer makes a stink.

I'm about to send a price increase notice to [ACCOUNT]. The increase
is [X%], effective [DATE]. The new pricing reflects [reasons:
inflation, new AI features, expanded support, etc.].

Tasks:
1. Draft a price-increase letter that:
   - Leads with the value delivered, not the price
   - Cites 3 concrete product investments in the last 12 months
   - Anchors the new price to outcomes, not features
   - Offers a 30-day grace period and a multi-year lock-in discount
2. Anticipate the 3 most likely objections and write 1-line responses.
3. Draft a 2-line Slack message I can send my manager pre-empting
   any pushback.
4. Identify 1 trade-off the customer might ask for (more seats,
   longer term, training credits) and pre-decide whether to offer it.
5. End with a single sentence: "If they push back, here is the one
   thing I can give them."

Example output (truncated):

Letter opening: “Hi Maria, before any pricing conversation, I want to lead with the data. In the last 12 months, your team has saved 14 hours/week, cut QA cycle time 60%, and grown seat adoption to 92%. The 8% price increase reflects the AI module, expanded SOC 2 coverage, and 24/7 support - three investments you asked for.” If they push back: “I’ll extend the current price for 6 months in exchange for a 2-year commitment.”

Pro tips:

Always lead with value, end with the ask. Order matters.
Pre-approve the “if they push back” trade-off with your manager before you send the letter.

Prompt 24: Multi-year renewal with expansion

Purpose/context: A 3-year deal is the best deal. This prompt structures the offer.

Account [NAME] is up for renewal in [X] days. Current ARR $[X].
Goal: close a 3-year renewal with a [Y%] ARR uplift and add [MODULE].

Tasks:
1. Build a 3-tier offer: 1-year / 2-year / 3-year, with the discount
   and lock-in terms for each.
2. For the 3-year option, list the 4 non-financial benefits the
   customer gets (roadmap input, exec access, beta features,
   dedicated CSM).
3. Draft a 6-sentence email to the CFO framing the 3-year option as
   the "predictable cost" choice.
4. Identify 2 risks of a 1-year deal that I can honestly raise (price
   exposure, support tier downgrade, roadmap lock-out).
5. Write a 3-line LinkedIn DM to the CFO as a backup channel.
6. End with the single most important question to ask in the renewal
   call.

Example output (truncated):

3-year offer: $130K/year (vs. $144K list), total $390K, +5 seats, beta access to AI module, named CSM. CFO email: “Hi Daniel, locking in for 3 years gives you predictable cost, beta access, and 5 additional seats - at a 10% discount versus annual. Worth a 20-min call to walk through the option?”

Pro tips:

Multi-year deals don’t just lock in revenue - they lock out competitors. Make this clear in the email.
The “named CSM” benefit often closes deals at the C-suite level. Don’t underestimate it.

Prompt 25: Customer reference and case study ask

Purpose/context: Every expansion deal is easier when the prospect talks to a peer. Get the case study before you need it.

Account [NAME] is a 4-year customer, $120K ARR, NPS 9, and just
expanded to the EU team. I want to ask for a case study + reference
customer agreement.

Tasks:
1. Identify the 2-3 strongest stories from their tenure (with usage
   data as evidence).
2. Build a 1-page case study outline with these sections: customer
   background, the problem, the solution, the outcome (with numbers),
   a quote, and a forward-looking statement.
3. Draft a 4-sentence email to the champion asking for a 30-min
   interview. Frame it as a chance for *them* to get credit, not as a
   favor to us.
4. Suggest 3 "low-effort" ways they can participate (10-min Loom, a
   200-word written quote, a 30-min Zoom).
5. Identify the 2 people inside the account most likely to say yes
   and the 1 person most likely to say no.
6. Predict what they'll ask in return and pre-decide what we can
   offer (G2 review, case study co-branding, conference speaking slot).

Example output (truncated):

Email: “Hi Priya, what your team has done with [Product] over the last 4 quarters is one of the best stories I get to tell. Would you be open to a 30-min interview? I can write the first draft and only ask for your sign-off. In return, we’d love to feature you in our customer spotlight and offer a speaking slot at our user conference in October.”

Pro tips:

“I can write the first draft” is the magic phrase. It removes 90% of the friction.
Pair this with Prompt 13 (champion enablement kit) to give the champion an internal pitch they can use to say yes.

SECTION 6: Save-the-customer prompts (Prompts 26–30)

The “Move” stage, but the move is retention. Sometimes saving the account IS the expansion - because the alternative is zero.

Prompt 26: Save play for a Yellow account

Purpose/context: You have 30 days. The account is drifting. This is the prompt.

Account [NAME] is Yellow on health, with usage down 18% over 60 days.
Champion is still engaged. CFO is unknown to me. Last QBR was 6 months
ago. 1 open P1 support ticket.

Tasks:
1. Diagnose the 3 most likely root causes, ranked.
2. Build a 30-day save plan, day-by-day for weeks 1 and 2, with weekly
   milestones for week 4.
3. Identify the 3 internal people I need to engage (CSM, support lead,
   executive sponsor on our side).
4. Draft a 5-sentence outreach email to the champion that:
   - Acknowledges the gap without panic
   - Names the specific issues
   - Proposes a recovery plan
   - Asks for a 30-min call
5. Identify the 1 piece of "quick win" content I can deliver in the
   next 7 days to rebuild momentum (a tailored ROI recap, a
   competitive teardown, a peer benchmark).
6. If the champion doesn't respond in 7 days, what is the next move?

Cite the usage data driving the diagnosis.

Example output (truncated):

Root cause #1: new admin never onboarded. Day 1: schedule call with new admin. Day 3: send tailored ROI recap. Day 7: book 30-min with CFO (intro via champion). Day 14: joint success plan signed. Quick win: send a 1-pager showing 3 peer accounts that hit 90% activation in 60 days. If champion doesn’t respond: engage their manager via LinkedIn, share the quick-win asset, offer a no-strings-attached workshop.

Pro tips:

The “30-min with CFO” is the move that saves most Yellow accounts. The champion’s manager has budget authority and is a fresh pair of eyes.
Always have a “no strings attached workshop” as a backup offer. It removes the customer’s “this is a sales call” reflex.

Prompt 27: Save play for a Red account (last chance)

Purpose/context: 60 days from renewal, health is red, usage is collapsing. The save prompt is different. It’s honest.

Account [NAME] is Red on health. Usage down 35% in 90 days. Last
login from champion 21 days ago. 4 open P1 tickets. Renewal is 60
days out.

Tasks:
1. Decide: is this account savable, or is it time to negotiate an
   orderly exit? Give me the honest call with a 2-paragraph rationale.
2. If savable: build a 60-day, 6-step save plan. Each step must have
   a specific deliverable, a date, and a named owner.
3. If not savable: draft a graceful "offboarding" email that preserves
   the relationship for a possible 2027 re-engagement.
4. Identify the 1 person (inside or outside my company) most likely
   to turn this around.
5. Draft a 5-sentence "executive-to-executive" outreach to the
   customer's C-suite. Tone: candid, respectful, low-pressure.
6. End with a 1-paragraph internal postmortem I can share with my
   team so we learn from this regardless of outcome.

No sugar-coating. Be direct.

Example output (truncated):

Honest call: savable, but 50/50. Usage drop is real but the contract is large and the champion hasn’t quit. 60-day plan: Day 1 - exec-to-exec call. Day 7 - close all 4 P1 tickets personally. Day 14 - joint success plan with new admin. Day 30 - usage review with CFO. Day 45 - expansion offer (last attempt). Day 60 - renewal close. Exec email: “Hi Carla, I want to be candid. Your team’s usage has dropped 35% and there are 4 open tickets I’d like to close personally this week. Can I have 20 minutes with you and Daniel?”

Pro tips:

“Be direct” is the instruction ChatGPT needs to skip the corporate fluff.
The internal postmortem is the most valuable output. Even if you lose the account, you learn.

Prompt 28: Competitive displacement defense

Purpose/context: Your customer just told you they’re “evaluating alternatives.” Time to defend without being defensive.

Account [NAME] just told me they are evaluating [COMPETITOR]. I have
30 days to defend.

Tasks:
1. Predict the 3 most likely reasons they're evaluating (price, missing
   feature, champion change, internal politics).
2. For each, write a 2-paragraph response I can use in a call. The
   response must be honest, not disparaging of the competitor.
3. Build a "why us" comparison with 4-5 dimensions most relevant to
   this customer's stated priorities.
4. Identify 3 customer references I can offer for peer conversations.
5. Draft a 4-sentence email acknowledging the evaluation and proposing
   a joint 30-min "honest comparison" call.
6. If I lose this account, what is the offboarding motion that
   preserves the door for 2027?

Cite the customer's own stated goals in the response.

Example output (truncated):

Likely reasons: (1) pricing - competitor is 30% cheaper. (2) AI features - competitor launched an AI module last quarter. (3) champion change - new admin is being courted. Response on price: “I hear you on price. Before we compare line items, can we look at total cost of ownership? Most of our customers find we’re 18% cheaper once support time and integration costs are factored in.”

Pro tips:

Never disparage the competitor. It signals desperation.
Peer references close deals that comparisons can’t. Always offer 3.

Prompt 29: Save play when the champion has left

Purpose/context: The single most common save scenario in 2026. Your champion just changed roles. Use this prompt to rebuild.

Account [NAME]'s champion [NAME] just left for a new role. I'm
inheriting a new stakeholder [NAME] who is unknown to me.

Tasks:
1. Build a 3-paragraph "warm re-intro" email I can send the new
   stakeholder. Frame it as a continuation, not a sales pitch.
2. Identify the 3 questions I should ask the new stakeholder in our
   first 30-min call.
3. List the 3 assets I should send in advance (a 1-pager on their
   current usage, a peer benchmark, a roadmap preview).
4. Predict the 2 most likely objections the new stakeholder will
   raise ("why are we paying for this?" / "what was the previous
   champion's view?") and write 1-line responses.
5. Identify 1 person at the new stakeholder's old company (or in their
   network) who could give me a warm intro.
6. Draft a 2-line LinkedIn DM as a backup.

Tone: warm, low-friction, helpful. No selling in the first touch.

Example output (truncated):

Email: “Hi Daniel, I just learned you’re stepping into Maria’s role - congratulations. I’d love to introduce myself and bring you up to speed on the work Maria’s team has done. I have a 1-pager on your current usage and a few peer benchmarks I think will be useful. No pitch in this email - just 30 minutes to get oriented. Worth a call next week?”

Pro tips:

The “no pitch in this email” line is the single most important phrase in champion-replacement saves.
Always ask the outgoing champion (politely) for a 5-min intro. A warm handoff beats a cold start.

Prompt 30: The end-of-quarter save-everything sprint prompt

Purpose/context: It’s the last week of the quarter. You’ve got 5 Yellow accounts, 2 Reds, and a CRO who wants to know the number. This prompt is your triage.

End of quarter. I have 7 at-risk accounts: 5 Yellow, 2 Red.
Total ARR at risk: $[X]. Renewal dates: [list].

Tasks:
1. Rank the 7 accounts by save probability.
2. For the top 3 (highest save probability), give me a specific
   next-action-with-a-date for each.
3. For the bottom 2 (lowest save probability), draft a graceful
   "we tried" email that preserves the relationship.
4. For the middle 2, draft a "we're here, let's talk" 3-sentence
   outreach.
5. Tell me the realistic renewal forecast for the group: floor,
   base, upside, with assumptions.
6. Identify the 1 thing I should ask my manager for *today* that
   would materially help (exec air cover, pricing flexibility,
   product escalation).
7. End with the single sentence I should say to my CRO.

Be direct. Make a call.

Example output (truncated):

Save probability ranking: (1) Acme 65%, (2) Initech 55%, (3) Globex 50%, (4) Stark 40%, (5) Wayne 35%, (6) Oscorp 25%, (7) Umbrella 20%. Top 3 next actions: (1) Acme - exec call by Friday, (2) Initech - close 2 P1 tickets today, (3) Globex - book joint success plan with new admin. Forecast: floor $182K, base $241K, upside $310K (out of $385K total). Ask manager for: exec air cover for the Acme call. CRO sentence: “We’re going to land between $241K and $310K of the $385K. Top save is Acme; biggest miss risk is Umbrella.”

Pro tips:

The “single sentence to my CRO” forces clarity. Save it for the Friday forecast.
“Be direct. Make a call” is the prompt that turns a CSM into a sales leader.

Comparison table: prompt category vs. stage vs. output

Use this table as a quick reference when you need to pick the right prompt fast.

#	Prompt category	Stage	Primary input	Primary output	Best for
1	Daily usage triage digest	Mine	Mixpanel/Amplitude CSV	Sorted account list + actions	Monday morning
2	Feature adoption gap finder	Mine	Feature-level export	Expansion hypotheses	Under-used premium features
3	Power-user detector	Mine	Event-level export	Stakeholder map	Multi-threading
4	API rate-limit alert	Mine	API usage report	Tier-upgrade emails	Quota-stressed accounts
5	Dormant seat reactivation	Mine	Seat-level export	Admin outreach	Renewal protection
6	Health score builder	Measure	Multi-signal data	0–100 score	Scoring without Gainsight
7	Churn radar	Measure	90-day trend	Save play	At-risk accounts
8	Expansion propensity	Measure	12-month view	$ estimate + outreach	Pipeline prioritization
9	NPS verbatim analyzer	Measure	NPS comments	Clustered themes	Roadmap + expansion
10	Health downgrade postmortem	Measure	30-day delta	Recovery plan	Yellow account triage
11	Multi-threading map	Message	CRM + product data	Stakeholder map	Single-threaded accounts
12	Executive sponsor reactivation	Message	CRM + inbox	Re-engagement email	Cold exec contacts
13	Champion enablement kit	Message	Champion verbatim	Internal pitch deck	Internal selling support
14	Power vs ghost team	Message	Team-level usage	Activation play	Cross-department expansion
15	Joint roadmap session	Message	Account profile	Meeting invite	Top-20 accounts
16	QBR deck builder	Message	Year of data	12-slide deck	Quarterly review
17	ROI calculator	Message	ROI inputs	CFO-ready number	EBR / CFO meeting
18	EBR narrative arc	Message	Year of data	3-act story	Executive meeting
19	Q&A prep for tough QBRs	Message	Yellow account data	Hard Q&A drill	At-risk QBR
20	QBR follow-up email	Message	QBR notes	Action-item recap	Post-meeting
21	Renewal forecast	Move	Renewal inputs	Probability + $ range	CRO forecast
22	Cross-sell engine	Move	Account profile	Top-3 cross-sell picks	Product catalog expansion
23	Price increase letter	Move	Increase details	Justified notice	Annual price uplift
24	Multi-year renewal	Move	Renewal terms	3-tier offer	Locking in 3-year deals
25	Customer reference ask	Move	Champion profile	Case study outline	Building references
26	Yellow account save	Move	Yellow account data	30-day save plan	Drift accounts
27	Red account save	Move	Red account data	60-day save or graceful exit	Last-chance saves
28	Competitive displacement	Move	Competitor intel	Defense + comparison	Active evaluations
29	Champion replacement	Move	New stakeholder data	Warm re-intro	New admin / new buyer
30	End-of-quarter sprint	Move	7 at-risk accounts	Triage + forecast	Final week of quarter

A 30-day “expand 5 accounts” sprint

Here’s the exact 4-week plan I’d run if I had to deliver 5 expansion deals in 30 days using only these prompts.

Week 1 - Mine and measure.

Monday: run Prompt 1 (daily triage digest) on the full book.
Tuesday: pick the top 10 accounts with the highest expansion signals.
Wednesday: run Prompt 6 (health score) on each, then Prompt 8 (expansion propensity).
Friday: identify the 5 accounts that score highest on both. These are your sprint accounts.

Week 2 - Map the stakeholders.

Monday: run Prompt 11 (multi-threading map) and Prompt 3 (power-user detector) on each sprint account.
Wednesday: book 1 multi-threading intro per account.
Friday: for accounts with weak exec sponsors, run Prompt 12 (executive sponsor reactivation) and send.

Week 3 - QBR and message.

Monday: book 5 QBRs using Prompt 15 (joint roadmap session) framing.
Tuesday–Thursday: run Prompt 16 (QBR deck) and Prompt 17 (ROI calc) for each.
Friday: hold 2 QBRs. Send Prompts 20 (follow-up) within 4 hours.

Week 4 - Move and close.

Monday: run Prompt 22 (cross-sell engine) and Prompt 24 (multi-year renewal) on each.
Tuesday: send expansion emails to the 3 highest-scoring accounts.
Wednesday: hold remaining QBRs.
Thursday: run Prompt 25 (case study ask) on the 2 most likely yeses.
Friday: hold 30-min close calls. Send Prompt 21 (renewal forecast) to your CRO.

Expected outcome: 5 QBRs booked, 3 expansion conversations live, 1 closed-won, 2 in the pipeline for next month. The ChurnZero study I cited earlier (74% of revenue from existing customers) says the math is on your side - you just have to run the plays.

Common mistakes to avoid

A few traps I see account managers fall into with these prompts:

Pasting data you don’t understand. If you don’t know what an “API rate limit” means, don’t paste it. Garbage in, garbage out.
Skipping the verification step. ChatGPT hallucinates numbers. Always read the output and check it against the source. Every. Single. Time.
Prompting in the abstract. “Write an upsell email” produces nothing useful. “Write an upsell email for [Account] citing these 3 usage data points” produces a draft you can send.
Forgetting the human. The prompt writes the first draft. The CSM adds the judgment, the relationship context, the tone. Don’t let the model ghost-write the relationship.
Saving the prompt, not the play. A prompt is a tool. A play is a tool plus a workflow plus a CTA in Gainsight. Use the prompts to feed your plays, not replace them.
Running prompts without a data governance review. If your company has rules about pasting customer data into third-party AI tools, follow them. Anonymize where possible. Prompt in a privacy-safe way.
Forgetting that the customer’s data is the customer’s. Even if you can paste it, you should always ask: would the customer be okay with this? If not, summarize instead.

Final word

You don’t need another tool. You need to run the same 30 plays, in the same order, on the same cadence, every quarter. ChatGPT makes the writing fast. Your judgment makes the selling real.

The 2026 numbers don’t lie. Existing-customer revenue is the biggest game in B2B SaaS. Usage data is the only honest signal. Your existing customers are telling you what they want next - in logins, in API calls, in dormant seats, in NPS verbatims. These 30 prompts just turn that signal into action.

Pick 5 accounts. Run the 4-stage framework. Run the prompts. Close the deals.

I’ll see you in the QBR.

Reader disclosure & educational-purpose notice

This page is published by SuperFreshAI for general informational and educational purposes only. By reading it, you agree to the points below.

Editorial independence. All reviews, guides, and recommendations are written by our editorial team based on hands-on use. Some links on this site are affiliate links, and some articles are produced as partner content — both are always clearly labeled. Our editorial conclusions are never shaped by partners or affiliates.
Not professional advice. Nothing on this page constitutes legal, financial, medical, tax, or other professional advice. AI tools, pricing, and capabilities change quickly — always verify current information with the tool's official documentation before making a decision.
Educational purpose only. The content here is intended to help you learn about AI tools and workflows. It is not a guarantee of results, performance, fitness for a particular purpose, or suitability for your specific situation. Your results may vary.
No warranties. The site and its content are provided on an "as is" and "as available" basis. We make no warranties, express or implied, about accuracy, completeness, reliability, or availability. See our Terms and Privacy for the full legal terms.
Your responsibility. You are responsible for how you use the information on this page, including any decisions you make based on it. Always do your own research and consult a qualified professional when appropriate.
Affiliate & partner disclosure. When you click certain outbound links, we may earn a commission at no extra cost to you. When a piece of content is produced as partner content, it is labeled at the top of the page. See our Editorial Policy for the full standards we follow.

By continuing to read, you acknowledge that you have read and understood this notice.

30 ChatGPT prompts for account managers to upsell existing clients based on usage data

Quick answer: how do I use ChatGPT to upsell with usage data?

Why 70% of expansion revenue hides in usage data

The 4-stage usage-based upsell framework

The 2026 benchmarks your prompts need to beat

How to use the 30 prompts

SECTION 1: Usage-mining prompts (Prompts 1–5)

Prompt 1: Daily usage triage digest

Prompt 2: Feature adoption gap finder

Prompt 3: Power-user and champion detector

Prompt 4: API and rate-limit “about to break” alert

Prompt 5: Dormant-seat reactivation opportunity

SECTION 2: Health-score prompts (Prompts 6–10)

Prompt 6: Build a Customer Health Score from raw signals

Prompt 7: Early-warning churn radar

Prompt 8: Expansion propensity score

Prompt 9: NPS verbatim analyzer

Prompt 10: Health score downgrade postmortem

SECTION 3: Stakeholder expansion prompts (Prompts 11–15)

Prompt 11: Multi-threading map generator

Prompt 12: Executive sponsor reactivation

Prompt 13: Champion enablement kit

Prompt 14: Power-vs-power user mismatch (organizational redesign)

Prompt 15: Joint roadmap session invitation

SECTION 4: QBR / EBR prompts (Prompts 16–20)

Prompt 16: QBR deck builder (one prompt, full deck)

Prompt 17: ROI calculator for the EBR

Prompt 18: Executive Business Review (EBR) narrative arc

Prompt 19: Customer success Q&A prep (for tough QBRs)

Prompt 20: QBR follow-up email (within 24 hours)

SECTION 5: Renewal & cross-sell prompts (Prompts 21–25)

Prompt 21: Renewal forecast narrative

Prompt 22: Cross-sell engine (the “next product” suggestion)

Prompt 23: Price increase justification letter

Prompt 24: Multi-year renewal with expansion

Prompt 25: Customer reference and case study ask

SECTION 6: Save-the-customer prompts (Prompts 26–30)

Prompt 26: Save play for a Yellow account

Prompt 27: Save play for a Red account (last chance)

Prompt 28: Competitive displacement defense

Prompt 29: Save play when the champion has left

Prompt 30: The end-of-quarter save-everything sprint prompt

Comparison table: prompt category vs. stage vs. output

People Also Ask: 10 questions account managers ask about usage-based upsell with ChatGPT

1. What are the best ChatGPT prompts for account manager upsell using usage data?

2. How do I use usage data to upsell without sounding salesy?

3. What’s a good NRR benchmark for SaaS in 2026?

4. How do I find expansion opportunities in Mixpanel or Amplitude?

5. What’s the difference between QBR and EBR?

6. How do I write a usage-based upsell email?

7. How can ChatGPT help me prepare for a tough QBR?

8. What’s a customer health score in 2026?

9. How do I multi-thread an account that’s single-threaded?

10. How do I run a 30-day expansion sprint with these prompts?

A 30-day “expand 5 accounts” sprint

Common mistakes to avoid

Final word

SuperFresh AI

43 ChatGPT prompts for non-native English speakers to polish interview answers

41 ChatGPT prompts for SaaS founders in San Francisco to map local partnership opportunities

How to Detect AI-Generated Content

What Is the Best AI Tool for Writing?

AI Newsletter Writing Guide

Get practical AI insights in your inbox