An end-of-year seasonal conversion campaign required personalized outreach for 170+ accounts — impossible to do manually. I built a Claude Project + Google Services integration that turned one prompt into a complete, on-brand prep pack for each account: a personalized deck, an admin report, and a stakeholder email. Here is how it works.
Each spring, the customer success team needed to review the year with every account and set up the next cycle — conversion season and back-to-school success depended on it. The prep behind each call was substantial: pull the data, build a deck, write a report, draft the stakeholder email, keep brand and messaging standards consistent. Multiplied across 170+ accounts and a team of 4, the production burden became the constraint.
“Less time prepping means more time strategizing.”
Director of Partnerships · Team Feedback
Team members never start from a blank page. They work from a triage dashboard that shows their entire portfolio by status. Each card generates a ready-made prompt. Paste it in, and the right workflow runs automatically.
Beyond triage, the dashboard reads each district's insurance gap, the share of uninsured and out-of-network students, against a state-specific threshold. Districts that clear the bar are flagged as upsell candidates right in the view. That signal then shapes the build: the deck and report frame the coverage gap and point to sponsored plan options at renewal.
The hardest part of any reporting workflow is trustworthy data. The suite automatically synthesizes utilization data from Looker, clinical outcomes data, and qualitative data from Google Sheets CSAT surveys — then layers in email and Gong call context to customize messaging for each district. It pulls from live sources every time, normalizes messy district names, and falls back gracefully when a number is too thin to be honest about.
Every prep pack stitches together a portfolio of sources, automatically:
These are real, lightly edited pages from a generated prep pack. Notice the personalization, the consistent palette, and the forward-looking language. Nothing here was hand-formatted.
Note: Account names, district details, and team member names in sample outputs have been anonymized for this portfolio presentation.
A presentation that mirrors the actual call: review the year, plan the next season, complete next-cycle onboarding live. Every slide carries the account name, the owner's correct title, and tier-appropriate links — with no manual formatting required.
"Customer Success Manager," never "CSM." Titles are enforced from a single source.
The attendance slide renders only for states where it applies, then renumbers.
Toolkit and summer-kit links route by contract type, never named in the copy.
A clean 6-page PDF a district leader can read on their own: referrals, school-site activity, staff engagement, and clinical outcomes. It is the artifact that outlives the meeting.
The full-year funnel and referral mix, benchmarked against the Daybreak average.
Signed by the owner with their real title and contact.
The same build drafts the summer email to the shared Drive folder.
Built-in, system-level enforcement keeps every output aligned across the team.
Approved-language rules are applied to the content before render, keeping district-facing copy accurate and non-judgmental. A few enforced swaps:
Every build locks in the company brand system: approved palette, accent rules, typography standards, and clean logos — enforced automatically. Tier/contract type is never named in external-facing copy.
Outcomes are framed against the study and the portfolio, never promised as guaranteed. Thin assessment counts fall back to averages. Framing stays forward-looking.
Each session writes itself to a Notion log: task, date, CSM, district, outputs, and flags. No one maintains a tracker by hand.
I've used it for every meeting, and had it prep the EOY deck for every async meeting too. Minimal edits needed. Definitely a time saver.
It saved me countless hours putting these decks together, and it was easy to customize per district. People are loving the two pieces of collateral.
It helps with production and logistics, so we spend more time talking strategy than building decks.
I love how it takes the nuances and translates them into actionable insights, and adds the most current resources so you're not searching things up. Well-rounded analysis that highlights what's working. It demonstrates value.
Accounts are highly responsive to usage and outcomes numbers, and use them to tell the ROI story internally.
It starts a different kind of conversation and demonstrates real thought partnership, opening new opportunities.
It gives account leaders a clearer view of their organization's outcomes and needs than qualitative feedback alone.
The data points decision-makers directly to where added investment would have the most impact.
This started with one person solving one team's bottleneck before a critical seasonal conversion window. The pattern generalizes: any GTM team managing high-volume, repeating touchpoints can build internal tools that eliminate repetitive production work. Every hour freed from assembly is an hour pointed at strategy, relationship, and revenue.
Name the friction first, then automate exactly that. The system was designed around the bottleneck.
Pulling live sources automatically removed the main source of error. No re-exporting, no stale numbers.
Brand, voice, and clinical rules live in the build itself. That makes the output safe to trust at volume.
A single orchestrator running steps in a set sequence beats a pile of clever, fragile shortcuts.
Self-logging gives the team visibility into every run. That turns a personal tool into shared infrastructure.
The same skeleton of lookup, build, guardrails, and logging is ready to extend to renewals, content, and onboarding.
Using AI to automate prep work, to free up humans to focus on strategy and relationship building — the pieces that AI shouldn't touch.