Multi-Step College Advising Workflow: From Assessment to Final Letters

What This Builds

A structured, repeatable college advising workflow that uses AI at each stage, from analyzing a student's college list for fit, to drafting personalized letters, to generating college-specific talking points for counseling meetings, replacing hours of manual research with a 20-30 minute AI-assisted session per student that produces better, more personalized outcomes.

Prerequisites

• ChatGPT Plus account ($20/month)
• Student brag sheets or counseling notes for your seniors
• Access to Naviance or your school's college planning tool for student college lists
• Basic familiarity with ChatGPT file uploads (Level 3)

The Concept

Think of this as building an assembly line for college advising. Each station in the line does one thing well: the first station analyzes a student's college list for fit and balance, the second station generates college-specific talking points for your meeting, the third station drafts the recommendation letter, and the fourth station creates a student-facing summary they can use during applications. You run each student through the same stations, producing consistent quality at a pace that wasn't possible when every step was manual.

Build It Step by Step

Part 1: Set up your ChatGPT Plus college advising conversation

For each student, start a fresh ChatGPT conversation. Your opening message establishes the workflow context:

I'm a school counselor helping a senior plan their college applications. I'm going to share their profile and college list, and I need your help at several stages: (1) analyzing college list balance and fit, (2) drafting a recommendation letter, and (3) preparing talking points for our advising meeting. Keep all responses specific to this student's data. Do not use their real name. Refer to them as "the student."

This sets up the workflow before you share any student information.

Part 2: Station 1. College List Analysis

Upload or paste the student's college list (15-20 schools) and their academic profile (GPA, test scores if available, major of interest). Ask for a fit analysis:

What to type:

Here is the student's college list and academic profile:
GPA: [X], class rank: [X or "not ranked"], intended major: [X]
College list: [paste list]

Please analyze this list for:
1. Reach/Match/Safety balance (roughly what % of each?)
2. Any obvious gaps (all reaches? No matches?)
3. 3 schools on the list worth discussing in our advising meeting: either notably good fits I should highlight, or concerning fits I should address
4. Any schools where the major they want is unusually competitive (list-specific context)

What you should see: A structured analysis with honest fit assessment and specific observations about 3 schools. This takes you from "reviewing the list from memory" to "informed conversation partner" in 3 minutes.

Part 3: Station 2. Advising Meeting Talking Points

Before your meeting with the student, generate a focused set of conversation starters based on the college list analysis:

What to type:

Based on the list analysis, generate 4-5 counselor talking points for my meeting with this student. Each should be a conversation opener, not a lecture. Focus on: [pick 2-3 from: college list balance / major-specific admissions / essay strategy / financial aid considerations / timeline and deadlines]. Keep them concise. I'll be using these to guide conversation, not read from.

What you should see: 4-5 bulleted conversation starters that address the specific issues surfaced by the list analysis. Your meeting will be more focused and productive because you walked in knowing what to discuss.

Part 4: Station 3. Recommendation Letter Draft

After the meeting, with the student's brag sheet in hand:

What to type:

Now draft a 400-word recommendation letter for this student. They're applying to [top choice school or program type]. Use the profile information we've discussed. Flag 2-3 places where I should add a personal observation that only I could write.

Because you've already discussed the student's profile in stations 1 and 2, the recommendation letter draft builds on established context. This produces better letters than asking for a draft cold.

Part 5: Station 4. Student-Facing College Research Summary

For students who need help understanding why specific colleges are a good fit for them, generate a brief, student-readable summary for each college:

What to type:

For each of the 3 schools we flagged as worth discussing, write a 3-4 sentence summary explaining why this school specifically might be a good fit for this student, based on their profile and interests. Write in second person addressed to the student. Tone: encouraging and honest, not sales-y.

Print these summaries and give them to the student before they write their "Why This School" essays. Students often struggle to articulate fit in their own words; seeing it laid out helps.

Real Example: A Full Session Walkthrough

Student: 12th grader, GPA 3.4, interested in environmental science, college list of 12 schools, first-generation college student.

What you do before the meeting (20 minutes):

1. Open ChatGPT Plus → paste opening message
2. Upload or type the student's profile and college list
3. Station 1: Ask for list balance analysis → get back: "7 reaches, 3 matches, 2 safeties. This list is top-heavy. Two environmental science programs on the list are particularly competitive (Bowdoin, UVM) and warrant discussion about backup plans."
4. Station 2: Request talking points → get: "Ask why she hasn't included any of the large state university environmental programs. They often have strong research opportunities and better scholarship money for her profile. Discuss financial aid strategy for first-gen students."

During the 30-minute meeting:

• You walk in with specific, informed questions
• Student feels heard because you've clearly thought about her specific situation
• Meeting covers what actually matters, not generic college advice

After the meeting (20 minutes): 5. Station 3: Draft recommendation letter with brag sheet → edit → finalize in 25 minutes instead of 90 6. Station 4: Generate 3 college summaries for her strongest matches → print for student to use in "Why This School" essays

Total AI-assisted time: 40 minutes for a complete, personalized college advising package per student. Without this workflow: 3-4 hours.

What to Do When It Breaks

• ChatGPT confuses two students' data → Always start a fresh conversation for each student; don't run multiple students through the same long conversation
• College list analysis is inaccurate for a specific school → Ask ChatGPT to clarify its reasoning; supplement with your own knowledge of schools in your state and region where AI data may be less reliable
• Recommendation letter misses the student's voice → Your editing in Station 3 is where you add the personal layer. The AI draft handles structure; you add soul
• Student summaries feel generic → Ask ChatGPT to "be more specific about how this school's [specific program/culture/resource] matches what this student said she wants in our meeting"; specificity requires specific input

Variations

• Simpler version: Use just Station 3 (recommendation letter) and skip the rest. This alone saves 60+ minutes per student during letter season
• Extended version: Add a Station 5 for each student's Common App personal statement. Paste a rough draft and ask for structural feedback and 3 suggestions for making it more specific and vivid

What to Do Next

• This week: Run one current senior through the full 4-station workflow and evaluate whether the quality and time savings justify the approach
• This month: Refine which stations produce the most value for your specific student population and school context
• Advanced: Combine this workflow with the Claude Project system (Level 4, Guide 1) to pre-load school context so you don't need to re-establish it each conversation

Advanced guide for school counselor professionals. ChatGPT's knowledge of specific colleges may be inconsistent. Verify any college-specific claims against the school's actual current website.