Step 4: Tailor packaging to the solicitation

Most solicitations specify page limits and a structure (e.g., “Provide up to 3 references; include scope, dollar value, period, customer contact”). Your system should let you generate tailored write-ups by filtering your reference library by:

• Similar agency/mission
• Similar security posture
• Similar scale (users, sites, $)
• Similar technical pattern (cloud migration, SOC, data platform, etc.)

This is where “why” matters: you’re not just telling a story—you’re proving reduced performance risk.

Practical application: Templates, examples, and code to operationalize it

A strong past performance write-up template (proposal-ready)

Use this structure for each reference unless the RFP dictates otherwise:

(1) Contract Snapshot

• Customer/Office:
• Contract/TO:
• Period of Performance:
• Contract Type:
• Total Value / Your Value:
• Your Role (Prime/Sub):
• Customer Contact (permission status):

(2) Relevance to this Solicitation

• Scope similarity:
• Complexity similarity:
• Magnitude similarity:
• Compliance/security similarity:

(3) What We Did (Technical)

• Key tasks performed
• Tools/stack (only what matters)
• Interfaces/integrations

(4) Results (Quantified)

• Outcome metrics (with baselines)
• SLA attainment
• Delivery milestones

(5) Proof / Validation

• CPARS rating/narrative excerpt (if allowed)
• Acceptance evidence type (email, signed form)
• Awards/recognition (if applicable)

(6) Lessons Learned / Risk Reduction

• What we improved
• How it reduces risk for this new requirement

Example: “Before/after” results that read like engineering, not marketing

Weak: “Improved system performance significantly.”

Better:

• “Reduced P95 API latency from 900ms to 220ms by implementing caching and query optimization; verified via CloudWatch dashboards and monthly SLA reports.”
• “Achieved 99.95% uptime over 10 months; documented in monthly availability reports submitted to the COR.”

A simple reference library in YAML (portable, versionable)

Store references as structured files so you can search and reuse them. Example:

id: pp-2024-07-cloud-migration
customer:
  agency: "Federal Civilian Agency"
  office: "Office of Digital Services"
contract:
  vehicle: "IDIQ"
  task_order: "TO-1234"  # omit if sensitive
  type: "T&M"
period:
  start: "2023-01-15"
  end: "2024-01-14"
value:
  total: 8500000
  your_share: 3200000
role:
  prime_or_sub: "Sub"
  responsibilities:
    - "Led landing zone design and IaC"
    - "Implemented CI/CD with security gates"
relevance_tags:
  - "aws"
  - "cloud-migration"
  - "zero-trust"
  - "nist-800-53"
metrics:
  - name: "Uptime"
    value: "99.95%"
    evidence: "Monthly SLA reports"
  - name: "Deployment frequency"
    value: "+3x"
    evidence: "CI/CD pipeline logs"
proof:
  cpars: "Very Good"  # if applicable
  acceptance: "Signed deliverable receipt"
customer_contact:
  name: "Jane Doe"
  title: "COR"
  email: "[email protected]"
  permission_status: "Approved for contact"
notes:
  sanitization: "No Controlled Unclassified Information (CUI) included; tool names generalized"

Quick search script (Python) to find relevant references

import glob, yaml

def load_refs(path="references/*.yml"):
    refs = []
    for f in glob.glob(path):
        with open(f, "r", encoding="utf-8") as fh:
            refs.append(yaml.safe_load(fh))
    return refs

def score(ref, required_tags):
    tags = set(ref.get("relevance_tags", []))
    return len(tags.intersection(required_tags))

required = {"zero-trust", "nist-800-53", "cloud-migration"}
refs = load_refs()
ranked = sorted(refs, key=lambda r: score(r, required), reverse=True)

for r in ranked[:5]:
    print(r["id"], score(r, required), r["customer"]["agency"], r["value"]["total"])

This is intentionally simple: the point is to make relevance retrieval repeatable.

Diagram (described): “Evidence-to-Proposal Assembly”

A diagram showing three boxes left-to-right: 1) “Evidence Store” (SLA PDFs, acceptance emails, dashboards) → 2) “Reference Record” (YAML/CRM fields) → 3) “Proposal Output” (tailored 1-page PP write-up). Arrows indicate automation opportunities (search, templating, redaction).

Best practices: Patterns that consistently improve evaluation outcomes

1) Lead with relevance, not chronology

Evaluators often scan for similarity first. Put a short “Relevance to this Solicitation” section near the top, using the RFP’s own language.

2) Quantify outcomes with baselines and measurement method

“Improved by 30%” is incomplete without:

• Improved what?
• From what baseline?
• Measured how and when?

Add one line of measurement provenance: “Measured via monthly SLA reports submitted to COR.”

3) Be explicit about your role (especially as a subcontractor)

If you were a sub, avoid vague phrasing like “supported.” Use crisp responsibility statements:

• “Owned design and implementation of X”
• “Provided Tier 2/3 operations for Y”
• “Authored System Security Plan (SSP) sections for Z controls”

4) Manage customer contacts like a professional program

Create a cadence:

• Ask permission to use them as a reference
• Confirm contact info quarterly
• Give them a heads-up when an RFP is in play

This reduces the risk of “unreachable reference” during evaluations.

5) Sanitize and classify your artifacts

Never drop CUI, sensitive architecture diagrams, or proprietary partner info into a generic library. Maintain:

• A sanitized version for proposals
• A restricted version for internal use

If you work with controlled data, align your handling to NIST guidance:

• NIST SP 800-171 overview: https://csrc.nist.gov/publications/detail/sp/800-171/rev-2/final

6) Treat CPARS as a living input, not a surprise

If your contracts receive CPARS evaluations, establish internal review and response procedures. CPARS narratives are often more valuable than the numeric rating because they explain context and corrective actions.

7) Build “relevance coverage” across your portfolio

A mature approach is to maintain a heatmap of which references cover which common federal needs (e.g., cloud, cybersecurity, service desk, data engineering, zero trust). When you see gaps, you can pursue targeted work (even smaller tasks) to fill them.

Limitations: Where past performance documentation can still fail

1) You can’t document your way out of irrelevance. If your portfolio doesn’t match scope/scale, documentation helps, but won’t fully close the gap.

2) CPARS availability varies. Not all work yields CPARS, and some CPARS narratives are thin. Don’t rely exclusively on CPARS—build your own evidence.

3) Customer contact risk is real. People rotate. Offices reorganize. Mitigate by keeping multiple points of contact when allowed (e.g., COR + PM).

4) Over-sharing can backfire. If you include sensitive details, you may trigger compliance issues or force redactions that weaken clarity.

5) Metrics can be misleading without context. A “99.99% uptime” claim means little if the service window, exclusions, and measurement method aren’t stated.

Further reading: Authoritative resources and practical guides

• FAR Part 15 (source selection & evaluation): https://www.acquisition.gov/far/part-15
• FAR 42.15 (past performance information): https://www.acquisition.gov/far/part-42.15
• CPARS official site: https://www.cpars.gov/
• GAO Bid Protest decisions (search past performance cases): https://www.gao.gov/legal/bid-protests
• NIST SP 800-171 (protecting CUI): https://csrc.nist.gov/publications/detail/sp/800-171/rev-2/final
• NIST Risk Management Framework (RMF) overview: https://csrc.nist.gov/projects/risk-management

Related Reading

• CUI-Safe CRM: The Complete Guide for Defense Contractors

Conclusion: Turn past performance into a compounding asset

The teams that consistently win federal work don’t “write past performance” under deadline—they operate a past performance system. Capture evidence at milestones, normalize it into reference records, and package it in a way that makes relevance and quality obvious to evaluators.

Actionable next steps: 1) Create a standard reference record schema (start with the YAML fields above). 2) Implement a milestone closeout checklist to capture acceptance + metrics. 3) Build 3–6 proposal-ready reference sheets mapped to your most common opportunity types. 4) Set a quarterly reference contact verification cadence.

If you want to accelerate this, cabrillo_club can help you design a past performance library, sanitization workflow, and proposal-ready templates tailored to your federal pipeline.