Problem
Where they started
The package looked technically intact, but it had only 1 recommendation and a 6-question fallback set. That was not enough for a strong client handoff.
Back to case studies
Quality gate and calibration
IT services firm: catching a weak package before handoff
This package had clean files, but the story was not strong enough to send. New Reward caught the weak recommendation depth and fallback question set before the package could be mistaken for complete proof.
Work
What New Reward did
New Reward checked file quality, mapped competitors and content gaps, flagged the weak recommendation set, and created a staging calibration plan.
Result
What changed
The package was stopped from being oversold. The next work was clear: improve recommendation depth and validate scoring before production rollout.
Client context
Anonymized IT services client
- Industry
- Information technology and services
- Timeframe
- February and March 2026 validation planning
- Starting point
- The package had 0/100 measured AI visibility, a 6-question fallback benchmark set, and only 1 research recommendation.
Why it matters
Quality control protects trust. A weak package can hurt the client relationship even when the files are technically valid.
Proof boundary
This is a quality-gate proof point. It is not a positive results case study and should not be framed as client growth.
AI visibility baseline
Measured starting score at audit time.
Benchmark questions
Fallback question set; needs expansion.
Competitors mapped
Competitive set included in the package.
Content gaps
Gaps found for follow-up work.
Recommendations
Insufficient depth; caught before handoff.
Readiness asset groups
Schema, data, training, review, and site-root groups.
Bars show evidence coverage or measured baseline values. They are not revenue or ranking lift unless stated.
Work performed
- Checked required files, valid JSON, and clean client scoping.
- Mapped 8 competitors and 7 content gaps.
- Flagged recommendation depth as too thin for a final client package.
- Created a staging calibration plan for weighted scoring, retrieval metadata, and grounded retrieval.
Verified result
New Reward identified that the package was technically intact but not strong enough to send as a complete visibility story.
Source notes
Deep package audit
Internal package audit dated 2026-02-25
Staging calibration plan
Internal Solution Stream calibration plan dated 2026-03-05