Case Study

TQ Solutions: 91.2% of AI spend
identified as recoverable.

A TokenomicsIQ diagnostic on a live Future of Work consulting practice. 4 ranked inefficiencies. Efficiency score 0.9 out of 10. Here is what the report found and how TQ Solutions is acting on it.

Recoverable Spend

91.2%

Of current monthly AI outlay

Efficiency Score

0.9 / 10

Critical — immediate action flagged

Ranked Findings

Across 3 active workflows · 47-day period

Company

TQ Solutions

Industry

Future of Work & Workforce Research

Analysis Period

26 Mar – 12 May 2026 (47 days)

Report Generated

29 July 2026

Efficiency Score

Score 0.9: Critical

TQ Solutions' usage data scored 0.9 out of 10 on the TokenomicsIQ efficiency scale. A score in the Critical range means the majority of AI spend is attributable to structural patterns that can be fixed without changing the underlying workflows or their outputs. The lower the score, the higher the return on implementing recommendations.

Where TQ Solutions sits

Critical

Significant Waste

Moderate

Optimised

0 – 2.5

2.5 – 5.0

5.0 – 7.5

7.5 – 10

Score 0.9 Critical: 91.2% of spend is structurally recoverable

A Critical score does not mean the AI tools are wrong or the workflows are failing. It means the implementation patterns have not yet been optimised, which is normal at an early stage of AI adoption. The finding is that 91.2% of current spend can be recovered through 4 specific structural changes, without altering what any of these workflows produce.

Inefficiency Analysis

4 inefficiencies. Ranked by saving potential.

Each finding is ranked by the percentage of monthly spend recoverable through that fix. Difficulty reflects the estimated engineering effort to implement the change.

#	Pattern & Workflow	Saving Rate	Difficulty
1	System Prompt Bloat HR Service & Operating Model Planning workflow	47%	Low
2	Agent Loop Redundancy HR Service & Operating Model Planning workflow	100%	Medium
3	Prompt Cache Miss Talent Marketplace Research workflow	45%	Medium
4	Prompt Cache Miss Market Trends Visual workflow	56%	Medium

Finding 1 in detail: System Prompt Bloat

On 43 of 46 requests in the HR Service & Operating Model Planning workflow, prompt tokens exceeded 30% of total tokens per call. The average prompt share was 93%, meaning the system prompt alone was consuming nearly all the token budget on every request. A lean system prompt is typically 10 to 20% of tokens per call. Trimming accumulated instructions and injecting edge-case guidance conditionally is a 0.5 to 1 day fix that removes the waste without any measurable difference in output quality. This single change accounts for 52% of total identified savings.

Finding 2 in detail: Agent Loop Redundancy

19 API calls were made in quick succession (within 60 seconds) on the same workflow with near-identical content. This is the classic agent loop signature: retries, parallel sub-tasks that duplicate work, or polling via full API calls. Each redundant call is billed in full at 100% waste. A lightweight deduplication cache or prompt-caching guard resolves this in a 2 to 3 day implementation, with no change to workflow outputs.

Verdict

TQ Solutions' highest-impact action is to audit and compress the system prompt on the HR Service & Operating Model Planning workflow. This single change accounts for 52% of total identified savings and can typically be implemented in under a day.

“

As Founder and CEO, TokenomicsIQ provided a snapshot diagnostic on the efficiency of our AI usage. The report clearly identified waste in our workflow and, importantly, prioritised the biggest cost savings and how to action them. As we continue to build out our AI capabilities, having a grip on the AI costs without impacting our output becomes business critical. It is going to be much needed.

Gareth Flynn

Founder and CEO, TQ Solutions · Future of Work and Workforce Research

TQ Solutions: 91.2% of AI spendidentified as recoverable.

Score 0.9: Critical

4 inefficiencies. Ranked by saving potential.

Run the same check on your AI spend

TQ Solutions: 91.2% of AI spend
identified as recoverable.