AI Value Analysis

The list of ideas is rarely the hard part

One team wants a writing assistant. Another wants an agent to handle customer requests. Finance wants to know what the business will get back.

Those proposals cannot be compared on enthusiasm or hours saved alone. They need a defined workflow, an accountable owner, a quality threshold and the full cost of getting a usable result. I put those elements on one basis so leadership can make the trade offs openly.

What counts as value here

An accepted outcome is a finished piece of work that meets the agreed standard and is kept. For a support workflow, that could mean a request resolved correctly. A draft that needs rewriting has not yet reached that point.

I examine the benefit after checking and rework, alongside the cost of building and running the workflow. That includes integration, oversight and operating costs where relevant. If saved time creates capacity, we name the work that will use it. We do not quietly relabel it as a cash saving.

Where the business lacks a baseline, the model is labelled a scenario estimate. A promising estimate is a reason to design a test, rather than report a realised return.

Lite and Full answer different starting positions

Lite: you already have the ideas

Four weeks. ₹4,00,000, excluding applicable goods and services tax (GST).

I screen up to 40 business supplied use cases, score the top 10 to 15 and develop economics for the top three. You receive a fit assessment, economics and investment guardrails, a value and readiness map and a 60 minute sponsor readout.

The screen examines the route, essential gates, readiness to start and whether the work is worth doing. Cases can emerge as ready now, live but unguided, needing preparation or better left waiting. A high score does not override a missing owner, required data rights or an essential control.

Full: build the list with the functions

Six weeks. ₹9,50,000, excluding applicable GST.

Full adds up to six function workshops and a long list of up to 120 cases. Up to 25 are scored in detail, a shortlist of 10 to 15 is agreed with owners and the top five receive detailed economics. The six outputs include the use case register, fit assessment, economics and guardrails, value and readiness map, baseline packs for the top two workflows and an executive readout.

Screening, detailed scoring and pricing are separate levels of work. Every idea does not receive the same depth of analysis.

What you need to bring

The sponsor brings the decision and access to the people who own the workflows. Finance provides available cost and volume data. Case owners help define quality, failure and the current way of working.

I test assumptions with those people before the final readout. Where evidence is missing, the recommendation names the next test and its owner. Delivery is remote, with inputs and timing agreed at kickoff.

What the assessment case shows

For an anonymised technology enabled business, I separated direct AI spending from broader platforms, compared candidate uses on value and feasibility, and documented proceed, prepare or defer decisions. The work identified the cost, volume, cycle time and quality measures needed to assess economics.

The delivered output was a basis for investment decisions. It was not evidence that the recommendations had already produced savings. Read the case and its boundary.

Questions buyers ask

Do we need a readiness assessment first?

Not always. If you have credible readiness evidence, we can use it. If the wider business position is unclear, AI Readiness Assessment answers where AI is safe to scale. The two packages can be scoped together without repeating the evidence work.

Will you implement the shortlisted cases?

This package does not build tools, run pilots or measure them after launch. Those activities need their own scope. The analysis defines the evidence and conditions that would support the next investment.

Can you guarantee a return?

No. Forecast value depends on assumptions, adoption and operating results. Measured figures and scenarios stay separately labelled throughout the work.

Bring the ideas and the question behind them

Book a free 30 minute discovery conversation. We will establish whether the immediate need is screening an existing list, building one with the business or resolving readiness first.

I lead the work personally. Read about my technology and advisory experience.

Fees reflect the October 2026 package scope and apply to statements of work signed by 31 March 2027. Final scope, evidence dependencies and commercial terms are agreed in writing.

Clarity today. Momentum tomorrow.