Stonefield Intelligence™

Complexity in.
Clarity out.

Built by management consultants, for management consultants. Footprint, location and workforce advisory — in plain English, with the sourcing shown, not another chatbot guessing in the dark.

Unstructured signal A defensible answer

Not a mockup. This is what the screen actually shows.

Real interface, running on labelled demonstration data — every screen tells you exactly what it is and isn't.

Stonefield Location Intelligence screen
SLI — set your own priorities, watch the ranking move
Stonefield Company Intelligence screen
SCI — who's under pressure, and how sure we are
Stonefield Executive Intelligence screen
SEI — the eight questions an exec actually asks

Everyone knows AI matters. Almost nobody knows how to move without breaking what already works.

Most firms are stuck between two bad options: ignore AI and fall behind, or restructure the model that pays the bills to chase it.

34%
of AI spend duplicates tools the firm already pays for — nobody's tracking why
HelpNetSecurity, shadow AI spend research
94%
of enterprises show no measurable earnings impact, despite record AI spend
McKinsey 2026 survey, 1,719 executives
Almost all of that spend is aimed inward — making the back office faster. Very little of it is aimed at winning the next client.
And the real fear isn't AI itself — it's that fixing this means gutting the billable-analyst model that funds the firm.

In plain English: you don't have to break your business model to use AI properly. You have to aim it at the right target — and give your analysts a tool, not a replacement.

Everyone's poking at ChatGPT. We built the playbook.

Same method, every time — sourced, labelled, checkable. Not whoever's turn it is to prompt the chatbot this week.

In Plain English™

The philosophy underneath everything we build: if a client can't understand the answer in one read, it's not finished yet. No jargon, no black-box maths — just the logic, said plainly.

Behind the Curtain

Our house rule for AI: every step has a matching manual method sitting behind it. If you ever want to know how a human would've done it by hand, we can show you — and prove both routes land on the same answer.

The market today
  • ×Everyone prompts a chatbot their own way, every engagement
  • ×Real data and rough guesses, mixed together, unlabelled
  • ×The answer lives in one analyst's head, not a method anyone else can run
  • ×Source in German or Dutch? Better find someone who reads it, or pay a translator, before you even start
  • ×A signal turns up and nobody's sure whose deal it is
The Stonefield approach
  • One method, every tool, every analyst, every time
  • Real data labelled as real. Estimates labelled as estimates. Always
  • Every AI step maps to how a human would do it by hand — you can check our work
  • Reads 10+ European languages directly, whatever the source throws at it — no translator, no new hire
  • Every signal lands on the right desk automatically

Behind the Curtain: Director's Cut

Every figure on this site has a traceable path back to its source, its transformation, and the rule that produced it — not just a claim, an actual generated trace. Any number, on demand.

We didn't pre-build a flowchart for every figure across all six tools on the off-chance someone would ask. That's exactly the kind of speculative AI spend we built Stonefield to argue against. Instead, the logic lives as a structured recipe, and the trace is generated live, only when someone actually asks the question — a technical reviewer, a data scientist, you.

Three traces are drawn on the Director's Cut page — stills, so you can see what one looks like. The live version, where any figure in any tool draws itself the moment you ask, runs inside our own tooling and can't sit on a web page. We'd rather show you than claim it: see the stills, then book a session and pick any number. The logic framework does the thinking — 250+ engagements' worth. AI just draws the picture.

In plain English: we're transparent. Everything in plain English, nothing off the table, nothing hidden. Stills on the site; the live draw when we meet.

SCENE 01
Why does this score come out the way it does?
▶ See the trace
SCENE 02
Why does this get a pressure flag?
SCENE 03
Why "cannot tell you" instead of a guess?

We didn't need an outside example. We used ourselves.

The clearest proof of "generate on demand, not just in case" isn't a client story. It's how we built the thing you just saw.

01

What did the user actually need?

Not a library of pre-drawn flowcharts. Just the ability to get one, the moment they ask a specific question. That's the whole north star — the output, on demand, nothing more.

02

So why render everything in advance?

We didn't. Building a flowchart for every figure across six tools, whether or not anyone would ever look, is exactly the speculative spend this whole site argues against. So we asked what the smallest real build actually was.

03

One master file, written once

The logic behind each traceable number lives in a single spreadsheet — source, transformation, rule, output, one row per figure. No code. Anyone on the team can read it, extend it, correct it.

04

AI does only the last step

When someone actually asks, AI reads that one row and draws the trace — live inside our tooling, in about a minute. It doesn't invent the logic. It executes ours — the same distinction we make about every tool in the suite. The stills on this site were drawn the same way, once.

What it saved: no speculative coding, no pre-rendered library nobody asked for, no engineering time spent on the 90% of figures that might never come up. Just one reference file, and a generation step that runs exactly once per real question.

In plain English: we're not asking you to take our word for how we work. We just showed you, on our own tool, using our own logic, in front of you.

What this is actually worth to your firm

Not a general-purpose AI story. These are the concrete reasons a management consulting firm licenses this.

Deal generation

Pre-trigger signals surface real opportunities before a competitor's scoping call does — not after the RFP is already out.

Cross-line business development

The eight-line framework tells you which practice owns each signal automatically, so BD time goes into the pitch, not the routing.

Analyst augmentation, not replacement

Your existing team works faster and covers more ground. Nobody's billable model gets restructured to get there.

Fast onboarding

One alignment session calibrates the framework to your service lines. No months-long internal build, no new methodology to design.

Multilingual by default

Reads source material in 10+ European languages directly — no translation step slowing down the first pass.

Workforce-of-the-future conversations

Starts the discussion with clients on skills, not job titles — exactly the conversation boards are now asking for.

One real engagement, not a projection
6 months → 1 afternoon

Reaching a real skills starting point on one engagement took two senior consultants six months, full-time, on-site — 233 distinct skills found. That starting point is now an afternoon's work, not the interviews and judgement calls around it, which stay exactly where they belong: with your consultants.

What "always on" actually costs to staff
4.2 FTE, per role, per week

Basic staffing maths: keeping just one function on call 24 hours a day, 7 days a week takes a minimum of 4.2 full-time people, before sick days or holidays. The suite is on call, across every tool, with no rota required.

Six tools. One engine.

Use one, use all six. From the first footprint signal to the answer your exec team can actually act on. Click any tool to see what it actually answers.

Key questions it answers: Which of our target accounts is under real pressure right now, and how confident should we be? Is this a footprint story worth raising with the client this quarter?

Every score is sourced and flagged with a confidence range, not presented as a bare number — so you know exactly how much weight to put behind it in the room.

Key questions it answers: Where should this role actually sit — and what does that decision cost, in real labour and regulatory terms, not a rule of thumb? How does one capital city really compare to another once you strip out the guesswork?

Weightings are yours to move — the tool shows the ranking shift live, so the client sees their own priorities driving the answer, not a black box.

Key questions it answers: Which roles in this business are genuinely AI-augmentable, and which aren't? If we planned around skills instead of job titles, where would we actually need to invest first? What does the workforce of the future look like for this specific client, not the market in general?

Coverage runs across 10+ European languages — screening skills and occupation data at a scale we don't publish the source of, since that mapping is proprietary. What we will say: [jobs and skills coverage figures — confirming exact numbers].

Key questions it answers: Has anything material changed in this client's workforce since we last looked — restructuring, hiring freezes, union activity? Do we need to be back in the room before they call us?

Runs as a standing monitor, not a one-off report, so a live signal doesn't sit unnoticed for a quarter.

Key questions it answers: Of everything the suite has flagged, what actually needs an executive decision this week? Where can we honestly say "we don't know yet" instead of guessing to fill a slide?

Every answer is labelled answered, partial, or "cannot tell you" — the gap is named, never quietly estimated around.

Key questions it answers: Where does everything the suite has produced for this client actually live, so the next analyst on the account isn't starting from zero?

One record per engagement, shared across the tools — not six separate exports living in six separate inboxes.

The design bar is unreasonably strict, on purpose.

The logic and frameworks behind each tool don't ship because they seem right. They ship because they survived people whose job is to find the hole in them.

150+
user testing sessions and counting
4
sectors sense-checked at C-suite level: asset management, private equity, real estate advisory, transaction advisory
every input challenged by external data scientists and industry specialists before it ships
DO
250+
location strategy programmes
40+
countries
16
years in consulting
8
sectors

Built by the person who did the work 250 times over.

David Osborne — Founder & CTO, Stonefield Advisory

Sixteen years in international consulting, thirteen of them in footprint optimisation. Two hundred and fifty-plus location strategy programmes across 40+ countries and 8 sectors, including co-leading a 50-person European hub at a major human capital advisory firm. Stonefield is that methodology, turned into six tools — not a generic AI wrapper guessing at what consultants need.

For strategy consulting firms specifically, that means:

  • Deal generation — pre-trigger signals surface opportunities before a competitor's scoping call does
  • Cross-line business development — the eight-line framework tells you which practice owns each signal, automatically
  • A real methodology underneath — not a chatbot improvising, but 250+ engagements' worth of logic, encoded and checkable
  • Your analysts, not instead of them — it gets your point of view into the room faster, it doesn't replace the person forming it

Enterprise-plus capability. Boutique speed.

The big platforms have depth but move slowly. Generic AI tools move fast but have no methodology underneath. Stonefield is built to sit in neither camp — enterprise-grade rigor, 250+ engagements deep, delivered with the speed and hands-on attention only a boutique can give.

Stonefield
Typical AI tool
Who built it
An industry practitioner who's run and sold 250+ of these engagements himself
× A technologist or software team with no consulting delivery track record
Can they deliver the work
Designed from real client engagements, by someone who knows the actual deliverable
× Often nobody behind the tool has actually run the engagement
Compliance design
Built around EU AI Act principles from day one — sourced, labelled, human-in-the-loop, auditable
× Compliance bolted on later, if it's addressed at all
TRY IT YOURSELF

Copy the block below into ChatGPT, Gemini, or Claude — it's the whole pitch, pre-loaded, so the AI has something real to judge instead of guessing.

In plain English: we're not asking you to trust a tagline. Paste our own pitch into an AI and make it argue with us.

Evidence, not enthusiasm

Real research on AI and consulting work — run against what your own numbers actually look like.

25%
faster, with AI vs. without
Harvard Business School × BCG field study
30–40%
more efficient on repeatable work — junior staff
BCG 750-person controlled experiment
-23%
worse, when AI runs loose on complex work with no review
Same BCG experiment
~97%
cheaper reading the source directly instead of translating first
Smartling / Lokalise 2026 data

In plain English: AI genuinely helps on the repeatable stuff, and genuinely hurts when nobody's checking the complex stuff. That's the whole reason Stonefield has a human in the loop, not just a chatbot.

Deals move at the speed of the signal

It's not just hours saved. It's showing up before anyone else does.

1

Pre-trigger signal detection

Restructuring, lease events, workforce shifts — caught from official sources before it's public. Not discovered in a scoping call.

2

Eight-line routing

Every signal already knows which service line it belongs to. No meeting spent working out who owns it.

3

Alignment session, not a cold start

One workshop calibrates the framework to your own service lines. No months-long build, no methodology from scratch.

4

Fewer scanning meetings

Meetings stop being "let's scan the account list" and start being "here's the shortlist."

In plain English: you find out about the opportunity before your competitor does, and you already know which partner should be in the room.

Watch the suite, tool by tool

Quick walkthroughs on demo data — nothing client-specific, ever. Each clip says exactly what kind of data it's showing.

All six tools, fast · want to see more? get in touch
SCI · Company Intelligence
SLI · Location Intelligence
SPI · People Intelligence
SHI · Human Intelligence

Built for consultants to actually use — inside your own client work.

Not a black box. Not a six-month build. One session and you're live.

Get in touch

Also — always up for a coffee

If you're into AI, structured thinking, or just figuring out how the two turn into better business decisions, get in touch. Coffee, a workout, a walk, whatever — happy to talk shop or nothing in particular.

Say hello →