Dossier annexes

The material that justifies the recommendation: working profile, questions, sources, interviews, prioritisation, assumptions and build plan.

Disclaimer

Demonstrator — synthetic data — Norevia Ingénierie is a fictional company and nothing shown here represents a real client outcome.

Annex 01

Norevia working profile

Our understanding of the business, model, targets and tooling. Every item carries its status; undisclosed fields show “Not disclosed”.

Identity

Identity
Field	Value	Status	Source
Name	Norevia Ingénierie	Synthetic data	Fictional company created for the demonstration
Country	France	Synthetic data	—
Headcount	46 employees	Synthetic data	—
Activity	Engineering consultancy in structure, MEP and energy, working on B2B projects.	Synthetic data	—
Engagement types represented	Studies · audits · design · technical analysis · project support, across buildings, logistics platforms, industrial sites and office assets.	Synthetic data	—
Exact revenue split by service line	Not established within the Diagnostic scope	To confirm	—

Business model

B2B engineering services sold primarily per project.Synthetic data

Projects with an indicative value of €20,000 to €250,000
Value driven by scope, complexity and engagement type
Revenue generated by studies, audits and engineering engagements
Heavy dependence on technical expertise and team capacity

Business model
Field	Value	Status	Source
Annual revenue	Not disclosed within the Diagnostic scope	Not disclosed	—
Contribution margin	Not shared	To confirm	—
Revenue split by service line	Not shared	To confirm	—
Client repeat rate	Calculation not run	To measure	—

Do not conflate the value of an opportunity, annual revenue, margin, and the potential value of the system. The Diagnostic only estimates the last one, and only as an order of magnitude.

Positioning

A multidisciplinary engineering consultancy able to bring structure, MEP and energy expertise to complex B2B projects.Synthetic data

Norevia works on bids that require several disciplines
Its projects produce reusable documents and references
Preparing a new bid depends heavily on the ability to recover context and comparable projects

Service lines

The real split of revenue and margin by service line must be confirmed.

Working commercial targets

B2B companies and principals
Logistics platforms
Industrial sites
Office buildings
Construction, extension, audit or retrofit projects
Stakeholders who need technical expertise and a documented decision

This target is inferred from the demonstrator's synthetic opportunities. It must be checked against Norevia's real portfolio.

Working segment

French Ops-Intensive SMBsHypothesis

Project-driven activity
Data spread across several sources
Heavy dependence on experts
Partly manual preparation process
Potential value tied to freed capacity and decision quality

Differentiation

A combination of structure, MEP and energy expertise, enriched by experience accumulated on comparable projects.Hypothesis

Multidisciplinary technical expertise
Ability to work on complex projects
Project references and archives
Understanding of industrial, logistics and office contexts
Internal know-how distributed across several people

The Diagnostic does not yet measure whether this differentiation is clearly perceived by prospects or clients.

How the firm works

A team of 46 people
Bids prepared by both commercial and technical profiles
Expertise distributed across several people
Project knowledge sometimes held by a few individuals
Human validation required before a recommendation becomes an action
The sales owner remains accountable for the opportunity
Technical experts confirm that the matches are relevant

Watch point: a recommendation without an identified owner produces no action.

Tech stack and data maturity

Tech stack and data maturity
Field	Value	Status	Source
CRM	An existing CRM, used daily	Observed fact	Sales lead interview
Name of the tool	Not disclosed	Not disclosed	—
Project archive	186 historical projects and 420 documents	Observed fact	Archive analysis
Shared identifiers CRM ↔ archive	Insufficiently linked within the scope analysed	Observed fact	Cross-analysis of both sources
API and access	Options not verified	To confirm	—
Current automation	Not examined in this first loop	Out of scope	—

Service lines — detail of the three families

Structure

Structural studies and analysis
Construction, extension or retrofit projects
Technical evidence from past projects

MEP

Design or analysis of mechanical and fluid systems
Projects requiring a coordinated technical approach
References from comparable projects

Energy

Energy audits
Analysis of industrial or office sites
Retrofit and energy performance improvement

Annex 02

Diagnostic questions & answers

The answers are synthesised for the demonstrator, never quotations. Each question carries its consequence for the Diagnostic.

Diagnostic questions
ID	Category	Question	Synthesised answer	Status	Consequence
Q-01	Value & measurement	How long does it take today to recover the context of an opportunity?	Several hours depending on the case, with no formal measurement. The duration varies widely with who handles the opportunity and how old the account is.	To measure	The business case rests on a 3.5-hour assumption that has never been measured. This is the first thing to time before any Build.
Q-02	Preparation process	How do you find a comparable project today?	From memory, then by browsing folders. Rarely by query. Comparable projects exist, but only a handful of project leads actually know about them.	Interview synthesis	The link between a CRM account and its historical projects has to be rebuilt: that is condition 2 of the decision.
Q-03	Decision & governance	Who decides whether to bid, and on what criteria?	Sales management decides, drawing on a project lead's opinion. The criteria are not written down.	To confirm	Conditions 3 and 4 of the decision: formally name the business owner and write the qualification rules.
Q-04	Differentiation & positioning	What do your clients specifically choose you for, rather than another engineering firm?	The combination of structure, MEP and energy on a single project, plus experience of comparable sites. This perception has not been checked with clients.	Hypothesis	The differentiation shown in the working profile remains a hypothesis. It is deliberately not used as an argument in the business case.
Q-05	Preparation process	What happens when the person who knows the comparable project isn't available?	Preparation is postponed, or the bid is handled without the reference. In the second case, the pricing is rebuilt from scratch.	Interview synthesis	This justifies the priority loop: it is exactly the “starting from scratch” the Diagnostic sets out to remove.
Q-06	Activity & model	What share of your activity comes from clients you already serve?	Not established within the Diagnostic scope. The CRM holds the necessary history, but no calculation was requested.	To measure	A high share would strengthen the loop's value. Worth computing before the Build — it is a simple CRM query.
Q-07	Activity & model	Which engagement type is most profitable for you?	Not disclosed. Margins by engagement type were not shared within the Diagnostic scope.	Not disclosed	The action sheet therefore does not prioritise by profitability, only by available context. This is an accepted limitation of the first loop.
Q-08	Data & tooling	Which CRM do you use, and since when?	A CRM is in place and used daily by the sales team. The name of the tool was not disclosed within the Diagnostic scope.	Not disclosed	CRM integration stays conditional in the build plan until the tool is identified.
Q-09	Data & tooling	Do your archived projects follow a common naming convention?	No. Naming vocabulary varies across periods and teams, and the engagement type is missing on part of the projects.	Observed fact	Condition 1 of the decision: standardise six project description fields before building.
Q-10	Data & tooling	Can a task be created in your CRM from an external system?	Access and integration options were not verified. No integration is delivered as part of the Diagnostic.	To confirm	If writing back proved impossible, the loop would stop at the documented decision. The value would remain real but reduced.
Q-11	Preparation process	At what point do you commit expert time?	Case by case, with no formal threshold. The decision mostly depends on whether the expert is available at that moment.	Interview synthesis	The action sheet shows the level of context recovered to support that judgement, without making it for the team.
Q-12	Preparation process	Who writes a proposal, and starting from what?	The sales owner writes it, with a project lead supporting the technical part. The starting point is usually an earlier proposal found by hand.	Interview synthesis	Action sheet evidence therefore includes commercial proposals, not only engagement reports.
Q-13	Value & measurement	How many opportunities do you actually handle per quarter?	Not established. The CRM holds 48 opportunities over twelve months; extrapolating to 24 per quarter is a working assumption.	To measure	If the real volume were half that, no scenario would reach the decision threshold. This is the most sensitive assumption in the business case.
Q-14	Decision & governance	Who can validate that a recommendation becomes an action?	The sales owner of the opportunity. That responsibility is exercised in practice but not formalised as a role.	Interview synthesis	Condition 3 of the decision: formally name the business owner before the Build.
Q-15	Decision & governance	What do you do today with a lost bid?	No usable trace was identified within the scope analysed. Loss reasons are not systematically captured in the CRM.	To confirm	Out of scope for the first loop. To revisit in a second loop, once the first one has been measured.

Annex 03

Source detail

Norevia CRM

Good

Sales database · 2019 → 2026 · 312 accounts · 1,240 contacts · 48 opportunities

Possible use

Provide the commercial context of an account
Act as the trigger: an opportunity created or updated
Identify the owner who will have to validate the action

Limits

Incomplete activity history on older accounts
No explicit link to the projects already delivered for the account

Norevia project archive

Medium

Documents and engagement references · 2019 → 2026 · 186 projects · 420 documents

Possible use

Find projects comparable to an opportunity
Link reusable evidence to a recommendation
Identify the project lead to involve

Limits

Naming vocabulary varies from one project to the next
Filing is irregular across periods and teams
Engagement type is missing on part of the projects

Usability matrix

Source usability matrix
Source	Information	Quality	Use	Blocker
CRM	Account, contacts, sales history	Good	Account context and loop trigger	History incomplete on accounts predating 2021
CRM	Opportunity, amount, stage, owner	Good	Qualification and designation of the validator	None
CRM	Link to projects delivered for the account	Low	Matching account ↔ project history	Account ↔ project link not reliable in the sample analysed
Project archive	Name, client, year, lead	Good	Identifying the comparable project	Inconsistent naming
Project archive	Engagement type and sector	Medium	Matching by engagement similarity	Type missing on roughly a third of projects
Project archive	Documents: proposals, reports, summaries	Medium	Evidence linked to the recommendation	Irregular filing, duplicate versions

What blocks today

Project names do not all follow the same vocabulary.
Some projects have no standardised engagement type.
Opportunities are not linked to the account's historical projects.
Critical information sometimes lives in an expert's memory.

Annex 04

Interviews

The findings below are synthesised restatements, not quotations. No real person is represented.

Executive management

D3 — 5 August 2026 · 60 minutes — Understand the value expected and the level of urgency.

Go/no-go decisions are sometimes taken without retrieving the relevant references.
The topic is seen as a priority, but no budget is committed before the readout.
Management expects a recommendation that clearly separates the certain from the hypothetical.

Sales lead

D3 — 5 August 2026 · 45 minutes — Understand how an incoming opportunity is actually handled.

Gathering context on an opportunity can take several hours.
Qualification depends on a project lead being available to confirm feasibility.
A recommendation with no identified owner leads to no action at all.

Head of projects

D4 — 6 August 2026 · 45 minutes — Understand how similar projects are found today.

Comparable projects exist, but only a few people know about them.
Searching is done from memory, then by browsing folders — rarely by query.
Reusable evidence is identified late, often after a proposal has been sent.

Cross-cutting findings

The problem is not missing data: both sources contain what matters.
The problem is the absence of an explicit link between the sources.
Users need linked evidence, not an answer stated without justification.
A recommendation without an owner produces no action.
Human validation is what makes the team trust the result.

Annex 05

Full prioritisation grid

Prioritisation of value possibilities
Priority	Possibility	Value	Data	Blocker	Owner	Metric
1	Prepare a bid without starting from scratch	High	CRM + project archive	Account ↔ project link to rebuild	Sales management	Time to a documented decision
1	Surface the missing information	High	CRM + project archive	None	Sales management	Missing items per opportunity
2	Retrieve comparable projects	High	Project archive	Engagement type missing on part of the projects	Technical management	Share of opportunities with a comparable project identified
2	Qualify incoming opportunities	High	CRM + project archive	Qualification criteria not written down	Sales management	Share of opportunities qualified within 48h
3	Link evidence to context	High	Project archive	Irregular filing across years	Technical management	Share of recommendations with linked evidence
3	Decide whether to pursue	High	CRM + project archive	Decision not traced today	Sales management	Share of decisions documented
4	Surface the opportunities worth revisiting	Medium	CRM only	Incomplete activity history	Sales management	Opportunities requalified per month
5	Reuse commercial proposals	Medium	Project archive	Duplicate versions never resolved	Technical management	Time to draft a proposal
6	Grow existing relationships	Medium	CRM	No shared prioritisation rule	Sales management	Accounts active over 12 months

Annex 06

Economic assumptions and scenarios

Read before the numbers

No assumption is validated to date. These scenarios are neither a forecast nor a guarantee of revenue or ROI.

Business case assumptions
Assumption	Value	Status	Source	Owner
Relevant opportunities over 90 days	24	Illustrative	Extrapolated from the 48 CRM opportunities over 12 months	Sales management
Current search time per opportunity	3.5 hours	To confirm	Sales lead interview — never measured	Sales lead
Internal loaded cost	€95/hour	To confirm	Sector order of magnitude — to be replaced by Norevia's own figure	Finance
Average deal value	€100,000	To confirm	Midpoint of the €20,000 — €250,000 range	Sales management
Contribution margin	30%	To confirm	Working assumption — not disclosed at this stage	Finance
Additional win rate — conservative	2%	Scenario	Lower bound used for the conservative scenario	NateSystem
Additional win rate — central	5%	Scenario	Central assumption used for the arbitration	NateSystem
Additional win rate — ambitious	8%	Scenario	Upper bound, not predictive	NateSystem

Scenarios
Scenario	Capacity	Illustrative commercial	Indicative total	Multiple	Status
Conservative	€4,560	€14,400	€18,960	1.35×	Below the decision threshold
Central	€7,980	€36,000	€43,980	3.14×	Build threshold met, on assumptions still to confirm
Ambitious	€11,400	€57,600	€69,000	4.93×	Upper scenario, not predictive

What remains to be measured

Real search time per opportunity
Real volume of relevant opportunities over 90 days
Average deal value
Actual contribution margin
Adoption rate by the sales team
Recommendation validation rate

Annex 07

Build plan

Loop steps
Step	Input	Processing	Output	Validation
New opportunity	An opportunity is created or updated in the CRM.	Check that the opportunity is sufficiently qualified and that an owner is identified.	Context request triggered.	None — automatic trigger step.
Account context recovered	Account and opportunity identifiers.	Gather the sales history, the contacts and the exchanges attached to the account.	The “context recovered” block of the action sheet.	Reviewed by the owner at validation time.
Comparable projects identified	Sector, engagement type, client, estimated amount.	Match against the project archive by client, then by type and sector.	A list of comparable projects with the reason for the match.	The owner can discard an irrelevant project.
Evidence linked	Comparable projects retained.	Select the reusable documents and attach them to their source.	A list of readable evidence, each linked to its source document.	The owner consults the evidence before validating.
Missing information flagged	Qualification grid and available data.	Compare the information required to decide against what is actually present.	An explicit list of items to confirm with the client.	The owner completes or corrects the list.
Human decision	The complete action sheet with its justification.	The owner validates, edits, or requests an additional check.	A documented, timestamped, attributed decision.	This is the validation step: nothing bypasses it.
CRM task created	Validated decision and the next action retained.	Write a task attached to the opportunity in the CRM.	An assigned task with a due date.	Only after explicit human validation.
Measurement	Timestamps of the opportunity appearing and of the decision.	Compute the delay and the secondary indicators.	A weekly measurement view.	Weekly review with the business owner.

Included

One business loop
Two to three sources
One CRM integration
One action sheet workflow
One human validation step
One task write-back to the CRM
One primary metric
Team training
Documentation

Not included

Automated campaign
Sending without validation
Exhaustive CRM clean-up
Full indexing of every document
CRM replacement
A complete client portal
A second business loop

Indicative Build: €14,000 over 90 days.

Annex 08

Acceptance criteria

Acceptance criteria for the first loop
Criterion	Scope
The selection rules produce understandable results.	Required
Every recommendation carries a justification.	Required
Every piece of evidence links to a source.	Required
Missing information is visible.	Required
The user can edit the recommendation.	Required
No external action goes out without human validation.	Required
The task is created in the CRM after validation.	Required
Blocking duplicates are handled.	To confirm
Blocking errors are reported to an owner.	To confirm
The team can use the workflow after training.	To confirm
The CRM remains the source of truth.	Required
Exhaustive indexing of every company document.	Out of scope
Automatic sending of messages to end clients.	Out of scope

Norevia_Dossier_de_Valeur_Active_Annexes — Demonstrator version — synthetic data — August 2026

Demonstrator — synthetic data — Norevia Ingénierie is a fictional company and nothing shown here represents a real client outcome.