Product Decision Review

An independent review of one product decision.

For the executive who owns a launch, platform, architecture or investment decision concerning an industrial product, a software-defined product or an industrial workflow. In two to three weeks you receive a written recommendation — go, redirect, sequence or stop — with the evidence behind it, the open questions and a prioritised 90-day plan. AI does not have to be part of it.

Discuss your project
Product Decision Review
  • Five decision moments5
  • Your Decision Dossier5
  • The 2–3 week review

The review at a glance

Is this for us?
A launch, platform, architecture or investment decision is unresolved, with a named decision owner and a deadline.
What do we receive?
A decision memorandum with a recommendation to go, redirect, sequence or stop; the evidence and risk map behind it; the questions that remain open; a 90-day plan with named actions and owners; and one executive readout.
What is the commitment?

Two to three weeks.

Individually scoped and quoted after a fit conversation.

The base scope is one decision, one organisation, one technical environment, up to five interviews and one workshop, delivered remotely. Anything beyond it is quoted before it starts.

What must our team provide?
An executive sponsor, access to existing technical and commercial documents, up to five interviewees, one workshop and one executive readout.
Can we inspect the output?

Yes. A specimen Decision Dossier, filled in for a fictional cobot cell, is further down this page, and the worked evidence pack behind it is public.

See the specimen dossier
Inspect a fictional worked decision

A method illustration with fictional cobot inputs. It is not a client deliverable, customer outcome or product release approval.

What does it not include?
Software implementation, a certification or conformity assessment, a legal opinion, a penetration test, a functional-safety assessment, or a guarantee that the initiative reaches production. Hyperion recommends; your organisation decides.

Situations this engagement can address

FICTIONAL SPECIMEN

A specimen Decision Dossier

The five parts of the Decision Dossier, filled in for the fictional cobot cell used in the worked evidence pack.

Prepared with AI assistance; not a client deliverable, not a client result. It shows the format of the dossier, not evidence of an outcome.

  1. FICTIONAL SPECIMEN · 01

    Decision statement & Product Bridge Contract

    Decision asked
    Release the assisted-assembly cell for operation across 8 units now, or prove it first on one bounded cell?
    Decision owner
    An illustrative accountable cell owner, acting as manufacturing product sponsor. A role, not a person: the scenario is invented.
    Business intent
    Flexible assisted assembly for a variable product mix, without shifting hidden burden to operators.
    Evidence that must return to the decision
    A measured human baseline, a safe restart after an interrupted cycle, and supervision evaluated at the level intended for operation.

  2. FICTIONAL SPECIMEN · 02

    Six-lens evidence & risk map

    Known, as modelled
    Availability of 99.3% against a 98% gate. The €2,100,000 budget covers 198.4% of the modelled upfront investment of €1,058,232.
    Assumed
    A lifecycle net value of €9,419,385 and a payback of 6.1 months. Both rest on invented inputs and on no measured human baseline.
    Missing
    3 of 9 completeness dimensions are declared gaps: coverage of part and tooling variation, safe restart after an interrupted cycle, and production acceptance and support commitments.
    Material risk
    The combined safety and evidence risk scores 63/100, and above 55 it blocks. The cell was evaluated under a continuous operator and is intended to run under monitoring only. Operating-envelope evidence stands at 79%, below the preferred 80%.
    What the evidence does not support
    The €9,419,385 figure does not support a release. It models 8 units, the comparison with manual assembly has not been observed, and 12 operator minutes per duty hour (12 ÷ 60 = 20% of one operator) contradict the monitoring-only intent.

  3. FICTIONAL SPECIMEN · 03

    Options & architecture record

    Option A: go, release 8 units
    Costs the modelled upfront investment of €1,058,232. Risk: 5 gates fail and 2 are conditional, rollback is not supported by evidence, and the supervision gap goes into production.
    Option B: sequence, one bounded cell first
    Costs about €132,279 (€1,058,232 ÷ 8 units; a derived share, not a quote) and moves the 38-week release forecast. Risk: one cell does not cover part and tooling variation unless the trial is designed for it.
    Option C: stop
    Costs nothing further and gives up a value case that is modelled, not observed. Risk: the manual burden the product was meant to relieve stays unmeasured.
    Architecture choice recorded
    The deterministic controller and its stop function stay the authority boundary, and the assembly policy acts inside it. The critic is not independent of the system it evaluates, so its scores are not accepted as release evidence.

  4. FICTIONAL SPECIMEN · 04

    Decision memorandum

    Recommendation
    Sequence. Authorise one bounded cell as an experiment; no operational release of the 8 units on this evidence.
    Conditions
    The cell runs under a continuous operator, the supervision that was evaluated. A human baseline is measured before any improvement is claimed. A named owner signs the production acceptance gate.
    Dissent
    The funding case argues for going now: 198.4% coverage and a payback of 6.1 months. The memorandum records that view and does not follow it, because both figures are modelled.
    Stop condition
    Stop if the cell cannot restart safely after an interrupted cycle, or if recoveries stay at the assumed 75% (100 − 75 = 25% of recovery attempts failing) with no owner for the remainder.
    Reopen when
    The policy, model, hardware, tooling, site or intended supervision changes.

  5. FICTIONAL SPECIMEN · 05

    90-day commitments & executive readout

    Days 1 to 10
    The cell owner defines one accepted assembly cycle and starts measuring the manual baseline: quality, cycle time, rework and operator burden.
    Days 11 to 20
    Engineering tests the restart after an interrupted cycle on the bounded cell. The cell supervisor owns recovery.
    Days 21 to 30
    The sponsor checks the baseline and the restart results against the stop condition, then confirms or ends the experiment.
    Days 31 to 90
    Committed only after the day-30 check: variation coverage across parts and tooling, and written acceptance and support terms.
    Executive readout
    One live session with the sponsor: the recommendation, the dissent, the conditions and the evidence still missing.

Every figure comes from the worked evidence pack for the same fictional cobot cell: invented inputs and modelled outputs. A figure derived here shows its formula.

Open the worked evidence pack

Ongoing product advice

Ongoing advice gives you an independent view on product choices, priorities and trade-offs, with practical next steps. Your leaders keep decision authority and delivery. Scope, outputs, cadence and fees are agreed separately.

Discuss ongoing product advice
Five decision moments

Five decision moments

One review format, five starting points — chosen by the decision in front of you.

  • Opportunity & Product Strategy

    Is this the right product opportunity, for the right customer, with a credible path to value?

  • Product & Production Readiness

    Can this prototype or pilot become a dependable, supportable product — and what blocks it?

  • Launch Readiness

    Can the product be sold, installed, accepted, supported and operated responsibly?

  • Product Family & Scale Readiness

    Do value, deployment and economics repeat across sites, machines, customers and variants?

  • Model, Knowledge & Architecture Decision

    Which system boundary, model strategy and architecture can meet the operating, evidence, safety and economic constraints?

When a review makes sense

  • A consequential product, architecture, launch or investment decision remains unresolved.
  • A pilot works in a demo or lab but can't get sign-off for production.
  • A board or executive sponsor wants the delivery risks quantified before releasing further budget.
  • A go-live slipped, and no one agrees on why.
  • OT/IT integration, reliability or safety questions are blocking the next step.
  • You're deciding whether to build, buy, partner or stop.
  • Software is forked for every variant, release recovery is unproven, and nobody owns what is shared.
  • An AI capability in your own operation is stuck between trial and daily use, and you need to decide whether to recover, replace or stop it.

Who it's for

  • Product, engineering, operations and AI leaders at European manufacturers, automotive and mobility companies, and energy and charging organisations.
  • Teams with a specific product, platform, workflow or deployment decision — not a request for generic advice.
  • Sponsors who want an independent, technical read before the next investment.
  • AI is not a condition: a software-defined product, platform or release decision qualifies, and so does a decision about one workflow in your own operation.

Who it's not for

  • Open-ended transformation work without one named decision, owner or decision date.
  • Pure research with no production intent.
  • Teams looking for a vendor to simply build, without first understanding the blockers.
What waiting costs

What waiting costs

A consequential decision does not stay neutral while it is unresolved. Teams keep building against the assumption nobody has tested, and every committed sprint makes reversing it more expensive.

  • Budget keeps flowing into an option the evidence may not support.
  • The decision is eventually forced by a deadline instead of settled by evidence.
  • Blockers found after commitment — safety, integration, unit economics — cost far more than blockers found before it.

Your Decision Dossier

  1. Decision statement & Product Bridge Contract

    The decision to close, who owns it, how customer and business intent constrains the product and development boundary, and which technical or field evidence must return to the business decision.

  2. Six-lens evidence & risk map

    Knowns, assumptions, missing evidence and material risks across customer value, economics, AI and data, architecture, safety and operations. Where material to the decision, the dossier can connect the affected role to available and conflicting evidence, with sources, then record the next test and owner, the decision date and the reversal condition. The Review recommends the next learning step; it does not recruit interviewees or run new customer research.

  3. Options & architecture record

    The viable options, trade-offs and recommended product or architecture choice, recorded so the reasoning remains inspectable. Where decisive, the same dossier can record product-economics options with sources, ranges and sensitivities, or complete-offer readiness by capability, owner, evidence, dependency, fallback and acceptance. It uses existing material and client-provided access within the same Review scope.

  4. Decision memorandum

    A clear recommendation to go, redirect, sequence or stop — including the evidence, dissent and conditions behind it.

  5. 90-day commitments & executive readout

    Named actions, owners, dependencies and kill criteria, followed by a live leadership readout of the decision and plan.

01—02

The 2–3 week review

Who conducts the review

The review is conducted personally by Hyperion's founder, Mohammed Cherifi. You get senior independent product judgment grounded in the six lenses, explicit evidence challenges, dissent and reversal conditions — not a junior checklist. Any later delivery mandate is separately decided and scoped; the review remains useful if you stop or work with someone else.

  1. 01

    Phase 1 — Frame & understand

    • Kickoff and scope confirmation
    • Document review (existing technical and business materials)
    • Up to five stakeholder interviews
    Your team
    Name the sponsor and decision; provide existing documents and up to five interviewees.
  2. 02

    Phase 2 — Assess, decide & commit

    • One working session / workshop
    • Evidence, option, architecture and operational-risk analysis
    • Decision Dossier, 90-day commitments and executive readout
    Your team
    Attend the workshop, challenge the assumptions and assign owners to the agreed next actions.

The bounded base scope

Remote by default. On-site work, additional sites or use cases, and specialist reviews are scoped separately (see below). The start date and the elapsed duration are confirmed in the proposal against the founder's available days.

Every review begins with the same written delivery boundary:

  • One product, platform, workflow, opportunity or consequential decision
  • One organization or business unit
  • One primary technical environment
  • Up to five stakeholder interviews
  • One primary workshop
  • A fixed document-review allowance
  • Remote delivery
  • One executive readout

Scope extensions

These additions change the written scope and quote, and proceed only by prior agreement:

  • Additional use cases or systems
  • Additional sites or business units
  • More than five interviews
  • On-site work
  • A specialist safety or cybersecurity review
  • Complex regulatory classification
  • Additional workshops
  • Urgent or compressed delivery
  • Documents substantially exceeding the agreed allowance

What this review is not

To keep it honest and focused, the review explicitly does not include:

  • Software implementation or model development
  • A formal certification or regulatory conformity assessment
  • A legal opinion
  • A penetration test
  • A full functional-safety assessment
  • A guarantee that the product or initiative will reach production

It is a strategic and technical readiness review — an independent decision aid, not an audit, certification or guarantee.

The operating envelope changes first.

The operating envelope changes first.

A demo shows behaviour under selected conditions. Operational use needs evidence for the intended machines, environments, failures and changes; a demonstration alone does not establish safety.

Representative transformation lanesDemo conditionOperational obligation
Scale
One configured device
A versioned fleet across sites
Field conditions
Chosen conditions
People, materials and environments keep changing
Failure response
An engineer intervenes
Degraded modes, rollback and recovery are designed in
Connectivity
A stable lab connection
Intermittency, latency and offline operation are expected
Evidence
The demo is watched live
Telemetry, audit trails and fleet health persist
Change control
A manual deployment
Signed releases, staged OTA and compatibility controls
Decision authority
A proposal is exercised under supervised demo conditions
Named authority, qualified control interfaces and independently assessed protective functions

Value on its own

Even if you decide to stop, or to continue without Hyperion, you leave with an independent evidence and risk map, a clear decision, and a 90-day plan your team can carry out. The review is designed to be worth its price regardless of what you do next.

If you decide to proceed

If the decision is to proceed, the 90-day roadmap becomes the basis for a Product Leadership Mission. The review stands on its own value.

FAQ

Is this a certification or audit?
No. It's a strategic and technical readiness review — an independent decision aid. It is not a certification, regulatory conformity assessment, legal opinion, penetration test or functional-safety audit.
Do you guarantee our pilot will reach production?
No. The review gives you an honest assessment and a plan; whether to proceed is your decision, and reaching production depends on execution and factors beyond the review.
Does the decision have to involve AI?
No. A platform, variant, release or operating-model decision for a software-defined product qualifies with no AI in it, and so does a decision about one workflow in your own operation. Where AI is proposed, the review compares it with the non-AI alternative first.
Is it remote or on-site?
Remote by default. On-site work is available and scoped separately.
How do we start?
With a 30-minute, no-obligation fit call to confirm the initiative — and Hyperion — are a good match. You don't need to complete any long intake form to book the call.
What do you need from us?
Access to the product, system or workflow under review and relevant documents, a few stakeholders for interviews, and a slot for the workshop and the readout.
What if we have little or no customer evidence?
Then the review says so. Missing evidence is recorded as missing, and the recommendation can be to sequence: build the evidence first, with the questions to answer, who to ask and what result would change the decision. The interviews (up to five) can include customers or operators you make available; the review does not recruit interviewees or run new customer research.

Individually scoped and quoted after a fit conversation.

Price & terms

  • Prices exclude VAT and any other applicable taxes. Tax treatment depends on the customer's location and tax status.
  • 50% at booking, 50% before the executive readout. Where Article L221-3 of the French Consumer Code applies (an off-premises contract: one concluded in person away from Hyperion's premises, or by email or other remote means immediately after the client was approached in person away from them, with a business of five employees or fewer, outside its main field of activity), the deposit is neither requested nor collected until seven days after the contract is concluded, the client has fourteen days to withdraw, and work starts within those fourteen days only at the client's express written request. If the client then withdraws, it owes only the part of the price that corresponds to the service already provided.
  • The report and roadmap are licensed to the client under the Terms of Sale: on full payment, a perpetual, non-exclusive licence to use them for its internal business purposes. The client may share them in confidence with its professional advisers, auditors, insurers, investors or lenders, and with suppliers it engages to implement or continue the work, each bound by a duty of confidentiality; any other sharing with third parties needs Hyperion's written consent. Where a deliverable is to go into a product or service the client supplies to others, the wider rights are set out in the proposal. Hyperion retains ownership of its pre-existing methodologies, templates, tools and general know-how.
  • Reschedule with at least five business days' notice at no charge.
  • You may stop the Review at any time by written notice: you then owe the work actually performed at the stop date, which Hyperion documents, and never more than the deposit before the readout date is confirmed in writing; once that date is confirmed, the balance is due in full only if the draft decision memorandum has been delivered, otherwise only for the work performed, and anything paid beyond what is owed is refunded within 30 days. If the founder cannot continue for more than ten consecutive working days, or in a way that makes an agreed date impossible to keep, you may suspend or end the Review immediately and fees paid in advance for work not performed are refunded within 30 days of the end of the engagement (Terms of Sale, section 11).

What stays with you

The review is bounded on both sides. Hyperion owns the analysis, the recommendation and the readout. These stay with your organisation:

  • One named decision owner, and an executive sponsor who will receive the readout
  • Access to the pilot or system and to the existing technical and commercial documents
  • Up to five stakeholders available for interview, and one slot each for the workshop and the readout
  • Test results, field data and incident history where they exist — and a plain statement where they do not
  • The decision itself: Hyperion recommends, your organisation decides

Relevant evidence

The judgement behind this review rests on a founder career record, one bounded client mandate and Hyperion-authored research references. Each is labelled with what it does and does not prove; modelled and simulated work is labelled as such and is never presented as a client outcome.

Review the evidence

Every engagement is led directly by Mohammed Cherifi. Specialist partners may be introduced for clearly defined work when required and disclosed; Hyperion is not a staffing or engineering-delivery bench.

Start with a conversation, not a commitment.

Tell me what you want to achieve, what is blocking progress and when you need help. A few sentences are enough.

Discuss your project

No sales funnel