Off-the-shelf content teaches the general case. Your people need the specific one.
Every enterprise already has training that is technically fine and practically useless, because it was written for a generic organisation. This page is about the other approach: content designed with your subject matter experts, in your business language, against your actual processes — and, where the behaviour you need to change is too important to leave to a video, simulated until it is practised. It also sets out honestly where the three available options — bespoke, off-the-shelf and generated — actually make sense.
Consulting first
Needs diagnosis and blueprint before a single slide is written
Simulation over video
MetaSim makes people practise the decision, not watch it described
Standards compliant
SCORM, xAPI, WCAG accessibility, multilingual, local regulatory wording
Scored as evidence
Simulation results feed capability records and compliance evidence
Bespoke, off-the-shelf, generated — and where each one is the right answer.
We are a bespoke courseware firm and we are also a courseware firm that will tell you not to buy bespoke courseware. Most programmes need a mixture. The table below is how we allocate the budget across it.
| Dimension | Bespoke | Off-the-shelf | Generated |
|---|---|---|---|
| What it is | Content designed with your SMEs against your process, policy and tools | Existing catalogue content, licensed | Content produced by a model from a prompt, usually reviewed afterwards |
| Strongest use | Regulatory obligations, product knowledge, sales process, system procedure, new role induction | Foundational skills, compliance breadth, professional development electives, induction at volume | First drafts, variant localisation, knowledge refresh, draft question banks, supplementary explainers |
| Realistic timescale | Several weeks per course, driven by review cycles and SME availability | Available immediately, subject to a licensing decision | Days, but with a review and revision load that follows |
| What it cannot do | Scale to a very large population cheaply, or survive without sustained SME involvement | Carry your specific process, terminology, judgement calls or escalation thresholds | Carry regulatory wording safely, and it does not substitute for an assessment design |
| Evidence quality | Full — assessment and simulation built in from the start | Variable — depends on the assessment the vendor included | Weak on its own; acceptable only after human review, versioning and assessment design |
| Our typical position | Default for anything you must certify against | Used for breadth and electives, with a gap review | Used inside bespoke work to compress production, never to certify on its own |
The rule we apply. If a learner completing this content would be making a decision that affects a customer, a payment, a disclosure or a person's rights, the content is bespoke and the assessment is designed with it. Everything else is a candidate for a catalogue or for a generated draft — and that allocation is usually where most of the money is saved.
Courseware in numbers
Used when they genuinely address the capability
Then reused across every region that runs the process
Real workflow, real assessment, real evidence
If there is no gap, we will tell you to use the catalogue
When custom courseware is the right answer, and when it is waste
Custom courseware is the most abused line in most learning budgets. This matrix is the argument we have with clients, and it is why we sometimes talk them out of building something.
Custom courseware
Specific workflow, meaningful population. Build it. This is the case bespoke courseware exists for, and it is genuinely worth it.
Bespoke light
Specific workflow, small population. Build a light version — a scenario, a simulation, a job aid — rather than a full course.
Configure the catalogue
General workflow, large population. Configure off-the-shelf with your own examples and terminology. Building here is usually waste.
Off-the-shelf
General workflow, small population. Take the catalogue. Do not let anyone build something because it feels more tailored.
Six stages, and the artefact each one produces.
The stages look like a production process, but the first two are the ones that determine whether the programme works. We have seen beautifully produced courses built on an incorrect diagnosis more than once.
Needs diagnosis
Interviews with the people doing the work and the people managing it, observation of the actual task, review of the current record, and a clear statement of the behaviour that has to be different afterwards. Output: a diagnosis that names the gap, the population, the evidence required and the business consequence of getting it wrong.
Blueprint
Module architecture, learning objectives written as observable behaviours, assessment design, simulation scenario design, delivery mode, sequencing against the calendar, and the exact evidence the programme will produce. Output: a blueprint signed off by the business owner, not only by the project sponsor.
Script and storyboard
Every module is scripted, not improvised: the narrative, the scenario, the questions, the feedback, the branch conditions, the local regulatory wording. This is the stage where subject matter experts contribute the parts only they can supply, and it is the longest review cycle in the project.
Production
Build in the agreed authoring standard, with accessibility and multilingual requirements designed in rather than retrofitted. Localisation happens on the structure, not on finished pixels, so a regulatory change in one market does not require re-shooting the course.
Pilot
A real cohort, in the real environment, with the real assessment — not a showreel. We watch where learners hesitate, where the scoring does not discriminate between a good and a poor answer, and where the scenario stops resembling the job. Output: a specific change list, not a satisfaction survey.
Iteration and operations
Performance data drives the revision: which items nobody gets right, which distractors are never chosen, which scenario branches are never reached. Content moves to a maintenance cycle with a named owner, an expiry and a review trigger tied to policy change.
How bespoke courseware is actually built
Six steps, and the first one is the one most programmes skip. Without a named capability gap, custom courseware is a discretionary purchase rather than an investment.
Capability gap, named
A specific gap, evidenced from the capability record — not a request from a vendor's catalogue.
Workflow captured
Observed on the floor, including the exceptions that never appear in the procedure manual.
Assessment standard set
What a competent person can do, stated in observable terms, agreed with the business before authoring begins.
Authored and piloted
Built with real practitioners, piloted with a real cohort, revised against what they actually struggled with.
Delivered and evidenced
Into the LMS with completion, assessment result and validity period — retrievable by risk.
Measured six months later
Is capability holding? Is the exception rate falling? This is the only evidence that the courseware worked.
Business and sales simulation: the part that actually changes behaviour.
A course tells you what should happen. A simulation makes you decide, under time and information constraints, and then shows you the cost. For any behaviour you intend to rely on, the second is worth considerably more than the first.
Scenario architecture
A simulation is a sequence of decision points, not a quiz. At each point the learner sees a situation built from your real cases, chooses from options that reflect the decisions people actually face, and gets a consequence that is internally consistent with your policy and economics. Information is deliberately incomplete, because complete information is not what a working day looks like.
Scoring design
Scores are computed against a rubric written by your experts, not a generic pass mark. A critical decision weighted differently from a cosmetic one. Partial credit is used where a partly-correct answer is genuinely instructive, and penalties are applied where a wrong decision has a real cost. The score has to mean something to the business, or it will not be used.
Feedback and debrief
Immediate feedback explains the reasoning, not just the answer, and the guided debrief is where the learning consolidates: the facilitator walks the cohort through the decision points, the alternative paths, and the moments where reasonable people would disagree. That last part is how you teach judgement rather than rules.
The same engine, three uses
Learners use it as assessment. Facilitators use it to practise a discussion before it happens with a customer. And an agent uses it as a rehearsal environment — running scenarios in a sandbox where the cost of a wrong decision is a score rather than a customer, before the agent is allowed near live work.
Three design rules we do not break
Every one of these exists because we have seen the alternative fail. Together they are the difference between courseware that changes capability and courseware that fills a folder.
Rule 1 — Build against a real workflow
The course is designed from an actual process, with the actual exceptions and the actual systems. A course built from a process diagram someone drew from memory will not survive contact with the floor.
Rule 2 — Assess against a standard
The assessment tests whether someone can do the thing, by someone who knows the domain. Completion is recorded separately and never presented as capability.
Rule 3 — Evidence from day one
The completion, the assessment result and the validity period are recorded in the LMS in a form your risk function can retrieve. Retrofitting evidence is a different and more expensive exercise.
Content that survives a technical and a legal review.
Courseware is reviewed by two audiences who rarely talk to each other. The technical reviewer asks whether the package will run, track and integrate. The legal or regulatory reviewer asks whether the wording is right, current and defensible. Content built for one audience routinely fails the other.
- SCORM and xAPI — packaged for your LMS with correct manifests and identifiers, or delivered as a tracked web experience with xAPI statements for detailed activity data
- WCAG accessibility — keyboard operability, focus order, captions and transcripts, contrast, alternative text, screen-reader semantics; tested rather than asserted
- Multilingual and localisation — translation with in-context review by a subject matter expert, and structure designed so a market-specific policy variant is a content change rather than a rebuild
- Local regulatory wording — jurisdiction-specific obligations stated in the jurisdiction's own terms, with the source and effective date recorded against each requirement
- Version control — every release versioned with an effective date, an owner and a change note, and every completion record pointing at the version the learner actually saw
- Device and bandwidth — delivery tested on the hardware and connections your workforce actually has, including shared and mobile devices in the field
What a learner receives
Every bespoke module ships as a set of artefacts, not a file. The business sponsor gets a capability; the compliance function gets a record; the facilitator gets the debrief guide; the platform gets the tracking.
The module, packaged
Scored, branchable where the process branches, accessible, translation-ready, versioned.
The assessment instruments
Items, distractors, weights, critical-fail conditions and pass thresholds — documented for audit.
The facilitator guide
Debrief structure, the reasoning behind each scored decision, and the questions worth asking.
The maintenance pack
What triggers a review, who owns it, and what changes when policy does.
Where to take it off-the-shelf, and where to build
Read across for the topic, down for the depth. Most organisations need the top-left region only, and buy custom content they should have configured.
| AI fundamentals | Prompt & context | Agent operations | Data & ontology | AI governance | Agentic sales | Leadership | |
|---|---|---|---|---|---|---|---|
| Off-the-shelf (Academy) | |||||||
| Configure with your context | |||||||
| Custom courseware | |||||||
| Assessment standard |
How a simulation score becomes a compliance record.
The value of a course is only realised when its result means something outside the course. Here is the chain, and it is the reason bespoke courseware and a system of record are sold together rather than separately.
Activity recorded
The simulation run is logged against the individual: which scenario version, when, with what score, and which decision points were taken. The record points at the material version, so a later policy change does not rewrite history.
Rubric applied
The score is mapped against the role's capability profile, with the weighting agreed at blueprint stage. A high overall score with a failed critical decision does not certify — critical-fail conditions are configured to be blocking.
Capability updated
The result writes to the capability record with an issue date and an expiry, contributing to a coverage view by team, site, tier and jurisdiction. This is the same record that holds course completions and, for agents, evaluation results.
Renewal triggered
Where the simulation is a certification, the expiry triggers reassessment automatically. Where it is a continuing development activity, the hours are derived from the completed record rather than claimed by the learner.
Why this matters for agents. The same chain runs for a digital employee: scenario run, rubric applied, capability recorded, expiry enforced. It is why an agent can hold a "certification" in the same register as a person, and why the question "is this agent still qualified to do that task" has a single answer in your system rather than two.
Four evidence states, and only the last one is useful for a decision
If your LMS cannot distinguish between these four, it is storing course history rather than capability. That distinction is the whole design problem.
Never claimed
No record at all. Most of your workforce on a given capability sits here, and most organisations do not know it.
Claimed
A self-declaration, or a completion record. Useful for finding the population; not evidence of capability.
Demonstrated
Completed a real task from their own domain to a stated standard, assessed by someone who knows the domain.
Currently valid
Demonstrated, in scope, within its validity period, with no conflicting change since. This is the only state you can make decisions on.
We do not reskin a template and call it bespoke.
A template with your logo, your colour palette and your company name is not custom courseware. It is a licensing arrangement with a presentation layer, and it cannot teach a process that does not exist in the template.
We do not produce content without a subject matter expert in the room. If your organisation cannot commit that time, we will tell you that the programme will be weaker and quote accordingly — we will not absorb the gap and let the client discover it in a pilot.
We also do not certify learners on generated content that has not been reviewed by a named owner, and we do not produce training whose only purpose is to generate a completion figure. If a business wants a number rather than a capability, that is a conversation we can have — but we will not present it as capability work.
If a course is largely unchanged except for the branding, it is off-the-shelf content and should be bought as such. Selling it as bespoke is a waste of your budget and dishonest about what the learner will get.
The accuracy of a course is bounded by the knowledge contributed to it. Without your experts, we are guessing at your process — and you would be certifying against our guess.
Generated material can speed up drafting and localisation, but it carries no review, no version and no authority. It enters a regulated programme only after a named owner has read it and taken responsibility for it.
Every programme is piloted with a real cohort before it is certified against. The pilot is where you find out the assessment does not discriminate — and finding that out late is expensive.
When we will talk you out of custom courseware
We will not build custom courseware without a named capability gap, and we will not produce a course that nobody can be assessed against. We also will not build the same course twice — if you need it in five markets, we design it once and localise the examples, because five slightly different versions is five maintenance liabilities.
Five questions that tell you whether a learning system is actually working
Five yes-or-no answers. Three or more yes means the foundation is real. Fewer means the learning system is a reporting tool rather than a capability system — and an agentic business cannot be built on that.
- Can you name who can do what, and how you know that is still true this month?
- Does a regulator or auditor get an answer from records you already hold, or do you assemble it?
- When a policy changes, can you identify everyone affected and what they must retrain?
- Do external partners who touch your process appear in the same system as your employees?
- Do your digital employees have a training record, or only a version number?
Bespoke for the important, catalogue for the rest, and a factory that can keep up.
A large enterprise needs both. Our own Academy catalogue covers the broad, general-purpose ground — thousands of titles, available immediately — and the bespoke studio handles the small number of areas that actually carry your risk. The design principle is that the two never compete for the same content.
Academy catalogue
Thousands of general-purpose titles across professional development, compliance fundamentals, leadership and AI-related awareness, available on subscription and immediately assignable. Used where the generic case is good enough and the volume matters.
Bespoke studio
A smaller number of programmes with high consequence — regulatory obligation, your product, your sales motion, your systems, your new roles. Built with your experts, assessed against your rubric, evidenced in your record.
Manufacturing capacity
Once the blueprint is right, production is a repeatable operation — parallel authoring, structured review gates, a versioning pipeline, and a rollout plan across sites and languages. Quality is held by the process, not by the individual producer on the day.
What this page answers
Every topic below is covered in the sections above, and each links to the page that goes deeper on it.
Give us the one course everyone says is useless.
Every organisation has a mandatory programme that people complete and then ignore. Send it to us and we will run a two-week diagnosis: what the content gets wrong, what the process it describes has become, and what evidence would tell you whether it is working at all.