Key Highlights
- Every time your enterprise system changes, outsourced training content has to be rebuilt, and you pay for it again. Building the content in-house breaks that cycle.
- One finished hour of eLearning takes between 50 and more than 700 hours to produce, and for simulation-heavy systems training the figure sits at the top of that range (Chapman Alliance). At the interactive and simulation levels, that runs to tens of thousands of pounds per finished hour, and the figure excludes the two things that matter most: your experts’ time and ongoing maintenance.
- The knowledge that makes systems training work, how your organisation runs a process in practice, lives with your people, not with an external agency. An author network puts the people who hold it in charge of the content.
- Object-based capture and no-code editing have removed the old barrier to in-house authoring. At CHU Montpellier and Zuyderland, the customers’ own staff built the training, and Assima supplied the tooling.
Introduction
The invoice that arrives with every system update
A finance team moves to a new release. The screens shift, a couple of transactions change, and the carefully produced training library is suddenly out of date. If that library was built by an outside agency or bundled into a system integrator’s statement of work, bringing it current means a change request, a quote, and a wait. The same thing happens at the next release, and the one after that. The content is never quite finished, and the meter never quite stops.
The economics are unforgiving. According to the Chapman Alliance study that the industry still treats as its benchmark, one finished hour of eLearning takes anywhere from 50 to more than 700 hours to build, and the higher end is exactly where systems training sits, because interactive, simulation-based content is the most labour-intensive kind to produce. The study also put a price on it: development ran from around £8,000 per finished hour for the simplest courses to nearly £40,000 for simulation-level work, and those are conservative 2010 figures. What the quotes leave out matters more than what they include: not your subject-matter experts’ time, and not ongoing maintenance. For enterprise software, maintenance is not a footnote. It is the whole game, because the system changes several times a year.
So the outsourced model has two structural problems, and price is only the first.
The knowledge an agency cannot sell you
Here is the second, and it is the one that decides whether training works. A vendor or an integrator can teach the software. They know the screens and the standard configuration. What they do not know is how your organisation runs its own work inside that software: how a purchase requisition moves through your approvals, which exceptions your team hits on a Tuesday afternoon, how the handover between two departments happens here rather than in the reference build. That knowledge sits with your key users and your process owners. It is the difference between a course that shows people the buttons and a course that prepares them for the job.
When you outsource content creation, you are asking someone who lacks that knowledge to produce the material that depends on it most. The usual result is training that is technically correct and practically thin: fine for navigation, weak on the workflows where go-lives are won or lost. Your experts end up in long briefing calls trying to transfer what they know to an author who will still only approximate it, which erodes the time saving that outsourcing was supposed to deliver.
An in-house author network flips the arrangement. The people who hold the workflow knowledge become the people who build the content. Nothing gets lost in translation, because there is no translation.
Why in-house authoring is now realistic
There is a reason organisations outsourced this work for years. Traditional authoring was slow and technical. Building a realistic simulation meant screen recordings, screenshots, and developer time, and every system change meant starting much of it again. Asking a busy finance lead or a senior nurse to produce that was never going to fly.
Object-based capture changed the economics. Instead of flat screenshots, the Assima Capture Engine records each interface element as a separate, editable object, which produces a working clone of the application rather than a picture of it. Because the clone is made of objects, an author edits the thing that changed instead of re-recording the lesson. When a field moves or a button is renamed at the next release, you adjust that object and the change carries across the affected lessons. Assima describes this as roughly a 90 per cent reduction in rework, and the practical effect is that maintenance stops being a rebuild and becomes an edit.
That shift is what makes an author network possible, because it lets non-technical people do the work. Assima Train’s authoring is no-code, so a subject-matter expert can build and maintain simulations without depending on the development team. Assima Collaborate then gives the network somewhere to work: a shared library where authors create and reviewers approve, so a group of people can produce content together without it fragmenting. And where the same content has to run in several languages, the Total Language Localizer translates whole lessons without recapturing the process, so a Dutch or French site is not stuck bridging a language gap on top of everything else.
None of this removes the need for skill or standards. It removes the technical barrier that used to make in-house authoring impractical for the people best placed to do it.
Want to see what the switch is worth for your organisation? Our ROI calculator estimates the training time and cost you could recover by building and maintaining content in-house.
Teams that built their own, and what it bought them
Several of Assima’s customers already work this way.
At CHU Montpellier, the hospital’s own training team built more than 150 lessons across every staff profile for its electronic patient record, and cut the classroom time per course from three hours to 40 minutes. The content was theirs, made by people who understood the hospital’s processes, and it stayed current because they could edit it themselves.
Zuyderland, a Dutch group of four hospitals, went further. A team of its own nurses drew on their knowledge of patient-care processes to build around 300 exercises and prepare 2,500 staff for a new clinical system. Their manager reported that the migration stayed transparent to patients, who kept receiving care throughout, and the classroom programme roughly halved, from seven days to three and a half. Nurses built training for nurses, which is precisely the point.
The commercial version of the same story shows up at npower. When the UK energy company moved its mission-critical systems to SAP, replacing its old training client with Assima saved around £3 million over three years, cut new-starter onboarding at the contact centre from 33 days to 18, and let six people build what had previously taken more than thirty. Two or three of npower’s own developers learned the tool in a single session and then produced all the material the business needed. RSA, the UK insurer, did the same on a new underwriting system: two of its own developers built 23 interactive exercises on clones in four weeks, cut training per employee from ten days to seven, and saved around £300,000 a year. In each case Assima supplied the platform and the support while the customer supplied the content and owned it, which is the arrangement that keeps paying off after go-live rather than only during it.
Planning a major rollout and weighing build against buy? Our customer stories set out how these teams scaled training with their own people.
Standing up an author network without it turning into a second job
Owning the content does not mean overloading a handful of experts. A few decisions make the difference between a network that holds and one that quietly dies.
Start with the right people. The strongest authors are usually your key users and process owners, the ones who already answer their colleagues’ questions, not your most available staff. Give them the no-code tooling so authoring fits alongside the day job rather than replacing it. Set light governance early, using Assima Collaborate to define who authors, who reviews, and where the single approved version lives, so quality stays even as the group grows. Point the effort at the workflows that carry the most risk first, the tasks where a mistake in production is expensive, and build those to a high standard before filling in the rest. Then treat maintenance as a standing habit rather than a project, because the whole advantage of owning editable content is that a protocol change becomes a few clicks rather than a fresh procurement.
Done this way, the network compounds. Each release costs less to absorb than the last, the content reflects how the organisation works, and the knowledge stays in the building instead of leaving with an agency at the end of the engagement.
Conclusion
See it on your own systems. Book a demo and watch Assima turn your live apps into hands-on practice your team retains.
Frequently Asked Questions
Let’s Answer Some of Your Questions.
It is a group of your own subject-matter experts and key users who build and maintain your systems-training content, rather than outsourcing that work to an agency or system integrator. They hold the workflow knowledge the content depends on, and modern no-code authoring lets them produce it without needing developers.
Outsourced custom eLearning is priced per finished hour and runs into the tens of thousands of pounds at the simulation level, and those quotes usually exclude your experts’ time and ongoing maintenance. Because enterprise systems change often, the recurring maintenance cost is where outsourcing gets expensive, and it is the cost an in-house model removes.
No. Assima Train uses no-code, object-based authoring, so a finance lead or a clinician can build and edit simulations without development support. That is what makes an author network of subject-matter experts practical rather than aspirational.
Because Assima captures each interface element as an editable object, authors edit the part that changed instead of re-recording the whole lesson, and the change carries across affected content. This is what turns maintenance from a rebuild into a quick edit.
Governance. Assima Collaborate provides a shared library where authors create and reviewers approve, with a single approved version, so content stays consistent as the network grows across teams, sites, and languages.
Yes. The Total Language Localizer translates entire lessons without recapturing the process, so one authored simulation can be published across regions and languages from a single source, rather than rebuilt market by market.