October, and already the year's first deadline: your training centre has just asked you for a "dissertation topic" when you've barely got back into your workplace assignments. Don't panic — but don't procrastinate either, because right now is precisely when the real work happens.
In short: in an apprenticeship, the professional dissertation (or activity report, depending on the qualification framework) and its defence frequently account for 25 to 40% of the final grade, and sometimes for the entirety of certain competency units within the qualification. The difference between a successful defence and one you merely endure comes down to three things: a research question rooted in a genuine workplace assignment, quantified evidence gathered throughout the year, and out-loud rehearsal — not a silent read-through the night before.
Dissertation, report, professional portfolio: what are we talking about?
Not all apprentices submit the same document, and that's the first source of confusion. Check your qualification framework (available on the France Compétences record for your qualification, accessible from the France Compétences website using your RNCP code) to see what's expected of you.
| Type of deliverable | Qualifications concerned | Typical length | What the panel is looking for |
|---|---|---|---|
| Activity report | Vocational certificates, vocational baccalaureate, two-year technical diplomas (certain specialisms) | 15 to 30 pages | Accurate description of tasks, trade vocabulary, professional conduct |
| Professional portfolio (DP) | Ministry of Labour professional titles | Set template | Evidence of each competency in the framework, through examples |
| Professional dissertation | Vocational bachelor's, bachelor's, master's, MBA | 40 to 80 pages | Research question, methodology, analysis, recommendations |
| Professional thesis | Specialised master's degrees, business schools | 60 to 100 pages | Literature review + original contribution |
The distinction matters: an activity report recounts, a professional dissertation demonstrates. If you're on a two-year technical diploma and you write a theoretical essay, you'll be off-topic. If you're on a master's and you describe your daily routine without analysis, you'll be penalised for the same reason in reverse.
A useful habit from the start of term: ask your training centre for the assessment grid used by the panel. In the vast majority of centres, it can be shared. Working without knowing the graded criteria is like aiming at a target with your eyes closed.
Choosing a research question that will last the whole year
This is the most rewarding stage of the process, and the one most often rushed. A good apprenticeship research question meets four conditions simultaneously.
The four criteria of a good topic
- It starts from a real company problem. Not from a fashionable theme. "How can we reduce quotation processing time in our regional branch?" is a thousand times better than "Will artificial intelligence transform retail?".
- You have access to the data. If answering it requires confidential figures your workplace mentor will never give you, change your topic now, not in April.
- It falls within the scope of your assignments. The panel checks the consistency between your job description and your analysis.
- It can be phrased as a closed or measurable question. An infinitely open question ("What future for the sector?") produces no usable conclusion.
A simple test: try to phrase in a single sentence the final recommendation you envisage. If you can't even sketch it out, the question is too vague or too broad.
The topics that mixed panels like
The panel for an apprenticeship qualification almost always includes a practitioner — often an industry representative or a former apprenticeship mentor. That member isn't looking for literature: they want to know whether you're job-ready. Angles that work well:
- optimising a process you have carried out yourself (invoicing, onboarding, logistics for a given flow, quality monitoring);
- introducing a tool or an indicator and measuring its effects;
- analysing a gap between a theoretical procedure and observed practice;
- running a mini-project with budget and deadline constraints.
To structure this thinking, a reference read on professional dissertation methodology helps you avoid reinventing the wheel: methodology books in social sciences and management offer tried-and-tested outlines, from literature review to discussion of results.
Turning your assignments into raw material
The classic apprentice trap: arriving in March with twelve months of experience and nothing written down. You won't remember the figures, the dates, or the name of the tool you tested in November.
A logbook, your best insurance
Keep a weekly logbook, ten minutes every Friday. Three columns are enough:
| Week | What I did | Figure, tool or difficulty worth noting |
|---|---|---|
| W41 | Took over follow-up on customer payment reminders | 180 cases in the portfolio, average delay 32 days |
| W42 | Tested a reminder email template | Response rate up from 19% to 27% across 60 sends |
A simple A5 notebook kept in your bag does the job perfectly, and it has the advantage of not depending on computer access that may be locked down at your employer's. Those who prefer digital can replicate the table in a shared spreadsheet — provided they respect the company's confidentiality rules.
Requesting data, formally
Two golden rules on company data:
- Ask in writing, explaining the purpose ("for my end-of-year dissertation, defence in June, document reviewed by a three-person panel"). Your mentor will then be able to decide with peace of mind.
- Anonymise by default. Client names, detailed revenue figures, margins: replace them with "Client A", base-100 indices or orders of magnitude. Many companies will accept a dissertation provided it carries the note "confidential document — circulation restricted to the examination panel".
As for external sources, favour references the panel can verify: Dares and Céreq studies on employment and labour-market entry, Insee data, the France Travail Labour Needs surveys, publications from your industry body or your OPCO (skills operator). Three well-cited institutional sources are worth more than fifteen blog links.
A realistic year-long reverse schedule
The calendar below assumes a defence in June, the most common case. Shift it if your session falls in September or January.
| Period | Objective |
|---|---|
| October – November | Read the qualification framework and assessment grid; 3 possible topics approved with your workplace mentor and academic supervisor |
| December | Research question settled, detailed three-part outline, initial bibliography |
| January – February | Data collection, internal interviews, first results |
| March | Writing the main body of the dissertation (60% of the text) |
| April | Finalisation, review by your mentor, layout, appendices |
| May | Submission of the document, building the defence slide deck, minimum 3 rehearsals |
| June | Defence |
The breaking point is almost always March: that's the month when workplace assignments intensify and the writing falls behind. Block out two half-days of writing per month in your diary right now — and let your mentor know. Remember that time spent at the training centre and coursework linked to the programme count as effective working time: raising this with your employer is entirely legitimate.
To keep up this pace, equipment matters more than you'd think: a noise-cancelling headset turns an open-plan office or a train journey into a usable writing slot. And if you're juggling between the training centre's computer and the company's, an external SSD avoids the drama of a file lost in submission week — provided you also back up to your school's cloud.
Building the defence slide deck
Twenty minutes of presentation, ten to twenty minutes of questions: that's the most common format. For a level 6 or 7 qualification, plan on 12 to 18 slides, no more.
The outline that works
- Who you are, where you were (1 slide, 60 seconds): the company, the department, your role. No company history dating back to 1952.
- The problem (2 slides): the starting observation, with figures. This is where the panel decides whether or not it's going to be bored.
- The question and the method (2 slides): your research question, how you went about answering it, and the acknowledged limits of your method.
- The results (4 to 6 slides): one chart per idea, not one chart per Excel file.
- The recommendations (2 to 3 slides): prioritised, with estimated cost and timeframe. A practitioner panel loves a recommendation that's "doable Monday morning".
- What the apprenticeship taught me (1 slide): the reflective step back expected on your skills development, linked to the qualification framework.
One slide = one idea = one assertive title sentence. "Average processing time fell from 32 to 24 days" is a title. "Results" is not.
The mistakes that cost you marks
- Reading from your notes. The panel spots it within ten seconds and your communication mark collapses.
- Running over time. Most grids explicitly penalise overruns. Time yourself.
- In-house jargon. Your internal acronyms mean nothing to an external panel: spell them out once.
- Criticising your employer. You speak of "areas for improvement identified", never of "things don't work well over there".
- An illegible deck. Font below 20 points, blurry screenshots, 40-row tables: avoid.
Rehearsing: the part nobody does enough of
Three full rehearsals, out loud, standing up, stopwatch in hand. That's the minimum.
- Rehearsal 1 — alone. Goal: hold the thread without notes. Record yourself: a simple USB microphone or your phone will do. You'll discover your verbal tics and your empty passages.
- Rehearsal 2 — in front of a non-specialist friend or relative. If they understand your research question, a panel will too. Ask them to note every moment they lose interest.
- Rehearsal 3 — in front of your workplace mentor or a colleague. They're the one who'll spot the factual error or the data point best left off the slide.
Also prepare a list of likely questions along with three-sentence answers: "Why this choice of method?", "How much does your recommendation cost?", "What would you have done differently?", "Is this solution transferable to another branch?". An honest answer such as "I didn't have access to that data, so I worked on a sample of 60 cases" goes down very well; a made-up answer is spotted immediately.
On the day, take care of the practical details: deck emailed to yourself and saved on a USB stick, a PDF version as backup, and a printed copy of the dissertation with your annotations. An understated laptop bag and an outfit in keeping with your sector's conventions complete the picture: the defence is also an exercise in professional bearing.
What if it goes wrong?
Two situations are worth knowing about in advance.
You didn't get the data. That's not a dealbreaker. Own it in a "limitations of the study" section and compensate with internal qualitative interviews (4 to 6 documented semi-structured interviews make an acceptable corpus) or with public sector-wide data.
Your contract was terminated during the year. The rules for sitting the examination depend on the qualification framework and on your training centre, which must support you — in particular during the 6-month window in which you can remain in training after a termination while looking for a new employer. Tell your programme manager immediately: accommodations exist, but they're negotiated early. If you're in this situation, our article on terminating an apprenticeship contract sets out your rights and the procedure.
In the event of partial failure, qualifications structured in competency units allow you to retain the units you've passed and resit only the missing one at a later session. Check this option on the RNCP record for your qualification.
The 7-days-before checklist
- Deck finalised, exported as PDF, tested on another computer
- Speaking time measured and under the limit (aim for 90% of the allocated time)
- List of 10 likely questions + three-sentence answers
- Copies of the dissertation submitted as instructed by your training centre (format, date, number)
- Written approval from the company on the data being shared
- Journey and timings checked, 30 minutes' leeway
- A full night's sleep: no late-night revision has ever saved an oral
The defence isn't a dissertation exam, it's a professional interview in disguise. The panel isn't checking that you've read everything: they're checking that you can frame a problem, tackle it methodically and talk about it clearly to people who weren't there. Those three skills are exactly what your future hiring will be judged on. You may as well start working on them now — and, if the year ends well, move straight on to negotiating a permanent contract.