May a candidate do a trial working day unpaid?

Only as long as he is not working. What matters is not the label but how it is actually carried out: anyone who works subject to instructions, on someone else's terms and in personal dependency is in an employment relationship under § 611a (1) BGB — expressly regardless of what the whole thing was called in the contract. Then § 612 (1) BGB applies, remuneration is deemed to have been tacitly agreed, and the minimum wage is the floor. Merely watching and tagging along, with no obligation to work and no revenue responsibility of one's own, is something different — the line is narrow, and in practice it runs through what happens, not through what is on paper.

How long should a trial working day in sales be?

One day is enough if it is built properly: two to three real customer contacts, a preparation phase and a debrief phase. Shorter and you only see nervousness, longer and you see nothing further while taking on effort and legal risk. Anyone who needs several days actually needs a fixed-term probationary period — and § 622 (3) BGB with its two-week notice period provides the right framework for that.

What if the candidate performs badly on the trial day but is good?

That is the most common objection, and it has a true core: an unfamiliar product, an unfamiliar market and an observer at your back are not a fair starting position. That is why the trial day does not assess the result but the behaviour — preparation, questions, handling a no, the debrief. Someone who says afterwards what he would do differently next time has shown more than someone who happened to rescue an appointment.

Hiring salespeople: a trial working day says more than three interviews

The short answer

A hiring interview with a salesperson is a sales conversation in which the applicant is the product. Precisely the ability you want to test is being used against you. The trial working day turns that around: the candidate comes along to real customers, prepares, leads or accompanies, follows up. What is assessed is the behaviour, not the result. Legally, the rule is: the moment the candidate works subject to instructions, it is an employment relationship with everything that comes with it — the cleanest route is simply to pay for the day.

Why interviews measure especially badly with salespeople

For every role an interview measures only indirectly. In sales it measures the wrong thing directly. The core competence is establishing trust in a first-time situation and staying calm under pressure — and the hiring interview is itself a first-time situation under pressure. Whoever wins it has proven that he can win it. That says little about the work with customers.

On top of that come three distortions I see again and again in selection rounds. The narrated success cannot be checked: nobody can recalculate the last employer's numbers, and whoever brings them brings the good ones. The likeability trap: salespeople are hired for likeability and dismissed for results — by the same people, six months apart. The self-description bonus: asked how they handle rejection, every applicant answers correctly. The answer is part of the craft, not proof of it.

What a trial day measures — and what it does not

Observable on the trial day	Not observable
Preparation: what does he know about the customer before he walks in?	Resilience over months
Question quality and listening behaviour in the conversation	Being a good colleague in the team
Reaction to a clear no	Administrative discipline (CRM, follow-through)
Debrief: what does he write down, what does he notice?	Loyalty, handling a dry spell
Self-assessment after the day	Depth of expertise in a product that needs explaining

The right-hand column matters more than it looks. A trial day does not replace a probationary period — it only decides whether there will be one.

The format that proves something

Briefing, 30 minutes. Product, price, typical objection, goal for the day. No training — the candidate should go into the conversation incompletely informed, because that is the normal case.
First appointment: he listens. The manager or a colleague leads, the candidate takes notes. The reference point for everything that follows.
Second appointment: he leads the opening. Greeting, situation questions, handover. No more. Anyone who realises here that he does not know the need has realised something right.
Third appointment: he leads, someone sits alongside. Intervene only if the customer risks harm — not when it gets uncomfortable.
Debrief, 45 minutes. Three questions in this order: “What happened in the third appointment?” — “What would you do differently next time?” — “What would have to be in place here for you still to be around in two years?”

The third question reveals the most. It takes the candidate out of the applicant role and forces him to formulate a condition. Anyone who has none has either not thought about it or is not saying what he thinks.

Assessment: four criteria, fixed before the day

A trial day without criteria fixed in advance is an expensive gut feeling with witnesses. Four axes are enough, each in three levels and in writing: preparation, conversation management, handling resistance, self-assessment. The order matters: each observer first notes for himself, then they compare. If people talk first, the second one is assessing the first as well.

What does not belong on the form: whether the candidate sold anything that day. The close depends on product, price and customer — on everything he cannot help.

For the conversation management itself the same grid applies as in the first meeting: anyone who talks about the product straight away in the trial appointment will do so when it counts, too.

The legal line: when getting to know each other becomes work

This is where it gets serious, and where small businesses make the expensive mistakes. The yardstick is § 611a (1) BGB: anyone obliged “to perform work subject to instructions and determined by others in personal dependency” is an employee. Sentence 6 of the provision expressly closes the back door: “If the actual performance of the contractual relationship shows that it is an employment relationship, the designation in the contract is irrelevant.” A sheet of paper headed “non-binding taster” therefore does not help if the candidate spends half a day working through appointments.

If an employment relationship exists, the rest follows automatically: under § 612 (1) BGB remuneration is deemed to have been tacitly agreed if the service “is only to be expected in return for remuneration in the circumstances”. The floor is set by the statutory minimum wage — since 1 January 2026, 13.90 euros gross per hour, from 1 January 2027, 14.60 euros (§ 1 MiLoV5).

Three points that usually get lost:

Accident insurance. § 2 (2) sentence 1 SGB VII also covers people who “work as insured persons within the meaning of subsection 1 no. 1”. Whether a specific trial day falls under this is decided in a dispute by the Berufsgenossenschaft (statutory accident insurance institution) — which is why the question belongs settled beforehand and not after the rear-end collision on the way to the second appointment.
AGG. The rejection after the trial day is a selection decision like any other. § 15 (2) AGG (German General Equal Treatment Act) caps compensation for a non-hiring at three months' salary where the applicant would not have been hired even under a discrimination-free selection; under § 15 (4) AGG it has to be asserted in writing within two months. The completed assessment form is then the best document you can have — provided it contains nothing that has nothing to do with suitability.
Data protection. Under § 26 (8) sentence 2 BDSG applicants count as employees. Assessment forms are applicant data: purpose-bound, limited to what is necessary, with a deletion deadline.

The way out of this whole question of classification is unspectacular: treat the day as a paid working day, in writing, with an hourly rate. That costs less than any hour you later spend on the question of what it actually was.

What the day costs

The following calculation is a model calculation with freely chosen figures, not a measurement:

Item	Assumption
Candidate remuneration for the trial day (8 hrs)	€200
Accompaniment by a manager	1 day
Total per candidate, with 3 candidates	around €600 plus 3 manager-days
Bad hire: fixed salary for 6 months at €3,500	€21,000
plus onboarding, burned territories, replacement hire	not quantified, but not zero

The point is the ratio, not the figure. Three trial days cost roughly one per cent of what six months of a bad hire costs — and they cost it beforehand.

Action list

Set out an assessment form with four criteria and three levels in writing before the first candidate arrives. One page is enough.
Set the day up as a paid working day: a short written agreement, an hourly rate at least at minimum wage level, date, scope.
Before the day, clarify with the Berufsgenossenschaft how the candidate is insured — and write the answer down.
Inform customers in advance that an applicant is coming along.
Schedule three real appointments, not three role plays. Role plays measure role-play ability.
After the day, assess separately, and only then talk it through together.
Give the rejection within 48 hours, reasoned against the four criteria. Salespeople know one another, and rejections get talked about more than offers.