Skip to content
Perspectives

StandpointStandard Terms

Standard terms or individually negotiated?

An individually negotiated clause takes more than negotiating. Why hardly any company gets there, and what actually works.

Peter Poleacov15 min

Yellow and black road sign at the roadside with arrows pointing both ways, desert and mountains behind it

German courts review pre-formulated contract terms strictly in the event of a dispute, between businesses too; Swiss law, for example, allows more there.1 An individually negotiated clause is not subject to that review. Only the general limits remain: statutory prohibitions, immorality, and good faith.2 That difference decides whether your clause holds. And it hangs on one word.

The future operator of a waste incineration plant offers a customer a contract running 17 years. The customer is to deliver a fixed quantity of waste every quarter. If it delivers less and does not make up the shortfall in the first month of the following quarter, it pays for the full quantity anyway. The operator declares this clause non-negotiable, because the financing of its plant depends on it. Two companies negotiate this contract for more than seven months. The customer objects to the clause, obtains a shorter term, has the contract intensively reviewed by lawyers, and signs with the clause in it. For the first year of deliveries the operator invoices around 710,000 euros for shortfall quantities.

The clause gets it nothing. The clause is a standard term, and as a standard term it is invalid (BGH, judgment of 22 November 2012 – VII ZR 222/12). The customer had signed, and it was still free to rely on the invalidity. That it had accepted the clause after lengthy discussion does not by itself make that bad faith.3 The operator had negotiated for months. It just had not negotiated the clause individually.

Three years later, the same panel of the court, again a waste contract, again a payment for shortfall quantities. This time the customer had achieved something in the negotiations: instead of 115 euros, only 30 euros per missing ton. The court told the court of appeal that this does not suffice either. Whoever softens a clause but never puts its core up for negotiation has not negotiated it individually (BGH, judgment of 22 October 2015 – VII ZR 58/14).

Both times the other party had a say, both times it obtained concessions, and both times two companies will have weighed the contract as a whole, the way any merchant does. And both times the one thing that counts was missing: the opportunity to get rid of the clause itself.

Negotiating is not the same as negotiating individually

The statute takes a term out of standard-terms control to the extent it has been individually negotiated (Section 305(1) sentence 3 of the German Civil Code (Bürgerliches Gesetzbuch, BGB)). What that means, the Federal Court of Justice has laid down in a formula it has repeated for decades.4

In plain words: negotiating individually means more than merely negotiating. It requires that the user of the terms seriously puts up for negotiation the core of the clause that departs from the statute, and grants the other party the freedom to shape it in its own interest, with at least an effective opportunity to influence the substance of the terms. The user has to declare clearly and seriously that it is prepared to change individual clauses.

Three things follow from that, and each is overlooked in practice.

The test applies to the individual clause. The statute says “individually” and “to the extent”. A hard-fought provision does not rub off on its neighbor. The parties may have assessed their economic position as one package, the Federal Court of Justice says; that does not justify treating a clause that was not specifically negotiated and was adopted unchanged as individually negotiated. Whoever has fought over price and term has not thereby negotiated the liability clause.4

The user has to be able to give up the core. Not the wording, not the amount, but the point at which the clause departs from the statute. For a limitation of liability, that is the limitation itself.

And the user has to prove it. It bears the burden of pleading and proof for each of these circumstances.4 What it cannot substantiate did not happen, as far as the court is concerned.

Four routes that lead nowhere

The case law since 2012 reads like a catalog of attempts to get around the hurdle. Each fails at one of three gates. Four attempts are typical; the choice between two variants and the missing records follow below in the line of thinking.

The cover letter. The user sends the contract text ready for signature and asks the other party to communicate comments or change requests. That does not take away its status as the user of the terms. At most, it signals a certain willingness to negotiate.5 And a generally declared willingness to change something is not an individual negotiation of the one clause that later matters.4

The softening. The customer negotiates the lump sum down to 30 euros per missing ton, a fraction of the fee. That is not enough as long as the user has not itself put up for negotiation the principle of paying without fault and without loss.6

The review by the other party. The customer criticizes the clause, has the contract reviewed by lawyers, and signs. The court sees no indication in that of the customer having abandoned its objections and adopted the clause as its own.7 What the other party does is not decisive. What the user gives up is.

The confirmation. The other party confirms in the negotiation record that every clause “was negotiated extensively and seriously”. The court does not let that sentence alone establish individual negotiation. Sections 305 et seq. BGB are mandatory law, between businesses too.8 Why such clauses keep coming back anyway is covered in the article “No individual agreement”.

What the courts demand is a costly signal

There is a mental model that reduces this case law to one sentence: the costly signal. A signal is credible only if it costs the sender something. A cover letter costs nothing. A confirmation clause costs nothing. Even a discount on the amount costs little as long as the principle stays. That is why the courts do not count them.

What counts is the offer that hurts. For a limitation of liability, the user would have to offer unlimited statutory liability as the alternative.9 In plant engineering, in the supply industry, in IT outsourcing that is not a serious option, and whoever did it would have to justify their diligence internally. That is exactly where the hurdle lies, and it is intentional. The statute is meant to catch the unilaterally imposed term, and a term its user is unwilling to give up remains unilaterally imposed, however much the parties talk about it.

That is the user’s dilemma, whatever the clause. For a supplier it is usually the limitation of liability, for a buyer more likely the contractual penalty or the long payment term. The user needs the clause, which is why the clause is in its terms. To take it out of standard-terms control, it would have to be genuinely prepared to lose it, and it cannot be, because it needs it. Add to that who sits at the table: the sales or purchasing employee has instructions, the external lawyer has a mandate, and as a rule neither may release their company’s or their client’s clause on their own authority. Whoever wants to put a clause genuinely up for negotiation has to decide on the loss beforehand, at the level where that decision may be taken.

Why practice does not notice

Businesses see it differently, and measurably so. In a 2014 survey of 1,220 company employees for the German Federal Ministry of Justice, 32.0 percent considered an agreement signed by both parties, stating that a limitation of liability was individually negotiated, to be sufficient. That is precisely the route the Federal Court of Justice has closed. Intensive contract negotiations were considered sufficient by 28.3 percent, and by 40.6 and 47.4 percent in the two groups of large companies.10 Those two groups comprise only 95 and 115 responses. The figures are an indication, not a measurement: the larger the company, the firmer the belief that negotiating is enough.

Where the error comes from is explained by Lars Leuschner, the author of the study: most disputes between businesses are heard by arbitral tribunals, and those do not publish their awards.11 What happens there stays invisible. What stays visible is only that it has gone well so far.

The counter-position: the courts are out of touch with reality

The strongest criticism comes not from academia but from an arbitral award. A machine manufacturer had limited its liability to five percent of the contract value and had presented the clause five times, in quotations, confirmations, and the specification. The buyer called it very harsh, asked for the reason, was told it could not be changed, and placed the order. An ICC tribunal held the clause to be individually negotiated: both were experienced companies, the seller was admittedly the only supplier of the machine, but that did not deprive the buyer of its freedom to decide, the clause had been a subject of the discussions, and the buyer had had a real opportunity to make a counter-proposal, which it did not use. The tribunal relied on commercial reality: in long contract documents, it said, by no means every clause is discussed, and merchants have to assume that every clause can be talked about.12

Paul Hobeck, then general counsel of an industrial group, named precisely this dilemma in his note on the award and drew the opposite conclusion from it: because the supplier cannot give up the liability clause it would have to put up for negotiation, the courts must relax their excessive requirements.1 Thomas Pfeiffer calls the Federal Court of Justice’s test a formal examination that does not sufficiently take into account the purpose of standard-terms control: whoever reads a processing fee of 44,520 euros in the contract and signs has adopted it into their contractual will, and nothing more should be needed.13

Both describe how businesses actually negotiate, and there is something to both. Only the state courts have not moved. The award dates from 2001, the decisions of the Federal Court of Justice above from 2012 to 2019, and Pfeiffer writes his criticism as a note on precisely the order that repeats the formula above. Before an arbitral tribunal the answer can therefore differ from the answer before a regional court. Whoever relies on that is betting on the forum, and the bet is decided by the arbitration clause, not by the negotiation.

Two doors are nonetheless open, and both are narrow. At most in special circumstances can a clause count as individually negotiated even where the text stays unchanged after thorough discussion.7 The court of appeal in the first waste case had given the trade-off elsewhere in the contract as an example, and the Federal Court of Justice let that stand; the operator simply did not get through, because it pleaded nothing to that effect. And Pfeiffer reads the case law to the effect that changes on central points can support the inference of a willingness to change other clauses that stayed unchanged.13 I would not build a contract on either.

What actually works

From all this follows a line of thinking that does not start with negotiating.

The first question is: who provides the text? Whoever presents the draft is the user of the terms and carries the burden of individual negotiation. If the other party presents the draft and a clause you need is missing from it, say the limitation of liability, ask the other party for a proposed wording. Then the other party has provided the text, not you. That is not certain, because what matters is not who drafted the clause but to whom it is attributable.5 Whether standard clauses of your own that you insert into the other party’s draft are themselves standard terms, the courts had not decided as of 2022. Johannes Teichmann and Daniel Bunsen say no, because the responding party imposes nothing but replies to an imposed draft, and both then negotiate.14 That is a well-reasoned view, but not a judgment.

The second question is: which clause do you genuinely put up for negotiation? The clauses whose loss you could bear, you name expressly, decided by whoever may decide that in your company, and you record it in writing.

With this letter you start; it proves nothing yet. A generally declared willingness to change clauses is precisely what the Federal Court of Justice does not let suffice.4 That is why the letter carries no general clause but the list: the clauses you name, and for each of them the willingness to delete it or to replace it with the other party’s wording, even if that costs you something. According to Pfeiffer, you may make the change conditional on concessions elsewhere.13 Offering two variants of your own to choose from is, as a rule, not enough. The other party must be able to bring in its own wording and push it through.15

The third question is: can you prove it? The burden of proof is yours. The negotiation record, the draft versions with their changes, the other party’s reply to your offer: that is the file which decides the clause in court. As a rule, individual negotiation shows up in changes to the text, and an unchanged text is the hardest case.4 Leuschner says practice meets the requirements at most in exceptional cases; Kappus, that the attempt regularly fails.16 From my own practice I know of no case in which the user of the terms has furnished this proof. If it pleads anything at all, it pleads in general terms: the contract was negotiated at length. Concrete pleading on the negotiation steps over the disputed clause is missing, and that is exactly what the courts then hold against it.47

How far an individually negotiated agreement reaches is shown by an order of November 2025. An individually negotiated framework supply agreement took priority over the standard terms, and it did so even vis-à-vis the affiliated companies to which it granted rights, because the group company that had concluded it had negotiated their interests as well.17

And because the route usually fails, the rule that comes before everything else applies: build every clause so that it also holds as a standard term. The individually negotiated agreement is the bonus, not the plan. What the fallback to the statute costs when it does not hold is worked out in the article “Why your limitation of liability in B2B contracts fails”.

The clauses that merely assert an individual agreement are covered in the article “No individual agreement”. The exit via choice of law and arbitration clause is the subject of a later article in this series.

Plain words

Under German law the individually negotiated agreement is not the result of a negotiation. It is a piece of evidence. It does not come into being because two companies talked for a long time, but because one of them can show that it was prepared to lose its clause.

So I do not advise negotiating contracts harder. I advise deciding beforehand which two or three clauses you really put at stake, and then doing it in a way a court can still read three years later. Everything else is negotiating. That is not nothing. It is just not negotiating individually.

Notes

  1. Hobeck, note, SchiedsVZ 2005, 108 (112). ↩1 2

  2. Kappus, NJW 2016, 33 (33). ↩

  3. Ulmer/Brandner/Hensen/Fuchs, AGB-Recht, 13th ed. 2022, Vor § 307 Rn. 65. ↩

  4. BGH, order of 19 March 2019 – XI ZR 9/18. ↩1 2 3 4 5 6 7

  5. BGH, judgment of 20 January 2016 – VIII ZR 26/15. ↩1 2

  6. BGH, judgment of 22 October 2015 – VII ZR 58/14. ↩

  7. BGH, judgment of 22 November 2012 – VII ZR 222/12. ↩1 2 3

  8. BGH, judgment of 20 March 2014 – VII ZR 248/13. ↩

  9. Leuschner, NJW 2016, 1222 (1223). ↩

  10. Leuschner/Meyer, AGB-Recht für Verträge zwischen Unternehmern, final report for the German Federal Ministry of Justice and Consumer Protection, 2014, pp. 163, 168 f. and p. 277 (question 29). ↩

  11. Leuschner, NJW 2016, 1222 (1225). ↩

  12. ICC International Court of Arbitration, interim and partial award of 29 January 2001 – Case No. 10279, SchiedsVZ 2005, 108 (110–112). ↩

  13. Pfeiffer, note on BGH, order of 19 March 2019 – XI ZR 9/18, NJW 2019, 2080 (2082 f.). ↩1 2 3

  14. Teichmann/Bunsen, ZVertriebsR 2022, 287 (287, 291 f.). ↩

  15. BGH, judgment of 13 March 2018 – XI ZR 291/16, decided on two loan variants offered to a consumer; the standard under Section 305(1) sentence 3 BGB is the same between businesses. ↩

  16. Leuschner, NJW 2016, 1222 (1223); Kappus, NJW 2016, 33 (33). ↩

  17. BGH, order of 20 November 2025 – I ZB 9/25. ↩

Reference: Poleacov, P. (2026). Standard terms or individually negotiated?. INN.LAW. https://inn.law/en/perspectives/negotiating-standard-terms/