Publicis DACH is part of the globally active Publicis Group. With 2,300 employees, it’s one of the largest network agencies in the region. Its promise: all communication services from a single source. And that’s no exaggeration – Publicis covers everything that clients need for modern communication, from classic to digital.
As Chief Commercial Officer of Publicis, one of Norbert Knapp’s responsibilities is negotiating agency contracts. These include numerous non-disclosure agreements (NDA), which oblige all parties to treat the contents of the contract confidentially.
In this interview, Norbert Knapp tells us how his employees review NDAs themselves with the help of Legartis and therefore save the internal legal department a great deal of work.
At Publicis, we prefer to deploy our resources in the operational side of business, where they can generate immediate added value. However, there’s always a whole slew of contracts waiting for us in our day-to-day. This ties up enormous resources in the back office.
Like all market players, we’re under enormous cost pressure. We can't hire the amount of people we actually need for our contract review. This can lead to bottlenecks in our central legal department, which becomes especially problematic when we have the very tight deadlines often given to us by the client.
Contract reviews are complex processes. Before Legartis, they cost us more resources in the back office than we could and wanted to spend.
For this reason, we had already been considering using artificial intelligence to relieve the burden on the legal department and the back office for some time. We were just missing the right solution. When our Swiss CFO met the CEO and co-founder of Legartis, David Alain Bloch, and his product, it quickly became clear to us: this is it! Working with Legartis meant a new opportunity for us. And with the backing of our legal team, there was no problem introducing the tool at Publicis DACH.
In every operating company there’s a team member who has a Legartis licence. This means that each agency is able to use Legartis independently for itself.
To start, we concentrated completely on the NDA contract type. These contracts are relatively straightforward – and are subject to certain legal standards in terms of structure and wording. The NDA does not form the basis for our subsequent contractual relationship with our clients and we’ve never had any legal disputes with it, as the topic of confidentiality is honoured very meticulously at all our agencies anyway. But of course, it still has to be formulated correctly and in a legally secure way.
Every agency that exchanges confidential data with clients needs an NDA. Every year, the legal department handles around 350 of these contracts – plus the number of unreported cases! Because in the past, due to time constraints, some agencies definitely reviewed the NDA themselves and concluded them with potential new clients.
To avoid exactly this, we now have Legartis. The software both reduces the risk of NDAs not being reviewed and at the same time increases the quality of our NDA contract review.
In the past, the all-important question was: "How do I get an NDA that comes from the client reviewed by the legal department within a few hours?" Now, each agency can simply review it themselves and doesn't have to keep their clients waiting any longer.
The contract playbook, i.e. the compilation of our contract guidelines, was drawn up by the legal department and is based on our compliance specifications. Every NDA review is based on these guidelines. This gives us the security that in the end, a "clean" document is going out to the client.
Setting up the contract playbook was quick. It only took about a day in total. We received a good NDA template from Legartis, which we then customised and supplemented as needed.
The legal software itself has been very well received! Within a very short time, all users were trained and ready to go. In the first feedback round, all users reported that they were already successful in reviewing NDAs with Legartis. We are now gradually adapting the compliance guidelines and incorporating user feedback into the playbook. This way, we’re making the process even smoother for our agencies.
As far as the time required to review an NDA contract is concerned: the AI is fast. We only have to check individual wordings afterwards and follow the AI's instructions for action.
The users overall have great trust in the AI. What remains is a given uncertainty about formulations, especially in the case of special reviews; where does this clause fit? Can I make this change like this? What do I do with an NDA from the US, which looks completely different from the usual NDAs?
We’re currently training our users with less legal expertise so that they can better classify individual formulations and thus gain even more security during the contract review.
Legartis has generally been very well received and is in the process of establishing itself in our company. Three things I’ve learned during the initial period:
There’s simply no reason to wait. The whole topic of resources is too important to let digital or AI-based opportunities pass by just to be able to deploy resources in the operational side of business.
The routine task of contract review is better solved with AI than by cost-intensive resources in the back office. Every hour invested on a customer is better than spending it on back office or admin functions. The cost pressure is simply too great for that.
We save time and money without having to compromise on quality.
Often, the business user only has their pitch in mind and forgets legal issues. This can backfire. With Legartis, we create legal certainty and accelerate the pitch process at the same time.
I think the NDA contract review will have become business as usual by then and will run as a matter of course. We’ll know exactly what we can resolve ourselves and where it’s better to get our legal department involved.
Our legal department has no fears of being replaced by Legartis AI. We need our lawyers now and will always need them in the future.
After all, an AI like this is not static. In five years, Legartis AI will have already learned a lot about Publicis, such as specific preferences and processes. Maybe it will even become smarter than we are? I can already see the day when it will come up with a formulation suggestion that we ourselves would never have thought of!
The interview was conducted verbally.