The badge is scanned, the LinkedIn connection added, the conversation already forgotten. In the aisles of VivaTech, CES or the RAISE Summit, networking can quickly resemble a production line: introduce yourself, exchange a few words, move on. Yet encounters that lead to crucial advice, a hire or a project often begin differently. With a precise question. With a few minutes of attention. Then with a message that proves you genuinely listened.
Three events, the same trap
These events are not interchangeable. In Las Vegas, CES offers an overview of consumer technologies and their industrial applications. In Paris, VivaTech brings together startups, major corporations, investors and public-sector organizations. Also held in Paris, the RAISE Summit focuses on artificial intelligence and its economic implications. The people you meet, their expectations and their availability therefore vary considerably from one event to another.
The trap remains the same: confusing a packed schedule with quality relationships. Between demonstrations, talks and evening events, everything encourages you to speed up. A long conversation can feel like it comes at the cost of several potential contacts. But ten generic exchanges do not necessarily amount to ten budding relationships. They may simply generate ten notifications that nobody follows up on.
As the business calendar picks up again in September 2026, the methods proposed here draw on established characteristics of these events and known technology trends. The developments envisaged are forward-looking, rather than a factual assessment of their 2026 editions.
Before the event, decide what you really want
“Expand my network” is too vague an intention to guide a day. Seeking to understand how a company evaluates its AI vendors, meeting product managers facing the same problem or testing a business hypothesis provides a much more actionable direction. A useful objective describes what you want to learn or move forward, not just whom you want to approach.
Preparation then involves selecting a few relevant people or organizations. A conference agenda, a product page, a recent interview or a job posting may be enough to identify a starting point. There is no need to compile an encyclopedic dossier: you need to understand the context well enough to avoid asking a question whose answer appears in the first line of the website.
Still, leave some room in your schedule. An overly rigid list turns the event into a series of appointments and rules out unexpected discoveries. The right balance is to prepare for priority meetings while remaining available for an unplanned conversation, particularly with someone who has neither an impressive title nor a long queue outside their booth.
The right question opens a conversation
Since the rise of generative AI, the promises have sounded much the same: save time, automate, personalize. Asking “What do you do in AI?” invites a standard presentation. A question about a real trade-off is more likely to elicit an interesting answer: which use case held up under testing? What constraint delayed deployment? How do you distinguish a compelling demonstration from a tool people actually adopt?
The question must, however, suit the person. A researcher, a procurement manager and a founder do not speak from the same vantage point. It must also be possible to answer without disclosing confidential information. Here are three possible conversation starters, to adapt rather than recite:
- At a booth: “In what circumstances would you advise a customer not to choose your solution?”
- After a talk: “You mentioned this barrier to deployment: what helps you spot it early enough?”
- Among professionals: “What recent decision made you change your mind about this market?”
These questions are not tricks to make you sound brilliant. Their value lies in moving beyond promotional talk and creating common ground. That also means accepting an unexpected answer, rather than immediately steering the discussion back toward your own pitch.
Spend less time introducing yourself and more time listening
A short introduction is still necessary: what you do, the problem you are working on and why you are there. But it should provide context, not dominate the exchange. If the other person describes a recruitment challenge, there is no point automatically launching into every feature of your product. Instead, ask which skills are missing and what has already been tried.
Active listening shows through a simple action: paraphrasing. “If I understand correctly, the bottleneck has less to do with the model than with integrating it into teams’ workflows?” This check reduces misunderstandings and shows that you are trying to understand, not simply waiting for your turn. It can also reveal that no immediate collaboration makes sense, without making the encounter a failure.
Respecting the context matters as much as the content. At a busy booth or after a presentation, suggesting that you continue later is often better than holding someone up. And exchanging contact details does not constitute consent to receive an automated sales sequence. Ask which channel suits them and what specific follow-up they have in mind.
The follow-up starts before you part
The best antidote to the “Great to meet you” message is to agree on a next step during the conversation. Sending a resource, passing along a technical question, arranging a short discussion: the follow-up should be proportionate. Someone who has given you a few minutes owes you neither a recommendation nor an introduction to their management team.
Immediately afterward, note the context, the topic and any commitment made. Avoid personal judgments or private details with no professional relevance. These few lines are there to capture the exchange accurately, not to build a profiling dossier.
A personalized follow-up can be just four sentences: “We spoke after the panel discussion on AI adoption. Your comment about team training interested me. Here is the case study I offered to share. If the topic remains a priority, would you like to compare our approaches on a brief call?” What matters most is its continuity with the encounter, not the elegance of the writing.
Measure progress, not accumulation
After the event, counting scanned badges is reassuring but tells you little. A more useful assessment examines lessons learned, commitments honored and conversations actually continued. Some encounters will lead to swift action; others will remain an occasional source of information. Not all are meant to become business opportunities.
This approach also benefits people who feel less comfortable at large gatherings. Preparing a question, favoring small groups and allowing time for breaks makes it possible to participate without adopting an extroverted persona. Networking is not about commanding the whole room: it is about creating a few exchanges that give everyone a reason to reconnect.
What next? If AI tools make preparation and follow-up writing even easier, polished but interchangeable messages could proliferate. The differentiating factor would then shift toward what is harder to automate: an astute observation, genuine attention and a promise kept. At the next event, it may be better to leave with fewer contacts but more conversations worth continuing.
