If you work in tech and you are thinking about your next role, you have probably heard the same things everywhere. We move fast. We innovate. We experiment. We are AI-first.
The more useful question is: what does innovation actually look like when deadlines are real and teams still need to deliver?
One useful example comes from Rotem Hazan, Data Engineer, and Yekaterina Kaluzhni, Product Manager, who were part of Teads Data Group AI hackathon. What stood out was not only what they built, but the kind of working environment that made it possible: collaborative, practical, and focused on solving something real.
It was not a large, highly polished initiative. It started small, and within a few weeks there were tools teams could use, time back in day-to-day work, and ideas that had moved beyond the concept stage into something concrete. From an AI-powered data mart creation tool to automated BigQuery access and cost cleanup, the projects focused on forms of invisible friction that quietly slow teams down.
Brainstorming, testing, and iteration all play a role, but they are only part of the process. Innovation becomes tangible when someone notices a problem, brings it up, tests something useful, and helps move it into the business. In most cases, that depends less on individual talent and more on whether the environment allows different kinds of people to contribute, build, and move ideas forward.
What tends to be true where innovation actually happens
1. Someone actually owns things
There is clarity on who is driving an initiative, which makes it easier for decisions to move and for work to progress without unnecessary friction.
In the hackathon, leads were defined early, which helped teams move faster and translate initial momentum into something usable.
Ownership is also not only about execution. It is about who gets to shape the direction of the work. Some people are invited to lead, while others are expected to support, and that distinction influences which ideas develop and which ones stay in the background.
2. You can start before everything is perfect
Many ideas do not move forward because people feel they need to be fully prepared before sharing them, with a complete plan and clear certainty.
In environments where innovation works well, people can begin with something partial, test it, and refine it as they go. This does not remove structure; it creates room for progress before everything is fully defined.
Not everyone approaches this in the same way. Some people think out loud and move quickly, while others take more time to reflect, test, or shape an idea before presenting it. Teams that recognize these differences tend to access a wider range of thinking earlier in the process.
3. People can actually say what is not working
The idea of psychological safety is often discussed in broad terms, but in practice it shows up in very simple ways.
Can people say that something is not working, that something does not make sense, or that something may be missing, and do so early enough to influence the outcome?
This becomes particularly relevant in areas like AI, product development, or experimentation, where uncertainty is part of the work.
In many teams, some ideas move faster not only because they are better, but because they are expressed in ways the organization already knows how to trust. Ideas that come with speed, confidence, and certainty often get traction more quickly. Others may be just as useful, but show up with more reflection, more testing, or less immediate polish. Over time, that shapes which ideas get airtime, which ones get developed, and which ones actually make it into the work.
4. There is structure, but not control
Innovation tends to be less effective at both extremes, either in complete openness or under heavy control.
The environments that sustain it combine direction with flexibility. There is a clear problem to solve, a defined timeframe, and regular check-ins that support progress rather than slow it down. There is also a willingness to adjust scope when needed.
The goal is not open-ended creativity, but useful outcomes.
Hackathons that lead to something concrete tend to follow this balance. They provide enough direction for teams to focus their energy while still allowing room to explore and iterate.
5. Leadership shows up in a real way
The presence of leadership is usually visible in how initiatives evolve.
When leaders engage with the work, ask questions, and help remove obstacles, projects tend to move with more clarity and consistency. When that presence is limited, the signal is also clear.
Support in this context is not about control, but about attention and continuity.
Leaders also influence how innovation is distributed within a team. They shape whose ideas are amplified, whose risks are supported, and who is seen as someone to invest in. Over time, these patterns affect not only outcomes, but also participation.
6. Different perspectives are not just “nice to have”
In practice, some of the most useful ideas come from identifying points of friction that have become normalized.
A junior team member noticing inefficiency. Someone in a different market identifying a blind spot. Someone questioning a process that works well for one group but not for others.
These inputs contribute directly to improving products, systems, and decisions.
In many cases, the challenge is not the absence of ideas, but how they move within the organization. Not all contributions are equally visible, supported, or developed.
If you are job hunting, pay attention to this
When you are interviewing, it helps to move beyond general statements and look for concrete examples.
- What is one idea your team tested recently that actually made it into the work?
- Can you give me an example of something that started small and became useful?
- How do you decide which early ideas are worth investing in?
- When people disagree on an idea early, how is that usually handled?
- How do you make sure useful ideas are not missed just because they come from quieter or less visible people?
The answers often provide a clearer picture of how work actually happens.
Why this matters more now
AI has made these dynamics more visible.
Many organizations are actively exploring new tools and approaches, but the outcomes still depend on how ideas are developed, shared, and implemented.
That includes:
- who contributes to the work
- how decisions are made
- whether different perspectives are considered early
- how ownership is distributed
- how ideas move from concept to execution
If the same types of voices consistently shape what gets built, the results will reflect that.
So when evaluating a role or a company, it is useful to look beyond whether innovation is part of the language, and instead consider whether the environment allows different kinds of people to shape what happens.
That is often where innovation becomes more consistent and more relevant.
If you want to hear how this actually played out
You can listen to the latest episode of Voices of AdTech here, where Rotem Hazan, Data Engineer, and Yekaterina Kaluzhni, Product Manager, walk through how the hackathon developed in practice, what triggered it, what needed adjustment along the way, and how ideas moved into something usable.