At NextWork, we help people learn AI skills by building real projects and showcasing it through a portfolio they can share. Our operating principles describe how we work together to make that experience better. They grew out of situations we kept encountering - work that needed another review, an idea waiting for someone’s approval, or a conversation that ended before anyone said what they really thought.
There are nine of these principles that we’ve printed and put up around the office. Occasionally they move closer to someone’s desk because that person is working on a particular principle. You’ll hear us bring them up in conversations too - we need the reminders most when following a principle asks more of us than letting something slide.
1. Learners come first, always
NextWork exists to solve a problem for people who want to learn. That gives us a way to judge the decisions we make: will this improve their experience? It is easy to lose sight of that when we are deep in product work or trying to move a growth metric. We might find a way to get more people to click a button, for example, while making the experience around it worse. A better number would not make that a good trade.
Spending time with learners keeps us connected to what we are here to do. We read their requests in our Discord community, talk to them about what they need and look at the projects they share. There was real excitement at seeing what people were building, and at finding a request we could finally mark complete.
That connection matters for the team as well. When you have been absorbed in your own work and start to lose a sense of its purpose, seeing someone use NextWork to build something brings it back. There is a person on the other side of the thing you are working on.
2. High standards are an act of care for our learners
This principle comes from a familiar moment in a review. Something has already gone back and forth five or six times. You still think it could be better, but you are wondering whether to bring it up again. Everyone has already put so much work into it. Another conversation would be uncomfortable, and it would be easier to let it go.
The learner will experience whatever we decide is good enough. If we notice a problem and stay quiet, we are leaving it for the learner to encounter.
That is why we expect people to give feedback beyond their own role. You can question a design without being the designer, or notice a problem in a project you did not write. It can feel awkward to challenge someone who knows more about that work, especially when they care deeply about it. Caring about the learner gives us a reason to have the conversation anyway.
It also gives us a way to check our own motives. We should be able to explain how the change helps the learner. Pushing for a higher standard takes care; insisting on our own preference simply because we want control does not.
3. Truth over harmony
We want to know when something is wrong, even when the conversation would be easier if nobody mentioned it.
This goes both ways with data. We need to look at the numbers when they challenge what we believe. We also need to question what we are measuring when the numbers look good but people are having a poor experience. A favourable metric cannot settle a concern about something it does not capture.
The point of raising a difficult question is to help us see the work truthfully. That requires being willing to hear an answer we do not like, including when it means changing our own mind.
4. Nothing is someone else’s problem
We want everyone at NextWork to feel proud of the whole product. That includes the projects, the app, the community, what we put out on social media, the events we run - everything. Our responsibilities overlap because the experience we are creating belongs to all of us.
Giving feedback is a good first step, and it is tempting to consider your part finished once you have told someone. But the problem may still be there. People are covering a lot of work, and a report can get lost among other priorities even when everyone involved cares. Checking back and offering help makes it more likely that something actually changes. You do not need to be able to fix every problem yourself to take responsibility for helping it get resolved.
5. Leaders don’t wait to be told
We want people who see something worth doing and get started. There will often be no one handing you an assignment, and we have limited time to learn what works. If an idea has been sitting with you for three months while you wait for someone to ask, we have lost three months in which we could have tried it.
This can take time to get comfortable with. Acting without full approval can feel risky, especially when you are new. You can talk it through with someone on the team and get a sense check - every question does not need to wait for Amber. We would rather have an attempt we can learn from, including one that needs changing, than an idea nobody acted on.
Keep the team in the loop as you go. Knowing that someone has tried something, seen a result and brought it back to the team is exciting. Initiative works better when the people around you can understand what you are doing and help.
6. Discomfort is a sign that it matters
You have spoken to a learner and understood what they were asking for. The team has built a solution, but you can see that it does not quite address the request. Now you have to tell people who have already done the work that something is still missing. That hesitation is exactly the kind of moment this principle is meant to help with.
We also felt it when people outside engineering started contributing to the codebase with AI tools. Engineers had concerns about quality and security. The people beginning to contribute wanted the codebase to better support how they were working. Those conversations went on for weeks. There were good points on both sides, and continuing to work through them helped us find a way of working that the team felt good about.
It would have been easy to back away because the discussion was difficult. We would also have left important questions unresolved. As the team grows, we will keep having to change how we work together. Small concerns need somewhere to go before they accumulate into frustration that seems to come out of nowhere.
7. Own the outcome, not the task
When something has taken a lot of effort and the result is disappointing, it is natural to explain how hard you worked. You put in the hours, gave it your attention and wanted it to succeed. Having someone question the result can feel as though they are dismissing all of that.
The effort can be real and the work can still have failed to do what we needed. We have to be able to discuss both honestly. Sometimes a small change makes a big difference; sometimes something we care deeply about takes months and gets us very little. Continuing because we have already invested so much can cost us even more. We need to keep looking up from what we are doing to see whether we are getting there. That can mean changing the task or reconsidering the goal itself.
We are here because we care about the problem NextWork is solving. We need enough commitment to that problem to change our approach when the current one is not working.
8. Say it now, not at retro
If you save a piece of feedback until the end of the week, you may have taken away a whole week in which someone could have done things better. They have carried on without information you already had. By the time you raise it, the chance to improve that particular piece of work may have passed.
It can feel disruptive to bring something up when you know a teammate has a full day planned. But waiting has a cost too. Give them the chance to understand the issue and respond while they can still do something about it.
Weekly retros help us look for patterns across the week. They work better when we have already been talking to each other, so we can spend the time understanding what keeps happening and how to improve it.
9. Commitment = conflict + clarity
People can understand a plan and still privately believe it is a bad idea. If nobody says so, the conversation looks settled. The doubt stays with them while they do the work, and may only come out later, when something goes wrong. By then, the team has lost the chance to consider a concern that existed from the beginning.
That is why commitment needs conflict. We need to put our concerns into the conversation while they can influence the decision. Otherwise, we are asking people to commit with parts of their reasoning still unspoken.
We also need clarity. A discussion can bring every disagreement into the open and still leave people unsure what they are meant to do next. Before we move on, we need to understand what we are doing, why we have chosen that direction and what happens next.
Clarity does not require everyone to prefer the same option. It means we have had the discussion and know the direction we are taking together. If you are leading a piece of work, make room for people to challenge it, then stay in the conversation long enough to reach that shared understanding. That gives the decision a much better chance of being carried through.
