Early career, evidence and the first professional steps

The assumption that it all begins after graduation is expensive, and it is expensive because it feels reasonable. Education is organised as a sequence of stages with a gate at the end of each one, and it is natural to expect the next stage to work the same way. Work has no gate. Nobody confirms that the transition has started, no letter arrives, and the absence of a signal is widely read as permission to postpone.

What happens instead is accumulation. Somebody spends a term running a small thing badly and then less badly. Somebody finishes a piece of work nobody asked for and puts it where it can be seen. Somebody holds a position in a society long enough to hand it over properly. None of that looks like a career while it is happening, and all of it is the difference between a final-year student who has something to point at and one who has a transcript and a set of intentions.

The two positions are not separated by ability, which is the part worth being clear about. They are separated by whether anything was produced and kept. A student who spent three years reading attentively and produced nothing external is not less capable; they are less visible, and the first stage of hiring runs almost entirely on visibility. The remedy is small and unglamorous, which is why it tends to be postponed rather than rejected.

A degree is a claim made about a person by an institution. It carries real weight, particularly where it operates as a threshold that applications are filtered against, and in some fields it is a formal requirement rather than a signal. What it cannot do is be specific. It records that a body of work was completed to a standard. It says nothing about what the person is like to work with, what they can produce without supervision, or what they do when a task arrives badly defined.

Evidence closes that gap by being external. It exists outside your account of yourself, which is exactly what makes it useful to somebody who has no reason yet to trust the account. A piece of work another person can open, a role that had a beginning and an end, a problem that changed shape because you were involved: each can be examined without taking anyone's word for anything, and each gives an interviewer something to ask about other than adjectives. Much of what a degree already produces can serve this purpose, and so can the roles and activities described on the student life pillar, provided something survives the end of term in a form that can be shown.

Evidence looks different in each kind of field, and the common mistake is to assume it always means a designed portfolio in a visual style. In most cases it means the ordinary artefact of that field, produced once, honestly attributed and available to look at without a request. The categories below are deliberately generic, because the specific form is set by the field itself and by what the people already working in it circulate among each other.

The word sounds like something a designer keeps and nobody else needs. In practice the question is narrower: when somebody becomes curious about you for a short moment, is there anything they can open. The failure mode is rarely having nothing. It is having a good deal and requiring an email, an attachment and a follow-up before any of it can be seen, by which point the curiosity has moved along to the next name on the list.

The second failure mode is presenting work without the problem attached. A screen, a document or a result on its own asks the viewer to reconstruct what was hard about it, and viewers do not reconstruct. Two or three sentences fix this: what the situation was, what had to be decided, what you did, and what happened afterwards. Where the work was collaborative, saying plainly which part was yours costs nothing and removes a doubt that would otherwise hang over the whole thing.

A portfolio does not need a custom domain, a visual identity or a section describing your philosophy. It needs to be legible to somebody skimming between two other tasks, and the properties below do most of that work. They apply equally to a single page, a shared folder, a public repository or a plain document, and the format matters far less than whether the thing opens. Choosing the format the field already uses saves an explanation.

The word is unhelpful because it describes a result and is then handed over as an instruction. Told to network, most people picture a room, a drink and an exchange of introductions, conclude they are bad at it, and stop. The relationships that actually produce openings form in a different way entirely: sideways, through work, over a period, and usually without either person thinking of it as networking while it is happening.

The mechanism is repeated contact around a shared task. Somebody who worked with you on a project knows what you are like when a deadline moves. Somebody who watched you run a society knows whether you follow up. Those are the people who offer a name when a colleague asks whether they know anyone, and offering a name carries a small risk for them, which is why it goes to people whose behaviour they have observed rather than to people who once handed over a contact detail at an event.

This is also why belonging and professional life are less separate than they appear. The groups a student already sits inside are where most of it begins, a point developed on the community pillar. The practical side of finding and building those groups is set out in the guide on why community matters for students. Keeping a relationship alive afterwards is mostly a matter of a short message when something relevant happens, sent with no request attached, which is the step most people skip.

An application is a matching process running under constraints, and the constraints are mostly invisible from outside. Roles get filled internally before they are advertised. A vacancy closes because a budget moved. A reader with a large pile spends a short time on each one and is looking for reasons to reduce the pile rather than reasons to keep it. None of that is a judgement about the person applying, and none of it appears anywhere in the reply.

The rate at which any single application converts is low in most fields, and it is low for the people who eventually get hired as well. That is worth stating plainly, because the arithmetic is what makes the process bearable. A sequence of attempts with a low per-attempt rate produces long runs of nothing, and a long run of nothing is the expected shape of that arithmetic rather than information about the applicant. Reading each rejection as information about your worth asks an ordinary distribution of outcomes to carry a meaning it cannot carry, since nothing in the process was measuring worth in the first place.

There is information in the volume, though it is narrower than it feels. Applications that never reach a conversation may be pointed at roles requiring something specific you have not shown, or may be missing the artefact that would let somebody check a claim quickly. Applications that reach conversations and stop there suggest something different. Keeping a plain record of what was sent, where and when turns an undifferentiated mass into something with a pattern in it, which is the only way that distinction becomes visible at all.

In the first months of paid work, the things colleagues notice are smaller than expected and almost none of them concern ability. Replying is the first. A message answered the same day, even with a holding line saying when a proper answer will arrive, marks somebody as a person who can hold a thread, and a team's tolerance for chasing is lower than most new arrivals assume it will be.

Finishing is the second. Work that reaches an end and gets handed over, including the parts that turned out badly, is rarer than work that is nearly done. The third is announcing lateness before it happens. A task that will miss its date is a problem; a task that misses its date silently is a larger one, because it removes everyone else's chance to rearrange anything around it. Saying so early costs a small amount of discomfort and buys a reputation that takes months to build in any other way.

Underneath all three sits the same behaviour: making yourself predictable to the people around you. Asking a question once rather than guessing twice belongs in the same category, as does writing down what was agreed after a conversation and sending it back. Early on, nobody can assess the quality of your judgement, so they assess the reliability of your conduct, and the two stay conflated long enough that the second becomes the foundation of the first.

Graduating with no destination in mind is common, and it is a workable starting position rather than a failure of preparation. Very few careers are chosen in advance; most are narrowed into. A person takes something available, discovers within months which parts of it they can tolerate and which they cannot, and moves in the direction that information points. That is a legitimate method, and it is often faster than deliberation because it runs on evidence rather than on imagination.

The risk in the position is not the absence of a plan. It is a year passing with no record of what was tried, so that the same uncertainty is still sitting there at the end of it with nothing added. A job that is not the destination still produces evidence, still builds relationships, and still teaches which conditions are workable and which are not. What it does not do by itself is get written down, and the writing down is what converts a year of employment into an account of a direction.

For students who studied outside their home country, what happens to permission to stay and work after a course ends is a separate question with its own timing, and it is set by that country's immigration authority rather than by an employer or an institution. The university's international office is the usual first stop, and the wider set of cross-border questions is covered on the study abroad pillar. Where the answer carries a deadline, it tends to fall earlier than people expect, which is an argument for asking well before the final term.