Becoming a Microsoft MVP

Most professional recognition in technology has a front door. An exam to sit, a form to submit, a fee to pay, a portfolio to upload. Effort goes in one end and a decision comes out the other, and the timing is largely under the candidate's control.

The Microsoft MVP Award has no front door. It cannot be applied for. That single structural fact determines everything else about it, and it is the fact most often skipped in write-ups that go straight to a list of tips.

The sequence is short and each step has a different owner.

Two consequences deserve to be sat with rather than skimmed.

The first step depends entirely on someone else. No amount of contribution triggers a referral automatically. Someone with standing has to notice, decide, and act — which means being visible in places where those people already are is not networking advice bolted onto the technical work. It is a structural requirement of the mechanism.

The award is a lease, not a purchase. Renewal depends on continued contribution. Anyone treating the award as a finish line has misread it; the same evaluation runs again every year against a fresh twelve months.

There is a nice piece of history sitting underneath all this. The program grew out of peer-support communities in the 1990s, and it was cancelled outright in October 1999 — then reinstated three days later after the community objected loudly enough to reach the top of the company. A recognition program that was once saved by the people it recognizes has a certain durability of purpose baked into it.

The award is not general. It is granted in a category aligned to a technology area — Azure, AI, Developer Technologies, Microsoft 365 and others — and each category has its own contribution areas beneath it. The population of MVPs runs into several thousand worldwide, spread across those categories.

This is where the mechanism rewards focus in a way that surprises people who expected breadth to help.

A year of scattered contribution across five product areas produces a thin record in each. The same effort concentrated in one category produces a record that a product group can recognize as sustained work in their area — and the product groups are part of the evaluation. Diffuse activity is not merely less efficient; it is harder for the process to see at all.

Choosing a category well is therefore a real decision, and the sharper version of it is choosing a lane inside a category. A crowded lane means competing for attention with many contributors covering the same surface. A sparse lane — a capability that is real, documented, shipping, and under-covered — means the contributions land in territory where they are genuinely useful rather than duplicative.

Concentrated contribution in one category is not narrower than broad contribution. It is the only kind the process is built to see.

Evaluation looks at the impact of activity over the previous twelve months. Not a career. Not potential. A rolling year.

That window has an underappreciated implication: it is unforgiving of bursts and kind to cadence. A concentrated month of output followed by ten quiet ones reads, at review time, as a thin year. Steady contribution across the same period reads as sustained presence, which is what the word impact is actually measuring.

The other implication is about evidence. A nomination form asks what was done, and the answer is a set of links. Which means the practical discipline is unglamorous and entirely mechanical: keep the record as it happens. A URL captured on the day costs nothing. The same URL reconstructed eleven months later — the discussion thread, the merged change, the recording, the answered question — costs an afternoon of archaeology and will be incomplete.

What tends to count, in rough order of how much work it represents:

• Code and artifacts that ship — merged contributions to relevant open-source repositories, published tooling, working samples other people use
• Answering other people's problems in the places where they ask — issue trackers, forums, community chat
• Published writing that explains something genuinely difficult, particularly where documentation is thin
• Speaking, from a user-group lightning talk upward
• Organizing — running the meetup, maintaining the resource, doing the work nobody gets a badge for

What does not count is easier to state: reactions, agreement, self-promotion, and volume without substance. The evaluation is about impact on a community, and a community can tell the difference immediately even when a metric cannot.

This deserves its own section because the confusion is common and expensive.

Certifications are not a path to the MVP award. They are a different instrument measuring a different thing. A certification attests that an individual has demonstrated defined skills against a defined objective domain. The MVP award recognizes impact on a community. One is about what a person knows; the other is about what a person has given away.

An engineer with a wall of badges and no public footprint is not a candidate. An engineer with no certifications and a year of substantial public contribution in a category is exactly a candidate.

None of which makes certifications pointless — they are frequently the right instrument for a different job, usually a role or an employer requirement, as the previous episode covers. The error is treating them as steps on the same ladder. They are separate ladders leaning against separate walls.

Given a mechanism where the first step belongs to somebody else, the review takes up to three months, the window is a rolling year, and the award has to be re-earned annually — what is a reasonable approach?

There is exactly one, and it is almost anticlimactic. Do work that would be worth doing if the award did not exist.

That is not a platitude dressed up as advice; it falls out of the structure. Contribution aimed at a nomination is legible as such, and the people whose referral is required have seen it many times. Contribution aimed at a real gap — a model that will not compile, a runtime behaviour nobody has written down, a question in a tracker that has sat unanswered for a fortnight — produces the same artifacts and is worth having produced regardless of what any committee decides.

The practical shape that follows:

• Pick one category, and one under-covered lane inside it. Depth is legible; breadth is noise.
• Set a cadence and hold it. A rolling twelve-month window rewards the ordinary month more than the spectacular one.
• Keep the evidence file as work happens. Links, dated, the day they land.
• Be present where the maintainers already are, and be useful there rather than visible there.
• Never lead with the ambition. Announcing a campaign for recognition is the fastest way to make every subsequent contribution read as instrumental.

The award is the lagging indicator. The lane, the cadence, and the record are the work — and they are worth having on their own terms, which is fortunate, because they are the only part anybody actually controls.