Competencies: how companies turn "good engineer" into a score

"Is this a good engineer?" is not a question five people can answer consistently. Companies solve that by breaking the question into named parts and scoring each one separately.

This lesson explains that decomposition in three steps. What a competency is and how it generates the questions you are asked. Which competencies appear at nearly every company under different names. And how to prepare for a company that publishes its own principles without sounding like you are reciting them.

What a competency is

The key word is observable. "Smart" is not a competency because two interviewers will apply it differently. "Sought out the reasoning behind a position they disagreed with" is, because two people watching the same answer will agree on whether it happened.

Why companies bother

Without a shared vocabulary, a loop produces five incomparable opinions:

"I liked them." · "Bit quiet." · "Strong on algorithms." · "Not sure they'd fit here." "Reminded me of Anya, who worked out well."

Every one of those is unfalsifiable, and the last one is how bias enters unchallenged. A competency framework forces each opinion into the form "on competency X, here is what they said, therefore this score" — which can be argued with.

The framework also spreads coverage. Without it, four interviewers all ask about the hardest technical problem and nobody asks about conflict. With it, the competencies are assigned in advance so the loop covers the ground once.

How the decomposition drives your interview

The competency list determines the questions. This is more mechanical than most candidates realise:

Competency: Earning Trust  Definition: builds working relationships that survive disagreement  Positive indicators:    - seeks the other party's reasoning before arguing    - separates the decision from the person    - commits fully after a decision goes against them  Negative indicators:    - characterises colleagues as unreasonable or incompetent    - avoids disagreement entirely    - relitigates a settled decision  Questions (pick 1-2):    - Tell me about a time you disagreed with a colleague.    - Tell me about a time you had to work with someone difficult.    - Tell me about feedback you disagreed with.

The interviewer picks a question, listens for the positive indicators, and writes down the negative ones. That is the whole mechanism.

What this means for your preparation

You are not preparing answers to questions. You are preparing evidence for indicators. The question is a doorway; several different questions open onto the same indicator list.

This is why the story bank works and the question list does not. Fourteen stories tagged against competencies cover hundreds of phrasings. Fourteen memorized answers cover fourteen phrasings.

The shared core across companies

If every company invented its own competencies from scratch, you would need a separate story bank per employer. They do not. Company competency frameworks look different and mostly are not. Five clusters appear nearly everywhere under different names.

Two more appear at senior levels and above:

The nine competency sections in this course (Sections 6–14) split these seven clusters slightly more finely, adding Customer Focus and Innovation, both of which appear frequently enough to deserve their own preparation.

Translating between vocabularies

When you get a company's competency list — recruiters often send it, and it is reasonable to ask — map it onto the clusters before preparing anything:

Their list (invented, but typical in shape): Customers First · Own the Outcome · Simplify Before You Build · Decide with Evidence · Stay Curious · Trust Is Earned · Understand the Details · Disagree, Then Commit

Maps to: Customer focus (Section 11) · Ownership (Section 6) · Innovation (Section 12) · Technical judgement (Section 8) · Growth (Section 10) · Collaboration (Section 9) · Technical judgement (Section 8) · Collaboration (Section 9)

Eight named principles, five underlying clusters, and two of them appear twice. Your existing story bank covers most of it. What you are adding is emphasis and phrasing, not new material.

Where companies genuinely differ

Three real differences, not merely naming:

Weighting. A company that ships consumer product weekly weights delivery and customer focus. A company building infrastructure weights technical judgement and long-horizon thinking. Ask the recruiter what the loop emphasises.

Threshold by level. Some employers expect strategic-thinking evidence at senior; others not until staff. This is the single most useful thing to ask about, and recruiters usually answer it plainly.

Culture-specific interpretations. "Ownership" at a company with strong team boundaries means owning your service well. At a company with weak boundaries it means picking up whatever is broken. The same story reads differently.

Company-specific value systems

Some companies go further: they publish their principles and interview against them explicitly. This changes your preparation less than candidates fear and more than they expect.

What "explicit principles" changes

Three things:

1. The questions are named. Instead of "tell me about a disagreement", you may get "tell me about a time you had backbone" or "give me an example of you being frugal". The vocabulary appears in the question itself.

2. Coverage is enforced. Each interviewer is assigned two or three principles, and the loop covers all of them. There is no principle nobody asks about.

3. The bar includes phrasing. Not that you must recite the principle, but that your story should visibly contain what the principle describes. If a principle is about acting on incomplete information, a story where you waited for full data scores badly even if the outcome was good.

How to prepare for one without becoming a recitation

The failure mode is real and interviewers name it in write-ups: "answers felt reverse-engineered from the principles".

It happens when candidates take a principle and go looking for a story to fit it. The story arrives thin, over-explained, and narrated in the company's vocabulary — "so this is where I showed real ownership" — which reads as performance.

Do the mapping in the other direction. Start from your real stories, then tag which principles each one demonstrates. You will usually find:

Four stories, ten principle tags. Now the gaps are visible, and you can go looking for one or two more stories rather than fourteen.

The vocabulary question

Should you use the company's words?

Once, lightly, if it is honest. "This is the one where I had to disagree and then commit" is a useful signpost that helps the note-taker. Saying it before every story sounds rehearsed.

Never as a substitute for evidence. "I showed a real bias for action here" is an assertion. "I shipped the fix behind a flag on day one and gathered the data for the proper decision over the following week" is the same claim with evidence attached, and it does not need the label.