Which sections should a CV have?

The question usually arrives like this: should I add a photo, should I write my hobbies, should I include references. The answer is not a matter of personal taste, it is set by what the side reading your CV can actually read. The GoTeam CV analysis engine carries ten fields out of a document: personal details, summary, experience, education, skills, languages, projects, certificates, achievements and publications. A field the schema does not recognise is not carried, it is dropped.

Whether the heading is in Turkish or English makes no difference; the extraction pipeline treats Deneyim and Experience, Projeler and Projects as the same section. What does make a difference is what is written under the heading. The engine awards no item for seeing the word Experience; the item is earned when a fact can be verified from the sentences beneath it.

The useful question is not which sections should be there, it is which section evidences what. The same fixed list of 49 items is applied to every CV, and 45 of them are computed directly in code. Every section below is tied to specific items on that list, and which one goes where is written down.

• Having the section is not enough. An item with no evidence does not count as half a point, it leaves the denominator; the absence of evidence is never counted against the candidate.
• The weights are within a dimension. A weight of 12 does not mean 12 out of the overall score, it means 12 units out of that dimension budget of 100 units.
• The same file always gets the same score. Reordering the sections does not move the result.

The table holds the ten fields the engine recognises. Each row writes the question that field answers and the item lost when the field stays empty. The numbers in brackets are item weights, and each dimension adds up to 100 on its own.

The table is not a list of advice. The rows do not say write this, they say if you do not write this, that question goes unanswered.

If an experience record has no start date, or the date cannot be read, that record cannot enter the duration calculation. The record is not ignored, the professional experience item is still earned. But because it is not added to the total duration, the three items looking at the 1, 2 and 4 year thresholds are affected.

If no month can be derived from any professional experience, the total duration comes back empty and all three threshold items are lost at once, with the reason that professional duration cannot be computed from the dates. That is how 28 weight units of the experience dimension become invisible. The cause is not a lack of experience, it is a date that cannot be read.

A range written with years only (2024 - 2026) is read, but the system pins it to the first month of the year and flags it with six months of uncertainty. Writing month and year earns its own item (10); having a start date on every record is a second item (6).

• Internships, volunteer work and personal projects do not count towards professional duration, they fall into separate classes. They do not need deleting; it is enough that they sit under a heading of their own, apart from professional experience.
• Part time roles enter the duration at half weight, freelance work counts as professional at full weight.
• Overlapping ranges are merged; the same month is never counted twice.

The skills section is the easiest part of a CV to lengthen, and that is exactly why it is the weakest evidence. The engine puts every skill into one of four classes: tied to a work record, tied to a project, tied to an education record, or list only. When a skill appears in more than one place, the strongest source is assigned.

Skills that stay list only are gathered under a separate heading in the report: claimed competencies. Next to it sits a single action sentence, write one line about a job where you used that skill. Lengthening the list does not solve this, it multiplies it.

The weights say the same thing. Listing a skill earns an item with a weight of 6, while the same skill appearing in an experience or project narrative is a separate item with a weight of 14. On top of that there is a 12 weight item asking that at least half of the listed skills appear in usage: inflate the list and that ratio drops, and the item is lost.

The education section produces not one item but nine, and all nine are deterministic: is there a record, is the degree name written, is it bachelor level or above, is there a postgraduate degree, is the department name written, is the school name written, is the year written, is the grade average written, was the degree completed.

The practical outcome: a line reading A University, 2019 - 2023 earns the school and year items and loses the department item (12). Adding Computer Engineering, BSc to the same line satisfies two more.

The grade average item (12) only looks at whether it is written, not at its value. The common advice is to hide a low average; the behaviour measured here contradicts it. The decision is yours, but an average left unwritten does not close that item.

The overall score is gathered not from one list but from five dimensions, and the dimension weights are the same for everyone. How much a section affects the overall score depends on two things: which dimensions it feeds and how much that dimension weighs.

A section appearing on more than one row is not a mistake. Certificates feed both the technical and the potential dimension, and so do projects; that is why the same record can be counted in two separate items.

• A dimension score is the ratio of the weight earned in that dimension to the weight still in the denominator. The denominator is not fixed: an item with no evidence leaves it.
• 45 of the 49 items are computed in code. Four items go to the model: growth in scope across roles, owning a piece of work end to end, leadership, and concrete impact. All four sit in the experience and potential dimensions.
• The result arrives with a band label: 85 and above Outstanding CV, 70 Strong CV, 55 Room to Grow, 40 Foundations in Place, below that Needs Rebuilding.

Some sections common in Turkish CV templates have no counterpart in this engine. A photo, date of birth, marital status, military service status, nationality, references, hobbies and interests are not in the extraction schema. The schema drops a field it does not recognise instead of carrying it, so these sections neither earn nor cost an item.

That does not add up to do not write them. What it adds up to is this: these sections take up space on the page and contribute nothing to the measured items. If your space is limited, what to put in place of what is no longer guesswork, it is a choice with a written consequence.

There is one more limit, and it is about the file rather than the score: on the candidate side only a PDF is accepted, at most 5 MB and by default 15 pages. A Word file is exported as PDF and then uploaded. The document passes an integrity gate first, so a file that is not a CV stops at the gate.

• The effect of a photo on the score has not been measured. The only thing that can be said is that a photo is not in the extraction schema and enters no evidence item.
• We have no item about how many pages a CV should be. The only rule that looks at page count is technical: a PDF longer than 15 pages stops at the gate.
• Reference details are not in the schema either. A line reading references available on request satisfies no item.