Beyond the CV, photo and job title: 10 signals for comparing specialists

In A Skills-First Labour Market, the OECD describes talent recognition approaches in which candidates are assessed and selected on the basis of demonstrated skills. Methods include structured assessments and competency-based interviews. [1]

Research by Sackett, Zhang, Berry and Lievens on employee selection methods found that, in their updated analyses, job-specific methods such as structured interviews, job knowledge tests and work samples ranked highly. The authors also explicitly warn against treating any one tool as universally best in every situation. [2]

This research concerns recruitment, not the specialist services marketplace directly. Its coefficients and rankings should not be transferred one-to-one to choosing a service provider.

The common principle is still useful: the more information relates to the actual task and the behaviour needed to perform it, the less you need to rely only on indirect signals of competence.

In World Bank-financed procurement, non-price criteria can include quality of methodology and work plan, risk management, past performance, delivery capacity and key personnel. Criteria should be tailored to the specific project, and bidders should provide evidence supporting their claims. [3]

The U.S. FAR likewise shows that the relative importance of price can vary with the nature of the purchase. When requirements are less definitive, more development work is needed or the risk of unsuccessful performance is higher, technical and past-performance considerations may become more important. [4]

Again, these are formal public procurement systems, not instructions for a private client selecting a specialist.

They nevertheless demonstrate an important way of thinking: price is one criterion. It does not have to replace the assessment of whether someone understands the problem and can realistically solve it.

Anonymous or partially anonymous first-stage assessment can reduce exposure to some information that is not directly related to the task.

It does not make the process automatically objective.

The OECD notes that algorithmic tools used in selection can reduce some forms of human bias, but can also reproduce or amplify bias depending on their design, data and use. [1]

A simple specialist comparison can also use poor criteria, overvalue portfolio aesthetics or favour people who describe their work better even when they do not necessarily perform it better.

Hiding the name and photo is therefore a design mechanism. It is not proof that the whole process is fair.

Do not start by asking how many skills a person lists.

Start with:

Which of their skills are actually needed for this assignment?

If you need a system migration, experience with data migration, integrations, testing and safe cutover may matter more than a long list of technologies.

If you need usability research, the ability to run research, synthesise observations and turn findings into decisions may matter more than familiarity with a design tool.

Fit does not automatically increase with the number of items on a skills list.

The statement "I can do this" and evidence "here is how I used it" are two different pieces of information.

Evidence can include:

• a work fragment,
• a sample completed for a similar task,
• a description of a problem solved,
• a repository,
• a report,
• a design,
• a recording,
• a result that can be independently checked,
• a reference, if it can be legally shared.

Research on employee selection is not research on provider portfolios, but the updated analysis by Sackett and colleagues highlights the value of methods closely connected to behaviours and knowledge required in a specific job, including work samples. [2]

A portfolio is not the same as a formal work-sample test. In both cases, however, it is useful to ask how closely the evidence relates to the real task.

Two projects can have the same label and be completely different.

"Online store" can mean a small catalogue with simple payment or a multi-country platform with complex logistics, returns and many integrations.

When comparing previous work, check:

• scale,
• complexity,
• constraints,
• type of users,
• the specialist's responsibility,
• project stage,
• time conditions,
• dependencies on other teams.

You do not necessarily need the same industry.

Look for similarity in the problem and the conditions that genuinely affect delivery difficulty.

The phrase "I worked on the project" does not explain what the person was responsible for.

Separate:

• what they did personally,
• what they were responsible for deciding,
• what was team work,
• which parts were already completed when they joined,
• what they did not do.

In large projects, a very good outcome can be the result of many people.

That does not reduce the value of the experience.

It simply lets you assess the right part of the experience instead of attributing the entire team's work to one person.

Not every project allows a measurable result to be shown publicly.

When a result is available, check:

• what exactly changed,
• whether there is a baseline,
• what part of the outcome can credibly be linked to the specialist's work,
• whether the data are measurable,
• whether they can be independently checked,
• whether the description avoids attributing a multi-factor outcome to a single intervention.

The absence of a public result does not mean the absence of competence. It only means that the independent verifiability of that particular evidence is lower.

Evidence from many years ago can still be valuable, especially for a durable skill.

In fast-changing fields, however, check whether the specialist has also used the competence recently.

Recency matters particularly when:

• tools or standards have changed,
• regulations have changed,
• the technical ecosystem has changed,
• user expectations have changed,
• the task requires knowledge of the current version of a system.

Do not ask only whether someone has done it before. Ask whether the evidence still represents their current ability to do the work.

A provider with an impressive portfolio can still be a poor fit for your specific problem.

Check whether they can:

• restate the objective in their own words,
• identify the main dependencies,
• distinguish a requirement from an assumption,
• identify missing information,
• propose logical stages,
• explain why they recommend that approach.

The World Bank's non-price criteria include quality of methodology and work plan as a common assessment area. [3]

The point is not to find one universally correct method.

The point is to see whether the proposal looks like an answer to your problem rather than the same template sent to every client.

A good proposal does not pretend uncertainty does not exist.

Look for whether the specialist can identify:

• what is not yet known,
• which assumptions affect price or timing,
• what could block the work,
• client-side dependencies,
• what needs to be checked first,
• actions that reduce the most important risks.

The World Bank lists risk management as a typical non-price assessment area. [3]

UK guidance on risk allocation also states that suppliers should be held accountable for results they can actually influence. [5]

Professionalism is not promising that nothing will go wrong. It also means recognising what can go wrong and who is able to manage it.

Competence and availability are not the same thing.

Someone can be an excellent technical fit but still be unable to deliver within the required timeframe.

Check:

• when they can start,
• how much time they can realistically allocate,
• whether they work alone or with a team,
• who will actually perform the critical parts,
• whether other commitments affect the schedule,
• how communication works during delivery.

You do not need full information about other clients.

You need enough information to judge whether the promised timeline has an operational basis.

The lowest price can be the best choice when proposals are genuinely comparable, the requirement is well defined and delivery risk is low.

That is not a universal rule.

The FAR explicitly states that the relative importance of price can vary depending on how definitive the requirement is, how much development work is needed and the risk of unsuccessful performance. [4]

Before comparing amounts, check:

• whether the scope is the same,
• whether similar outcomes are included,
• whether assumptions are similar,
• what additional costs may arise,
• who bears overrun risk,
• whether the more expensive proposal provides a concrete value you actually need.

Price is a strong economic signal. It is not standalone proof of competence or quality.

1. Job title.
The same title can represent completely different responsibilities in different organisations.

2. Number of years of experience by itself.
Ten years does not tell you how complex the work was, how quickly someone developed or how often they used the specific skill.

3. Brand of a former employer or client.
It can provide context, but does not by itself explain the person's responsibility.

4. Number of skills on a profile.
A long list is not automatically better than a short list of well-supported competencies.

5. Portfolio aesthetics.
Excellent presentation has value, but may be unrelated to the quality of the work you actually need to buy.

6. A high or low price by itself.
Price does not prove quality in either direction.

7. Photo, name and other traits unrelated to delivery.
They can affect first impressions, but the first competence assessment does not have to rely on them.

Structured comparison means asking each specialist a similar set of questions and evaluating similar types of evidence.

For each person, you might review:

• fit of key skills,
• strength of evidence,
• similarity of context,
• clarity of personal contribution,
• quality of the proposed approach,
• recognition of risk,
• capacity and availability,
• total cost.

Research in employee selection supports the value of job-related methods and greater process structure, but it does not create a universal scoring system for the services marketplace. [2]

You can therefore use descriptions such as:

strong evidence / partial evidence / insufficient evidence

instead of pretending that a difference between 8.2 and 8.4 has scientific meaning.

Structure helps consistency. It does not turn a qualitative decision into an objective laboratory measurement.

A skills-based assessment does not mean keeping people anonymous forever.

At a later stage, you may need:

• identity verification,
• information required to sign a contract,
• company or business details,
• legally required licences or professional authorisations,
• the actual team composition,
• time zone and work location when operationally relevant,
• information needed for security and access provisioning.

Some of this information may also be needed earlier if it is directly relevant to the requirement.

The rule is not: hide everything. The rule is: reveal information when it becomes necessary for a specific decision.

You can hide the photo and name and still build a poor process.

For example by:

• assessing evidence unrelated to the task,
• rewarding presentation style instead of solution quality,
• requiring a specific career history without justification,
• using criteria that were not known to everyone being assessed,
• changing the standard during comparison,
• assigning too much certainty to information that cannot be verified.

A useful test for every criterion is:

"How does this information help predict or assess the ability to perform this specific task?"

If the answer is weak, the criterion may be weak too.

1. Am I assessing the skills actually required for this task?
2. Are claims supported by evidence?
3. Does the evidence come from a similar context or problem?
4. Do I know what this person personally did?
5. Can the outcome be verified at least in part?
6. Is the evidence recent enough for this field?
7. Does the proposal demonstrate an understanding of my problem?
8. Does the specialist state assumptions and risks openly?
9. Does the declared availability support the proposed timeline?
10. Am I comparing price at a similar scope and level of responsibility?
11. Am I assessing everyone through similar questions and types of information?
12. Is any criterion unrelated to delivery having too much influence on the decision?

Comparing specialists without a CV, photo or job title does not mean pretending career history does not exist.

It changes the order.

First:

skill -> evidence -> context -> personal contribution -> outcome -> recency -> approach -> risk -> delivery capacity -> price.

Only later, information needed for further collaboration, verification and contracting.

This sequence does not guarantee a perfect decision.

It can, however, help you ask a better first question.

Not:

"Which profile looks most impressive?"

but:

"Which person presents the strongest, most relevant evidence that they can perform this specific work?"

[1] OECD - A Skills-First Labour Market: Promoting skills-first hiring and talent management, 2026
Open source

[2] Sackett, Zhang, Berry, Lievens - Revisiting the design of selection systems in light of new findings regarding the validity of widely used predictors, 2023
Open source

[3] World Bank - Rated Criteria
Open source

[4] U.S. Federal Acquisition Regulation - 15.101 Best value continuum
Open source

[5] UK Government - Risk Allocation and Pricing Approaches Guidance Note, 2026
Open source

Methodology note: sources [1] and [2] primarily concern recruitment and employee selection, while sources [3]-[5] concern formal procurement and contracting. This article uses only principles that can be cautiously applied to service comparison: task-related assessment of skills, the value of evidence, structured criteria, evaluation of methodology and risk, and separating price from other factors. The ten-signal model itself is an original editorial synthesis.