Idea

How to calculate how much salary to ask for in an offer

The idea came from the need to organize and explain the salary-expectation math. It became a live calculator, in Portuguese and English.

Érick Jesus6 min read

  • Product Design
  • Product Ownership
  • Applied AI
  • Salary negotiation
  • Career
  • Mobile

How the idea started

Negotiating salary looks simple and isn't. The recruiter asks "what is your salary expectation?" and many people don't know how to arrive at a number. People guess a figure, or accept the first one that shows up. And when an offer comes with a higher salary but without the benefits the person already had, it gets even harder to tell if it pays off.

I felt the need to organize this math and explain it in a clear way. The idea: a calculator that estimates how much to ask and, in a separate calculation, helps you see whether the offer makes sense.

The idea

The calculator was built for the Brazilian job market. On a single mobile screen, the person enters their current salary and how urgently they need to leave their current job. If they have an offer, they turn on the comparison and enter its values, including benefits such as meal voucher, food voucher, health plan, bonus and profit sharing (known in Brazil as PLR).

The result shows three values, a minimum, an ideal and an aggressive one, plus a counteroffer script to use in the conversation.

The product is live and you can try it right now, here or at quantopedir.com.br.

The real product, live. Try it with your salary.

The method in three steps

  1. Start from base salary. Negotiations usually center on it. If you already have an offer, the base is your current package minus the benefits the new company already pays.
  2. Apply a raise that varies by salary range. With a salary of R$ 6.000, for example, the calculator suggests +12% as the minimum, +18% as the ideal and +25% as the aggressive one.
  3. Adjust for urgency. The more urgently you need to leave, the less room you have to negotiate, and the suggested raise adjusts for that. The example above uses neutral urgency.

In the example, the ideal figure is R$ 7.080 (R$ 6.000 plus 18%; the product shows amounts in the Brazilian format, with a period for thousands). Whether an offer is worth it is a different calculation, one that compares the full package; I explain the difference below.

What the calculator does not do

The values are estimates, calculated from what you enter and from my own assumptions. They are not official market data, nor legal, labor or financial advice, and the three values are references, not a guarantee that an offer will be accepted. For important decisions, consult a professional.

The decision that changed the product most

The first version added salary and benefits together and used that total to suggest how much to ask. The number came out higher than seemed reasonable to me.

I fixed it by separating two calculations I had been mixing:

  • How much to ask starts from base salary, because negotiations usually center on it. When there is an offer, the base is your current package minus the benefits the new company already pays.
  • Is the offer worth it? compares the full package, salary and benefits.

In my assessment, that adjustment made the result simpler and more coherent.

The positioning

The product started with the question "how much should I ask?". I changed it to "understand whether your offer makes sense for you", with a sentence that sums up the intent: to help people decide with more information, not to accept an offer without comparing it.

I also thought about how much is reasonable to ask. I considered what raise the human resources (HR) team tends to accept and that is still good for the person. That led me to revisit the percentages, which today vary by salary range. This is my own assumption, not market data: I start from the idea that people who earn less have less room to negotiate.

Other decisions

  • Urgency: I assume that people who are more urgent to leave have less room to negotiate, so the urgency control adjusts the suggestion. In the first version it added or subtracted points and could produce negative percentages. I switched to a multiplicative factor and tested all five levels, with and without an offer.
  • No cents: values are entered without cents, to keep filling in simple.
  • The field being filled stays visible: on mobile, the keyboard and the fixed button covered the field being filled. I fixed it.
  • "I don't know": each benefit has a reference value for people who don't know the amount. The meal/food voucher and transport ones come from public 2026 sources; the others are my own estimates, declared on the screen itself.
  • Calculate again: a button at the end to start over with one tap.
  • Accessibility: the goal is level AAA. I reviewed labels and focus and checked contrast and touch-target size with an automated tool and pixel sampling, with Claude's help. No external audit.
  • Privacy: the measurement tools (Google Analytics and Microsoft Clarity) and AdSense only load after the person accepts, and declining has the same visual weight as accepting. The values you type are not sent.

Made with AI

I built the product in conversation with Claude, over a few days in April 2026. To check the calculation logic, I compared Claude's answer with another AI's. Claude indicated that the code already covered the other one's suggestion. The check was between AIs, not by a labor-law or finance specialist. I learned that comparing sources helps, and that the person who decides what goes into the product is still the designer.

In October 2026, before the latest release, a second AI agent reviewed the code without having seen the work and flagged real problems: an imprecise privacy text, insufficient contrast in some notices and a button that was too small. I fixed them all.

The product is lightweight: it runs on a single page, with no installation, in Portuguese and English.

Where this can go

People who search for "pretensão salarial" (salary expectation) on Google are probably at a decision point. In Google Analytics, over the last 90 days, the engagement rate was 72% for search visitors and 25% for direct visits. These numbers have no statistical test: they are an indication, not a conclusion. This suggests an audience that already wants to solve the problem, and that is where the idea can grow. It is a hypothesis.

  • For recruiters: job sites and HR platforms could offer the calculator to candidates, and the salary conversation could become more objective. A hypothesis to test.
  • For product builders: the idea of separating the base calculation from the full-package comparison could apply to other decisions with several numbers. It is a hypothesis, not yet tested.

What I would measure first: how many people reach the result, how many copy the counteroffer, and how many come back with a new offer. This can already be measured in the current product.

What do you think?

If you work with career, HR or recruiting products, or know someone who needs a tool like this, write to me. I would be glad to talk.

Contact

Want to talk about this idea?

Agree, disagree or see a better path? Write to me.

Only used to reply to you.

Up to 2000 characters.

Your details are used only to reply to this message. It is delivered through the Web3Forms service. Privacy policy