Should I take this job offer? A framework + AI second opinion

Job offers feel binary on the inside — yes or no — but on the outside they're seven different decisions wearing one envelope. The reason people regret either accepting or declining is almost always that they optimized for one dimension (usually money or "the brand") and ignored the other six. This is the framework that fixes that.

It takes about twenty minutes to run, and the output is a written list you can show your partner. If you want the second opinion in chat form, the AI career consultant will run the same seven points with you against your actual offer.

The 7-point offer evaluation

Score each on a simple green / yellow / red. No five-point scales — you'll spend twenty minutes arguing whether something is a 3 or a 4. Green means clearly good. Red means clearly bad. Yellow is everything else.

1. Compensation

Not just the base. Total comp = base + bonus (realistic, not max) + equity (vesting, refresh policy) + signing bonus (amortized over how long you'll stay) + benefits (insurance, pension, time off). Compare against the median for your role and location, not against your current package. If you don't know the median, public data from levels.fyi, Glassdoor, and local engineering communities gets you within 10%. Green: at or above median. Yellow: 5–15% below. Red: more than 15% below.

2. Growth trajectory

Will you be more hireable two years from now? The shortcut question: "What's the next role I'd grow into here, and how long does it usually take?" If they answer with a clear path and recent examples, green. If they say "we don't really do levels," that's usually a euphemism for "we promote nobody." Red.

3. The hiring manager

The single most important variable. Bad manager destroys good role; good manager rescues bad role. Specific signals: how do they give feedback, what was the last person they promoted, how do they handle disagreement. Ask to talk to one of their current direct reports. If you're not allowed, the answer is already red.

4. The actual role

Read the job description back to yourself. Is the day-to-day work you'd be doing actually what you want to do? "Senior engineer at FAANG" sounds great; "writing Terraform for the internal dev platform team" might not be what you signed up for. Ask: "What does my first 90 days look like?" — and listen for boring vs exciting honestly.

5. Market and runway

Is the company itself going to be around in two years? Public company: look at the last earnings call, revenue trajectory, recent layoffs. Startup: months of runway, revenue, last raise, burn rate. If the answer to "what's runway" is "we don't disclose that," translate that to "less than 12 months." Layoffs less than 6 months ago: yellow at best.

6. Family and life fit

Hours, location, commute, remote policy, on-call expectations, travel. The honest test: if you tell your partner the deal, do they look relieved or worried? Career decisions you make alone tend to get reversed within a year.

7. Gut check

Imagine it's six months from now. You're at your desk on a Wednesday afternoon in this new job. How does that feel? Gut isn't magic — it's pattern recognition from every signal you noticed during the interview process but didn't write down. If your gut is loud, listen to it. If it's quiet, ignore it.

Red flags that override the score

Some signals are so bad that they cancel everything else. Treat any of these as automatic red:

• Pressure to decide in under 48 hours. Real companies give you a weekend. "Exploding offers" are a tell.
• Refusal to put the offer in writing. If you can't get the full package in an email or PDF, walk.
• Last-minute change in role or level. "Actually we'd start you as a Senior 1 instead of Senior 2." That bait-and-switch repeats.
• Bad-mouthing the person you'd replace. You will be the next person they bad-mouth.
• You couldn't meet your manager during the loop. They're hiding something — manager doesn't know, doesn't care, or doesn't exist yet.
• The recruiter dodges every question about retention or attrition. If turnover is healthy, that's an easy answer.

When to negotiate vs decline

Most offers are not "yes or no." They're "yes if X." Sorting tells you which lever to pull.

• 5–7 greens, money is the issue: negotiate. Counter the base, ask for a sign-on, or push equity. See the dedicated salary negotiation script.
• 5–7 greens, role scope is the issue: negotiate scope, title, or team. Easier to get than money.
• 3–4 greens, manager is yellow: ask for a follow-up call with the manager and one of their reports. If still yellow, decline.
• 3–4 greens, comp is the issue: counter once, hard. If they hold, the offer is telling you what they think you're worth there.
• Two or fewer greens: decline politely and stay in touch with the recruiter — better roles often come from the same company twelve months later.

The mistake people make is negotiating offers they should be declining. Money rarely fixes a role that's wrong on three dimensions. You'll be back on the market in nine months, just with a worse story.

What to say when you decline

Decline in writing, briefly, and don't burn the bridge. Template that works:

"Thanks again for the offer and for the time the team put in. After thinking carefully, I'm not going to move forward right now — the timing and the specific scope aren't quite right for me. I'd love to stay in touch; I have a lot of respect for what your team is building."

No long explanations. No detailed feedback unless they explicitly ask (and even then, keep it gracious). The career world is small and recruiters remember who handled "no" well.

Frequently asked questions