Most market-data briefings fail not because the data is wrong, but because the data answers questions the hiring manager isn't asking. They show up wanting to know "why is this so hard" and leave with twelve charts about pool sizes. This guide is about closing that gap.
When this conversation matters
Not every market data conversation needs this much structure. The five scenarios where it does — because the meeting carries real weight on a decision:
- Opening a hard req where the comp band is below market. The manager wants the role filled. You need them to either approve the comp lift or reframe the role. The meeting is where that happens.
- A mid-funnel role that's been stuck 60+ days. They're frustrated. You need to land that the math is broken, not the recruiting effort, and propose what changes.
- Geo expansion or remote-vs-site debate. They want Austin again. You need to show them what the data says about Austin, and what the alternatives actually look like.
- Quarterly or pipeline-health TI refresh. Standing meeting where you give them market context they'll forget half of unless it's structured around their actual decisions.
- New initiative ramp-up. They're building a new team and need to know what's realistic on volume, comp, time-to-fill, and geo. This is where the early framing gets locked in.
Pre-meeting prep — what you bring
A hiring manager's tolerance for analyst presentation is roughly 90 seconds. Compress aggressively. The prep checklist that holds up in real semi/AI meetings:
One page
A single page they can scan in 60 seconds. No appendix. If they want depth, send the backing file after, not during.
Three numbers
Three numbers they will remember and repeat. Choose them deliberately. Everything else is supporting cast.
One chart, maybe
If a chart genuinely changes how they think — bring it. If it just decorates, leave it. A bar chart of pool sizes is rarely the chart that matters.
One recommendation
Named, specific, owned. "Here's what I think we should do differently." Without this, the meeting ends in nodding and nothing changes.
The deck
If you find yourself building a 15-slide deck, you are preparing for the wrong meeting. Send a doc after.
The methodology
They don't care how the BLS OES data was filtered. They care what the data says about their hire. Methodology lives in the appendix.
Hedging language
"It's complicated" and "there are many factors" are how analyst credibility dies. Be specific. Be wrong sometimes if you have to. Wrong-and-specific beats vague-and-safe.
HR framing
Avoid "EVP," "talent acquisition lifecycle," "candidate experience signal," "engagement KPI." Talk like an engineer who happens to know the market.
The 30-minute meeting structure
A working default flow for a recurring or initiative meeting. Adjust for shorter and longer slots, but keep the proportions. Most TI briefings overrun on Section 2 and skip Section 4, which is exactly backwards.
What's changed in the market since last time
One paragraph, max. New data point, new signal, new shift. Sets the frame: "the world is different than when we last talked, here's how." Skip if you've never met.
The answer to their question
The thing they asked you to investigate. Three numbers, one line of interpretation each. Lead with the punchline. Do not bury the answer in setup.
The thing they didn't ask but should have
The 2–3 second-order findings that change the picture. This is where TI earns its seat. Pace it — leave room for them to push back. If they don't push back, you didn't say anything sharp enough.
Your recommendation and the trade-off
"Here's what I'd do." Then the cost of doing it and the cost of not. They may disagree — that's fine, but the recommendation has to be on the table for them to react to.
What we commit to differently
Concrete. "I'll re-scope role X by Friday." "We'll open Austin posting for senior IC only." "I'll come back in three weeks with comp comparison vs Apple Silicon." Write it down in the room.
The 8 questions hiring managers actually ask
Eight questions show up in nearly every semi/AI hiring conversation. The data answer matters — but so does the translation. Here's the pattern for each: what they're really asking, and how to answer with market data without sliding into HR-speak.
Is recruiting screwing this up, or is the role genuinely scarce? They want a math answer they can defend to their leadership.
Tell me what to take to comp. They want a number — ideally one that won't get challenged by total rewards in five minutes.
I want to stop fighting for the same Bay Area candidates everyone else is fighting for. Is the data on my side?
Recruiters are giving me the same five companies' candidates. Are we missing people because we have a thin source list?
Is it comp? Is it our brand? Is it me? They want to know what to change, not a survey readout.
I'm getting heat about cycle time. I want to know what's realistic so I can either push recruiting to move faster or push leadership to be patient.
What does X have that I don't? I want to know whether to build that thing, position around it, or accept the loss.
Is there a faster path than slogging through individual recs? They want a yes/no with a name attached, or a clear no-it's-not-realistic.
Translation cheat sheet
The single most reliable predictor of whether a TI briefing lands is whether it's spoken in the language of the room. In a semi/AI hiring meeting, the room is engineers and engineering managers. Translate aggressively.
Post-meeting commit template
A briefing without a written commit is a briefing that gets re-litigated next month. Fill this in during the last 5 minutes of the meeting, share it within 24 hours, and reference it in the next session.
Guardrails
- Land one thing. If they remember one number, one insight, one action — pick which one before the meeting. Don't outsource that choice to them.
- Push back when you should. A briefing where you nodded the whole time was not a briefing. TI's job includes telling hiring managers when their assumption is wrong. Do it kindly. Do it.
- Never punt with "it's complicated." Even if it really is. Pick the most likely answer and label your confidence. "Best guess: X, ~70% confident, here's what would change it" is a usable line. "It's complicated" is not.
- Don't blame other functions in the room. If comp is the problem, say "comp is the problem" — not "it's a comp issue, that's not really our area." Own the analysis, even when the fix lives elsewhere.
- Write the commits down in the room. Not after. In the room. Out loud. Hiring managers who watched you type the commit are far more likely to honor it.
- Send the followup within 24 hours. 48 if there's no way around it. Anything longer and the data falls out of the manager's head and the meeting effectively didn't happen.
- Reference the prior commit at the start of the next session. "Last time we agreed X — here's where we landed on it." This is how TI gets a real seat at the table over time.