The Offer Letter Is a First Draft

Years ago, back when I was still new to managing people, I sent a young rep an offer that came in about twelve grand under what I was actually allowed to pay him. He said yes before I'd finished my coffee. The email was mostly logistics, his start date, could he expense a monitor, and it signed off with a line I still think about: "Thrilled, thank you so much!!" Two exclamation points. He was happy. He also had no idea he'd just left a used car sitting in the parking lot.

That twelve grand, by the way, wasn't a mistake. I built the offer that way on purpose, the same way you list a house a little high and wait for someone to talk you down. Everybody who writes these letters leaves room in them. The number you get is an asking price. And for reasons I genuinely cannot explain after seventeen years of watching it happen, the candidate is almost always the only person at the table who treats it like the final sale.

So let me tell you how the other side actually thinks, because I was the other side. Seventeen years in enterprise sales, the last chunk of it running comp for my own team and signing off on what we paid people. I wrote the lowballs. I also sat across from the handful of people brave enough to push back on one, and here's what I noticed: the ones who pushed almost always walked away with more. Not because they were smooth talkers. Because they asked. And I'd already set the money aside for anyone who did.

Now, the fear, because I know it's sitting right there in your chest: if I push back, will they yank the offer? In seventeen years, I saw that happen over a polite, reasonable counter exactly never. Companies don't torch weeks of recruiting to punish you for negotiating like an adult. Worst case, they say "that's really our best," and you're standing exactly where you started. The downside is a rounding error. The upside is a used car. Let's go get the car.

One quick note before the tactics. The dollar figures in the scripts below are made up, or really, they're the shape of deals I've actually run with the names filed off. When I quote an actual market number, an equity split, a band spread, I've linked where it comes from. Your real numbers depend on your field and your level. The moves, though, those travel anywhere.

Read the package before you touch it

An offer isn't one number. It's three things stapled together and handed to you as one: base salary, equity, and a sign-on bonus. Staring at the total, they blur. But each one comes out of a different pocket, and the company protects each for a different reason, and that, right there, is the whole game. You counter each piece differently because each piece hurts them differently.

Base salary is the one that matters most and moves least. It's recurring cash, so every dollar you add follows you around forever. It lifts every future raise, because raises are almost always a percentage of base, and it resets what your next employer assumes you're worth. So the company fights hardest right here. A ten-thousand-dollar bump to you is ten grand a year to them, forever, with raises stacking on top. When a recruiter tells you "there's just no room on base," they're often telling the truth about that one bucket, while sitting on plenty of room in the others.

Equity is the part people go all starry-eyed over before they understand a word of it. At a startup it's a lottery ticket that vests in slow motion, and the number they toss you, "0.2 percent," or "$180,000 over four years," honestly means close to nothing until you've asked a few pointed questions, which I'll hand you in a minute. For an early engineer it's often a huge slice of the package, and it's the slice you understand the least, which is a rough combination when you sit with it.

Sign-on is a one-time check, and that one-time nature is the whole reason it's so gettable. It doesn't recur, it doesn't compound, so it never becomes a permanent line on anybody's budget. The money usually comes out of some discretionary pool the hiring manager isn't personally guarding. Tuck that away, because it's the very first lever I reach for when base won't budge.

Once you see the three buckets for what they are, "can you do any better on the offer?" turns into "can you move the sign-on to twenty-five?" And that second question, unlike the first, is one a manager can actually say yes to without booking a week of meetings.

Countering base: anchor, then hold

Start with base, because honestly, the transparency laws have handed you a gift here. As of 2026, roughly eighteen states plus Washington, D.C. make employers put salary ranges right in the posting, or cough them up if you ask (Jackson Lewis, 2026). A remote job can often be done from several of those states at once, so the employer usually ends up following whichever rule is strictest. Which means the band is probably printed somewhere already, and if it isn't, asking for it doesn't make you look greedy. It makes you look like someone who's done this before.

And these bands are wide. For one role at one level, the spread from floor to ceiling usually runs 20 to 40 percent (Carta). So "the range" is real estate, and where you land in it is a decision a human being makes, not a law of nature. Most offers show up at or below the midpoint, because that leaves the company room and because you haven't yet given them a reason to reach higher. So give them the reason. Aim for the top third.

Here's roughly what I say, or what I coach people to say. Keep it warm, keep it specific, and do not apologize for asking. Say the offer landed at $150,000 and the posted band tops out around $175,000:

"I'm genuinely excited about this and I want to make it easy to say yes. Based on the scope we discussed and where I'm coming in, I was targeting a base of $168,000, which still sits inside your posted range. Can we get there?"

Look at everything that little paragraph is doing. "Make it easy to say yes" keeps it collaborative, so nobody's suddenly bracing for a brawl. The reason is scope, never need, because "the job is bigger than the title suggests" moves a manager and "I have a mortgage" just makes everyone stare at their shoes. The number sits near the top of their own posted band, so they can't call it unreasonable without insulting their own job listing. And it ends on a question they can answer in one word, which puts the next move back on them.

Then, and this is the hard part, you shut up. The single most useful thing you can do after saying your number is absolutely nothing. Let the silence stretch. Let it get a little uncomfortable. Whoever caves first and fills that quiet gives up a bit of ground, and there is no rule anywhere that says it has to be you.

Equity is where the fuzzy math lives

Don't counter an equity number until you actually understand it, because a percentage with no denominator is theater. When they float a grant at you, ask this, and ask it like it's the most ordinary thing in the world:

"To model this properly, can you share the total number of shares outstanding, the strike price and the current 409A valuation, the vesting schedule and cliff, and whether there's a refresh policy? I just want to understand what I'm being granted."

A company worth joining answers that without blinking. One that suddenly gets vague just told you something useful for free. Here's what you're really pricing out:

• The denominator. "0.2%" of what, exactly? Percentage of total shares is the only figure that means anything, and you can't get there without knowing how many shares even exist.
• Strike and 409A spread. The gap between your strike price and the current valuation is your paper gain. It's also, eventually, your tax headache.
• Vesting and cliff. The standard is four years with a one-year cliff, so you own nothing until month twelve. Hold that thought, because it matters a lot for the sign-on in a second.
• Refresh timing. Almost nobody counters on this, which is a shame, because it's basically free money hiding in plain sight. Your first grant vests down over four years, and if there's no refresh, your equity comp shrinks a little every year even as you take on more. Ask when refreshes happen and whether you're eligible at your first review instead of your fourth. Getting "eligible for a refresh at your first annual review" written down can be worth more than a few grand on base, and it's cheaper for them to promise, because it's future money that only pays out if you're still around to collect it.

On grant size, earlier-stage companies tend to weight equity heavier and leave more room to move it, while at the later-stage places the bands are standardized and the number often really is fixed (Carta's startup compensation data). When the grant itself won't budge, pivot to the terms around it, refresh eligibility, a longer window to exercise after you leave, and to the buckets that will move.

Sign-on is the pressure valve

If base is the hardest yes and equity is the murkiest, the sign-on is the pressure valve, the spot where all that stuck negotiation finally lets off some steam. It's a one-time check, so it never haunts next year's budget. The money usually sits in some discretionary pool the hiring manager isn't defending to the death. And it comes with a built-in excuse baked right in, which is exactly what makes it easy for them to hand over: it covers what you're giving up to walk out your current door.

So say that part out loud, and put a real number on what you're leaving behind:

"I'm walking away from an unvested equity grant and a bonus that pays out in March. To make this move whole in year one, I'd need a sign-on of $30,000. Given the vesting cliff on the new equity, that first year is where the gap is."

That lands because it's true and you can point at every piece of it. You're not begging for a bonus, you're documenting a real loss, and their own vesting cliff is out there making your argument for you, because they know as well as you do that year one is when your money runs thin. Sign-on numbers are all over the map by company and level, but the one constant is that they live outside the recurring budget, which is exactly what makes them the most flexible lever in the whole thing. I know, because I used to be the one signing off on them. When they genuinely can't move base, this is usually where "no" turns into "well, how much do you need." One warning, though: sign-ons often come with a clawback if you leave inside twelve or twenty-four months, so read that clause before you go pop any champagne.

The rest of the offer is negotiable too

Most people stop at base, maybe take a halfhearted swing at the sign-on, and call it a day. Which is a shame, because a lot of the money is hiding in the parts nobody thinks to touch.

Title and level. People sleep on this one, and they shouldn't, because your level usually decides which band you're even paid from, and that band steers your raises for years. Coming in as a Senior instead of a Mid, or a Staff instead of a Senior, can reroute your whole trajectory, not just your first paycheck. If the job is genuinely a senior job, say so, and ask for the title to match. Titles cost the company nothing to hand out and are worth a small fortune to you.

Band placement. Even at a locked level, you can ask where in the band they're slotting you, and push for the top third with the same scope argument you used on base. "I'd love to come in at the top of the level-4 band, given the scope" is a sentence a decent manager can actually act on.

Remote stipend. A one-time home-office setup stipend, usually somewhere in the $1,000 to $1,500 neighborhood, is pretty standard at remote-first companies. Higher at some, missing entirely at others. If it's not in the offer, just ask. It costs them next to nothing, and it's a real desk and a real chair to you, which your back will thank you for around month three.

Review and refresh timing. Ask for your first comp review at six months instead of the usual twelve. Reviews are typically annual and tied to performance (Carta), so pulling yours forward gets you to your first raise sooner, and if you're crushing it, it costs them nothing to say yes now.

Start date. The most human item on the whole list. Push your start out two or three weeks and take an actual break between jobs. You will not get this window again, and sprinting straight from one role into the next is how people show up on day one already running on empty. I've watched people fight like hell over money and then hand back the one thing money can't buy them, which is the pause. Take the pause. Ask for it.

The one email that ties it together

Send these asks one at a time and each one's easy to swat down on its own. Put them all in a single calm message, though, and they become a package somebody actually has to sit with. Here's the shape of what I'd send. Numbers made up, structure real:

Hi [Name],

Thank you for this. I'm really excited about the team and the problem, and I want to make this work. I've thought it through and here's where I'd need to land to say yes:

• Base: $168,000 (inside your posted range, and reflecting the scope we discussed)
• Sign-on: $30,000, to bridge the unvested equity and bonus I'm leaving, given the one-year cliff here
• Equity: I'd like to confirm refresh eligibility at my first annual review, and get the total-shares and strike figures so I can model the grant properly
• Title: Senior [Role], to match the scope
• First comp review at six months rather than twelve

Everything else in the offer looks good to me. If we can get to this, I'm ready to sign this week. Happy to hop on a call if that's easier.

That last line does more work than any of the numbers. You've named your price and told them a yes is one reply away. And because you bundled the asks, they can trade: maybe base holds but the sign-on grows, maybe the title moves and the review date carries the rest. The whole thing stays warm, which, honestly, is the only sane way to negotiate with people you'll be on Slack with come Monday morning.

One more thing that makes every script up there hit harder: all of it assumes you could walk. And you can only really walk when there's another conversation happening somewhere else. A single offer you can't afford to lose isn't a negotiation. It's a countdown with a signature at the end. So keep more than one pot on the stove while you interview. If that's the piece that always falls apart when you're busy and fried, an auto-apply tool like JobCopilot can keep the applications trickling out in the background, so you're picking between offers instead of clutching the only one you've got.

Knowing when to stop

One last thing, and it's the one that took me the longest to learn, and cost me the most in the learning. Somewhere in every negotiation you hit a number that makes the job a genuinely good deal, and from that point on the entire skill is noticing you've hit it and stopping. I could never stop. I chased the last dollar of every deal, because somewhere along the way the number had become the scoreboard, and let me tell you, the number is a miserable scoreboard for a life. It cost me a marriage and most of my late thirties before that lesson finally landed.

So here's my rule, and it's a plain one. Before you counter, pick two numbers: the one that'd make you genuinely glad to take the job, and the one below which you'd honestly rather keep looking. Counter toward the first. If they land you anywhere north of the second, take it and stop, even if you're pretty sure there's another two percent hiding somewhere in the room. That two percent is not worth starting a new job as the person who wrung out the last drop on day zero. And if they won't even clear your walk-away number? Then walk, and mean it, because a company that underpays you just to get you in the door almost never gets around to fixing it once you're inside.

They sent you a first draft. Send one back.