Every Product Decision Lands on Someone Else's Desk

Your head of sales has just told the leadership meeting that a deal fell through because product wouldn't build an integration, and everybody at the table has turned to look at you. You explain that product decided not to build that kind of thing, and that the decision is written down, with a reason. Your head of sales says nobody told him that was a decision. Then the founder asks whether you could make an exception for this one deal, just this once. The people around that table don't know what product is for, or why it said no.

Meridian's two pages

Meridian, the fictional composite company I've used through this series, is about to have that meeting. It has 40 people and 60 customers, and it sells shift scheduling software to logistics operators. Every Monday the founder, who is also the CEO, meets with the head of product, the head of sales and the lead engineer. In three weeks the founder has a board meeting.

Meridian keeps its strategy on one page. Its focus is logistics operators running 20 to 200 vehicles, and its bet is that an operator that size has no IT department, so it needs software that works on the first day with no project attached. The page also names one thing Meridian won't do: it doesn't build custom integrations for enterprise warehouse systems. This series calls a sentence like that a decline.

The head of product keeps a second page. Before anything ships, she writes down the number that would mean it worked. This series calls that page the scoreboard, and the number on it the bar.

This week the head of sales reports the lost deal: a prospect with 400 vehicles that wanted Meridian's schedules inside its own warehouse system.

Where a decision lands

The head of sales has a quota this quarter, and the no reached him as a deal he couldn't close. Every decision product makes lands on somebody else's desk. A bet you're unsure of is work your engineers might throw away. A date that slips breaks a promise the founder made to a customer. A result that falls short of the bar is money the board watched you spend.

When a decision arrives with no reason attached, it looks like product getting in the way, and people find a way around it. Meridian's head of product lost a decision that way in her sixth week. She turned down a custom integration for one prospect, and after a hallway conversation with the account executive, the founder put it back on the roadmap. The account executive was chasing a quota and the founder wanted to help. Neither of them knew why she'd said no.

The advice treats each relationship as a fight

Most of the advice I found about product and the other executives frames each relationship as a fight. It teaches you how to stop sales from overpromising, how to say no to the CEO, and how to stop wasting your board's time. You can win every one of those fights in the meeting and still have people go around you afterward.

The people going around you are the rest of the executive team. Melissa Perri wrote Escaping the Build Trap, and she sells a program for new chief product officers. She says a lot of first-timers in that job get pushed out, and one reason she gives is that they forget which team they're on: they give their time to their product team, forget that they're also members of the executive team, and then the executive team turns on them. Patrick Lencioni calls the team they forgot Team #1, and his firm, The Table Group, puts the order this way: "they need to prioritize their leadership team (Team #1) first, for the good of the organization." Perri also passes on advice from Gif Constable, who ran product at Meetup: expect to build trust with the other executives, and don't presume you already have it.

One sheet, four columns

You build that trust by asking what each of them needs from product to do their own job (the three people at the Monday table, and the board) and giving it to them before they need it. Put the answers on one sheet with four columns. Each column is a conversation you keep having with one of them. Most of what goes in them comes off the two pages Meridian already has, the strategy and the scoreboard. The layout is my own.

The sheet Sales: what it can promise this quarter, what product has said no to and why, and what would change that answer. The founder: the options, and what each one would delay, talked through with the whole leadership team. Engineering: which parts of the product are settled and which are bets. The board: the scoreboard, with the bar written before the work started and the result beside it, and any miss told to each board member before the meeting.

Sales

In part five I said the decline is the sentence naming what you will not do. At Meridian, the head of sales first met that sentence when a prospect asked for the very thing it rules out.

"Sales-led companies let their contracts define their product strategy," Melissa Perri writes in Escaping the Build Trap, about companies where sales sets the direction, though she also says plenty of small companies start out that way and it can be okay. Sales gets paid to close the account in front of it, and product has to serve the whole market it chose.

So sales needs three things from the sheet: what it can promise this quarter, what product has said no to and why, and what would change that answer. Meridian already knows the third. Three customers inside its focus would need to be struggling to get their schedules into another system, and so far only two of its 60 are.

Before the next Monday meeting, the head of product fills in the sales column and walks the head of sales through it, with the strategy page on the table. The lost prospect runs 400 vehicles and has its own IT department. It wanted a project, and Meridian decided not to be in the project business. Inside the focus, operators buy without a project attached, and that's where the head of sales can spend his week. He can promise what works today. He can't promise the templates Meridian is building for overtime rules, because they're still a bet that might not work. When a prospect asks for the integration, he can now say:

"We don't build custom warehouse integrations, because our customers need to be running on the first day. If three operators like you need the same thing, we'll look again."

The founder

When sales can't get past a decline, the next stop is the founder, who was already asking for an exception and can overrule you. In a 2019 episode of the Fearless Product Leadership series, six product leaders talked about saying no to a CEO. One product leader told a story about a quarterly call with their company's biggest customer. The CEO joined the call and, partway through, promised the customer a feature the product team had already decided against. The team had turned it down because nobody else in its market wanted it. After the call, the product leader sat down with the CEO and went through the work for other customers that would slip if the team built it. The CEO agreed, and the two of them called the customer and withdrew the promise, which the product leader says is never fun.

That fix came after the promise was already out. The episode's last speaker suggests having the conversation earlier and in the open: whenever the CEO wants something that breaks the plan, lay out the options and what each one would delay, and talk them through with the CEO's other direct reports in the room, so the whole leadership team picks which cost to pay. Rich Mironov, who coaches product executives, turns a request into the same trade-off by asking about the work already promised to the person asking. His sample script: "Roadmap items X and Y specifically benefit your functional group… Do you want to postpone/cancel those in favor of this?"

The founder's column holds options laid out that way, each with what it would delay. At Meridian, the head of product brings two of them to the next Monday meeting. Building the integration for the big prospect pushes the overtime templates back about three months. Keeping the decline means letting that prospect go while the head of sales chases operators inside the focus. The founder could still choose the integration. It's his company, and that's a legitimate call. If he makes it, he makes it in front of the team, knowing the cost, and everyone in the room hears the decline change.

The founder keeps the decline.

Engineering

The three months that building the integration would cost came from the lead engineer, and engineering is where an exception keeps costing time after it ships. Two of Meridian's six engineers already spend most of their week keeping custom integrations running for three customers, and every exception adds to that load.

The engineering column sorts the product into settled pieces and bets. The team builds a settled piece to last and builds a bet so it's cheap to change. At Meridian the scheduling core is settled: customers run their week on it, and the strategy depends on it working on the first day. The overtime templates are a bet. Their bar says five of the next six new customers will publish a first schedule inside a day, and since that might not happen, the engineers build the first version to be replaced. The head of product tells the lead engineer: "The scheduling core is settled for this year. The templates are a bet, so make them easy to change."

Product and engineering can have that conversation at Meridian because they sit in separate chairs. I've been the CTO that product reported to, and early on, one person holding both what customers need and what the engineers can build is the right setup. As the company grows, that person ends up deciding the trade-offs alone. Meridian's head of product and lead engineer can argue them out, as long as the lead engineer is in the room before the decision.

The board

Every hour of engineering time is money, and the board watches where the money goes. A young company lives on three numbers: revenue, costs and runway. I've preached that for years. Every quarter of product work spends runway, so the board needs to know what it bought. In part eight I said a scoreboard that tells the truth will be mostly misses, and the board meeting is where you report them.

Mark Suster, a venture investor, calls no surprises "the golden rule of board meetings." Fred Wilson, another investor, wants a call from you before the meeting, so no board member hears bad news for the first time in the room. Brett Fox, a startup coach, says the first job of the meeting is to report what you've done against what you said you'd do. All three write for CEOs; applying their advice to a head of product is my own step.

The board's column is the scoreboard, which already holds what you said you'd do and what you did: the bar is written before the work starts, and the result goes next to it. At Meridian's last board meeting the update was a list of what shipped. This time the head of product brings last quarter's scoreboard. The bar said four of six new customers would publish a first schedule inside a day without calling support, and three did.

A week before the meeting, the head of product calls each board member. "Before Thursday, I want you to hear this from me. We set a bar of four, we got three, and here's what we learned." Three days before the meeting she sends the scoreboard out. Brad Feld, another investor, wants board materials early so the meeting can be a discussion. "The day of the 'board update' is over," he writes. In the meeting she shows what this quarter's spending is for: the overtime templates, because the other three new customers all got stuck setting up overtime rules. The meeting turns to what the miss taught them.

A month later

A month after the board meeting, the head of sales reports a deal of his own at the Monday meeting. The customer runs 40 vehicles, has no IT department, and signed without anyone from product on the call. Later that week another customer asks him about the integration, and he explains the decline himself, with the reason and what would change it.

Every decision you make lands on somebody's desk, so send the reason with it.

What I can't tell you

No study I could find compares companies holding these four conversations with companies that don't, and there's no credible number on how often product leaders get pushed out. Before your next leadership meeting, pick one thing you've said no to and walk your head of sales through the reason, and what would change it.

Dave Saunders is the founder of Base Reality Group and a Fractional CPO for product companies. He was a founder and operator at Galen Robotics, where the surgical-robotics platform earned FDA De Novo authorization in 2023, and he managed a 35-patent portfolio licensed from Johns Hopkins. He wrote Founders Who Finish and publishes The Build. More about Dave →