If you came to a Plotline all hands and looked at the slide where we show the roadmap, you would think we don't know what we're doing. There are about 40 items on it. Some are quarters out. Some are years out. Some have no dates at all. There are arrows between things, scratch outs, items moved between columns. It looks like a working document because it is a working document.

This used to bother me. I would spend the Sunday before each all hands cleaning it up. Trim the items. Rationalize the dates. Make the arrows look neater. Then I'd present it Monday and a sales person would ask about a feature I had cut from the slide, and I'd realize the slide didn't actually represent what we were doing.

About 18 months ago I stopped cleaning it up. The slide we show is the real working doc. I will explain why.

The clean roadmap is a marketing artifact

The clean version of a product roadmap, the one that goes in a customer deck or a board update, is a marketing artifact. It says "we are confident about these things and we will deliver them in this order on this timeline." The audience for this artifact is people who need to be reassured.

The thing is, the people who actually build the product cannot use the marketing artifact to do their jobs. The marketing artifact is too clean. It hides the uncertainty. Engineers look at it and think "this is what I am committed to" when really the answer is "this is what we are pretty sure we want but also Q2 might change everything depending on what we learn from the multifamily pilot."

If you only have the clean version, the engineers build the wrong thing. Because they take the clean version literally.

The messy version forces conversation

The messy version, the one with 40 items and scratch outs and arrows, forces people to ask questions. "Why is this arrow here." "What's the status on this item." "Why did we move this back a quarter." Every question is a chance for the team to align on what is actually true, not what looked good when I cleaned up the slide on Sunday night.

The messy version also surfaces conflict early. If two PMs both think they own the same item, the mess shows that. If engineering has been quietly working on something that's not on the roadmap, the mess shows that. If a customer commitment was made without product knowing about it, the mess shows that.

I would rather have those conversations on a Monday than three months later when something has gone wrong.

The risk of messy

The obvious risk is that the team feels disorganized. We have had two engineers leave over the last two years and one of them, in his exit interview, said the roadmap felt chaotic and he wanted a more predictable environment. Fair feedback. He was right. The environment is less predictable than it would be at a 5000 person company.

The trade I have made is this. I would rather lose the engineer who needs the predictable environment than have the predictable environment and lose the customer who needs us to react to their actual problem in their actual timeframe.

I am not certain this trade is the right one. I think it is the right one for a Series B company. I am not sure it will be the right one at Series C, when there are 120 people instead of 48 and the cost of disorganization scales nonlinearly.

What we do that helps

A few things make the messy roadmap less chaotic than it sounds.

We have a 90 day commit list. The 6 to 8 items on the commit list are the things we have actually committed to ship in the next quarter. Those items are clean. Engineering knows them. Sales can promise them. Customer success can plan around them.

Everything beyond the 90 day commit is on the messy roadmap. Anyone in the company can look at it. We do not commit to those items externally. We do not put them in sales decks. We talk about them as directional, not promises.

The transition from "messy item" to "committed item" happens once a quarter at a meeting we call roadmap court. Anyone can propose moving an item up. Anyone can challenge the proposal. We make decisions in the room. We then publish the new commit list on Friday.

This is the part that took the longest to figure out. The messy roadmap by itself does not work. The messy roadmap plus a clean 90 day commit, with a transparent transition process, is what works.

I think.

Ask me again in a year.