A year ago, two roads diverged in a support-tool land — build or buy?

Most companies buy. It’s simpler, but the cookie-cutter experience boxes them into brittle, ticket-punching machines.

We don’t see support like that.

A user reaching out appreciating the product and support

Our mandate is to figure out what users want, deliver it with excellence. And we have to do that at tremendous speed and scale. So that’s what we did.

One year after shipping the original Help Station, we were supporting almost 1 million users, pushing 1500 conversations a week and seeing 50% upticks in daily thread volume.

But now people weren’t just asking us for help anymore — they wanted to share ideas, report bugs, and help us shape Railway. How did we get so lucky? We needed to push the boundaries of our support platform.

That brings us to today, where we’re introducing the all-new Railway Central Station. It’s no longer just for support (although of course we have that) but it’s also the beating heart of our community where everything happens from news to feedback to a brand-new public feature roadmap.

As our user base rapidly approaches the 1 million mark, the streams from which we ingest feedback are: 1) individually widening 2) multiplying in count

How do we know rather than guess what to build next?

We previously tried a roadmapping tool but it didn’t connect to Linear, our issue tracking tool, nor to our customer channels. We needed a way to condense the scattered channels of feedback and funnel them into a unified holding area for our team and users to look at

So, we built our own roadmap:

Railway’s public roadmap on Central Station

Since unweighted user upvotes aren’t granular enough to help us pick between many competing priorities, we needed primitives to weight our Priority Tetris board:

• How much do we stand to unlock, in revenue, when we ship a feature?
• How do we maintain a granular view into feedback as we scale 1 → 10 → 100 million users?

Knowing, rather than guessing, what to build buys us the leverage we need to scale faster, with less.

As a company grows, the product matures. Most of the “must haves” are shipped by the OG squad at lightning speed and with thunderous conviction because early on, it’s easy to know what to build. But then we run into the age-old question:

What’s the need of the hour?

A plethora of decisions, each with their own opportunity cost and there’s too many variables to juggle. If we have the deeply embedded context from building our own roadmap and we have all our tickets flowing from the product to this roadmap, could we automate the prioritization matrix and leapfrog the decision making?

Yes. Let me show you how. Here’s where we’re headed.

As a business, the answer to: “How much do we stand to unlock, in revenue, by shipping a feature?” is one of the star ingredients in our feature prioritization recipe. We track engagements (like upvotes and comments), subscriptions, team size, average platform spend over the last three months, and projected spend based on current usage. Then we tie these pieces together to derive the current financial upside to a feature. Allowing users to self-disclose how important a feature is to them further fine tunes the signal:

Not having Feature X…

• “…significantly blocks us from fully migrating to Railway”
• “…significantly blocks us from expanding on Railway”
• “…would be a nice to have / QoL Improvement”

Going a step further, opening the roadmap to the top of our funnel, factoring in any feature requests by ongoing sales engagements, indexed by their Total Addressable Spend helps pin “Feature X unlocks sales prospect A”. It doesn’t guarantee they’ll come onboard but lays down tracks to pave a path and unblock our GTM team to move the conversation forward with them.

Since our feature development is heavily demand based, what if we let users pre-order features? We share updates that get them excited, their excitement gives us joy as we build, and we bring them along the feature development journey with us.

Imagine live product development — you get to see the chefs cooking in the kitchen:

User suggests a feature → the community engages with it → feature is moved to “Planned”, we emit a notification → we create a Linear project, DRI gets cooking → the feature is moved to “In Progress”, another notification emitted → the DRI brings the feature to life → all subscribers find out immediately.

The result? Customer love.

When someone drops a comment, it’s piped into Linear where the project DRI can engage with you directly. The feature discussion thread serves as the glue for our design + engineering teams to keep their ears to the ground and know exactly what you’d like, while we build it for you.

Together, all this enables:

• Users submit tickets, we can map them onto roadmap items
• Users can upvote existing roadmap items (or create new ones)
• We aggregate them and sort them by how much users are spending
• As we move into building, the users who asked are now our beta groups
• We work with them as design partners, and then we roll out the feature

Layering in sales, we can add in “Deal X is blocked by feature Y”, which will further help us prioritize the roadmap.

We own the support stack and see the roadmap as an opportunity to create a crisp product development lifecycle to deliver a super high touch, gratifying experience to users. Building a solution from the ground up let us build a completely different organization. Where engineers and designers are empowered to work directly with users, to build something magical, and close the loop ASAP, at scale.

Central Station allows us to deeply nurture user connections — if you drop a support thread, someone who genuinely cares about your deploys, your workloads, your things, will jump in.

Why does it help us do that? Because of context aggregation and the fact that when we’re on support rotation, we are thinking “product” not “tickets”.

That’s our founder btw, not during their on-call rotation (it was mine)

Angelo reaching out to one of our users

Central Station is evolving from support to community, but this by no means is its final form, nor the final destination on this ride. Besides, it’s not about the destination, it’s about the quality of the journey — the one we’re on together with each other, with our users, and there are miles to go before we sleep.

So, come as you are, drop a feature into our roadmap, give us feedback and help make this better. Gather your teammates, friends, and put your weight upvote behind it. We’re super excited to be looking back a year onwards from today and seeing what we’ve brought from our humble farm to your table.