Crown Jewels: Why Your Most Important Feature Is Slowly Dying
Build the Best Damn Feature in the Universe
Every product has a thing. Maybe two things. The capability that customers actually care about. The problem that they really need to be solved. The feature that closes deals. The reason people chose you over the other three options on the shortlist.
And somewhere in your organisation, nobody is allocated to making it great.
I know this because I see it constantly in my advisory work. A product team with a dozen features on the roadmap, none of which are the thing that actually matters most to their customers. The crown jewel—the capability that disproportionately drives revenue, retention, and competitive differentiation—is sitting in a corner of the codebase, unchanged for years, slowly dying of neglect.
Not because anyone decided to neglect it. Because nobody decided not to.
TL;DR: I took over as CTO at a company where reporting was the single most important capability to our customer base. We hadn’t touched the reports in five years. They were ugly, inflexible, and functionally frozen. Meanwhile, a competitor built out significantly better reporting and dashboarding. We spent a year losing every deal.
The fix was disciplined iteration: a beautiful (hardcoded) dashboard with the six most important reports in March, four more in May, customisation in July. By September we’d stopped the bleeding. But it didn’t have to be that way. If we’d maintained a pilot light of investment in our crown jewel all along, we’d never have been in that position.
The Slow Death of Your Best Feature
Here’s how it happens. The feature was built years ago—maybe by the original team, maybe by people who have since left. It works. Customers use it daily. Nobody complains about it because it does what they expect.
And therein lies the trap: because existing customers aren’t comparing it to the competition and it works the way they expected, nobody raises an alarm. The feedback loop is silent. No support tickets. No feature requests. No urgency.
Meanwhile, the product team is focused on new capabilities. There are eight stakeholders clamouring for new features. The roadmap is packed. Every sprint is allocated to something shiny and new. The crown jewel doesn’t appear on anyone’s backlog because it “works fine.”
The Vicious Spiral
Here’s where it gets dangerous. Because nobody touches the feature, the code becomes legacy. Because the code is legacy, engineers don’t want to work on it. Because nobody works on it, the knowledge of how it works fades from the team’s collective memory. And because it’s been “good enough” for so long, it’s probably deeply integrated in the guts of the system… other capabilities depend on the way it works, its data structures, its APIs.
Now you’ve got a vicious spiral. The feature is simultaneously the most important thing in your product and the thing that’s hardest to change. Every month of neglect makes the eventual reckoning more expensive.
This is exactly like KTLO (Keeping the Lights On) . If you forget to allocate resource to it, the debt builds and builds until it bursts and your entire roadmap is blown while you scramble to catch up. The only difference is that KTLO is about infrastructure and operational health. Crown jewels are about your competitive position.
When the Competitor Arrives
The day of reckoning comes—and it always comes—when a competitor does the thing slightly better.
Not dramatically better. Not a revolutionary reimagining. Just… slightly better. A cleaner interface. Better filtering. A dashboard that answers the question in two clicks instead of five. A mobile view that actually works.
Your sales team notices first. Loss ratios start creeping up. The competitor’s name keeps appearing in deal post-mortems. Prospects are saying “we like your platform, but their reporting is better.” Your sales team becomes frustrated, then demoralised, then angry.
And then the panic sets in.
The Panic Response
When your loss ratio starts climbing, the board panics. And when the board panics, we’re straight into the world of priority whiplash —the engineering team is suddenly pulled onto something completely different that wasn’t a priority a week ago.
But it gets worse. You already have a full backlog of committed work. So someone suggests bringing in an outsource partner to help address the thing that’s right at the heart of your value proposition. And while we’re at it, let’s modernise to a new tech stack. And maybe we should integrate that third-party reporting tool that product marketing found…
Hello, frankenproduct.
I’ve lived this. At the company I mentioned above, the product marketing team had already implemented a “wheeze”—a third-party reporting tool bolted onto the side of our product. It wasn’t well integrated. The data was slightly different from what customers saw in the core product. The UX felt horrific. The already disillusioned customer base lost even more confidence in our ability to solve their problems.
The outsource-and-modernise approach is tempting because it feels decisive. But you’re handing the most important part of your product to people who don’t understand it, while simultaneously asking them to rewrite it in a new technology. Now you have two problems: a legacy system you can’t maintain and a new system that doesn’t work yet.
How I Fixed It
When I took over, the reporting capability was a disaster. Five years without investment. Ugly. Inflexible. Losing every competitive deal on this one dimension.
Here’s what we did—and it’s the same pattern I described in Ship It and Move On :
March: Built a simple, beautiful, hardcoded dashboard page with the six most important pre-canned reports. Not configurable. Not flexible. Just the six things that customers looked at every day, presented in a way that was aesthetically excellent and immediately useful.
May: Added the next four most important reports. Still focused on what customers actually need, not what would look impressive in a demo.
July: Added customisation—letting customers configure the reports to their specific needs. This is the point where we moved from “good enough” to “genuinely powerful.”
September: We’d stopped losing deals. The competitor’s advantage had evaporated—not because we’d built some revolutionary reporting engine, but because we’d invested steadily in making our version excellent.
Six months. That’s all it took to fix. But it took a year of losing deals before anyone prioritised it. The total cost wasn’t six months of engineering—it was six months of engineering plus twelve months of lost revenue plus immeasurable damage to customer confidence.
None of that had to happen.
The Iterative Approach Was Key
Notice the pattern. We didn’t try to build the ultimate reporting platform in one go. We planned the second act —and the third, and the fourth. Each iteration was informed by what we’d learned from the previous one. Each one was small enough to deliver quickly but meaningful enough to change the competitive dynamic.
This is how great products are built. Not in a single heroic effort, but through disciplined, visible, protected iteration.
Identifying Your Crown Jewels
Before you can protect something, you need to know what it is. Here’s how to identify the two or three capabilities that disproportionately matter.
Win/Loss Analysis
Your sales team knows. They might not articulate it in product terms, but they know exactly which features close deals and which features lose them. If you’re not doing structured win/loss analysis, start. If you are, look at the patterns. Which capabilities are mentioned most frequently in losses?
Customer Usage Data
Look at what customers actually use, not what they say they want. The feature with the highest daily active usage is probably a crown jewel. The feature that correlates most strongly with retention is definitely one.
The “If This Broke” Test
Ask yourself: if this feature went down for a week, how many customers would leave? The features that would cause an exodus are your crown jewels. The features that nobody would notice are not.
Competitive Benchmarking
Regularly evaluate your crown jewels against the competition. Not annually—quarterly. How does your reporting compare to theirs? Your onboarding? Your core workflow? If you’re falling behind, you need to know before the sales team starts losing deals.
The Fix: Permanent Pilot Light
The solution is to ensure that your crown jewels always have some allocation. Not a huge investment—a pilot light. Enough to keep them current, competitive, and improving.
Tag It on the Roadmap
In RoadmapOne , you can tag objectives as a North Star Metric —the one metric that captures the core value your product delivers to customers. Your crown jewels should be tagged similarly. When they’re tagged, they’re visible. When they’re visible, they can’t be silently deprioritised.
The beauty of making this visible on a squad-by-sprint grid is that everyone—product, engineering, leadership, the board—can see at a glance whether your most important capabilities are being invested in or neglected. If the grid shows twelve months with zero allocation to your crown jewel, that’s a conversation that needs to happen now, not after you’ve lost a year of deals.
Allocate a Minimum Capacity
Even if it’s just half a sprint per quarter, ensure there’s always something allocated to your crown jewels. Use that time for:
- Competitive benchmarking: How does your capability compare to alternatives right now?
- Customer feedback synthesis: What are power users asking for? What workarounds are they building?
- Incremental improvements: Small changes that compound over time
- Technical health: Keeping the codebase modern enough that bigger changes remain feasible
This is the KTLO principle applied to your competitive position. Just as you wouldn’t run a platform with zero allocation to infrastructure maintenance, you shouldn’t run a product with zero allocation to your most important features.
The Ambition: Best in the Universe
Here’s where most “protect your core” advice falls short. It’s not enough to maintain your crown jewels. The ambition should be to make them so good that competitors literally cannot come close.
The team allocated to your crown jewel should be challenged constantly: build the best damn version of this feature in the whole universe. Not “keep it working.” Not “keep it competitive.” Make it incredible. Make it the reason customers choose you and the reason competitors give up trying to match you.
This is a fundamentally different mindset from maintenance. It’s offensive, not defensive. It’s aspirational, not protective. And it’s the difference between a product that survives and a product that dominates.
If you’ve prioritised correctly and your crown jewel genuinely addresses your most important business objective, then making it extraordinary isn’t a luxury—it’s the highest-ROI investment you can make.
The Evernote Warning
If you want a cautionary tale, look at Evernote. Instead of doubling down on what made it great—simple, powerful note-taking—the company chased shiny objects. Physical products. A Work Chat feature nobody asked for. Feature after feature bolted onto a product that was losing its identity.
The core experience got bloated. Performance slipped. The thing that customers originally loved—fast, reliable, searchable notes—degraded under the weight of everything else. Users left for Notion, Apple Notes, Google Keep. Not because those products were more feature-rich, but because they did the core thing better.
Evernote forgot which feature was the crown jewel. By the time they remembered, the market had moved on.
Signs Your Crown Jewels Are Dying
Watch for these early warning signals:
The feature hasn’t been touched in over a year. Check your git history. If the code for your most important capability hasn’t had a meaningful commit in twelve months, it’s in decline—even if it still “works.”
Engineers avoid the codebase. When you ask someone to make a change to the crown jewel and the reaction is a grimace and a sharp intake of breath, you’ve got a legacy problem building.
Sales starts qualifying around it. If your sales team has started steering demos away from the crown jewel, or saying “that’s on the roadmap for improvement,” your competitive position is eroding.
Customer workarounds appear. When customers start exporting data to Excel to do something your product should handle, they’re telling you the feature isn’t good enough. They’re just not telling you directly.
The competitor appears in loss reports. By the time this happens, you’re already twelve months behind. The time to act was a year ago.
The Board Conversation
If you’re a CPO or CTPO, this is a board-level conversation. Not because the board needs to micromanage feature investment, but because they need to understand a fundamental principle: your competitive moat requires ongoing investment.
Frame it like this: “We have two or three capabilities that disproportionately drive our competitive position. These are the features that close deals, retain customers, and differentiate us from alternatives. I want to ensure we always have a minimum allocation to these capabilities—not just maintenance, but continuous improvement. The cost of neglecting them is measured in lost deals and customer confidence. The cost of investing in them is a small fraction of our total capacity.”
No reasonable board will say no to this. They understand moats. They understand competitive advantage. They just need someone to connect the dots between “feature investment” and “revenue protection.”
Conclusion
Every product has crown jewels—the two or three capabilities that disproportionately matter to customers, close deals, and define your competitive position. And in every organisation I work with, those capabilities are at risk of slow neglect.
The failure mode is always the same: the feature works, so nobody complains. Nobody complains, so nobody invests. Nobody invests, so it becomes legacy. It becomes legacy, so nobody wants to touch it. And one day a competitor does it slightly better and your loss ratio spikes.
The fix is simple but requires discipline:
- Identify your crown jewels through win/loss analysis, usage data, and competitive benchmarking
- Tag them on your roadmap so they’re permanently visible
- Allocate a minimum capacity—always, without exception
- Challenge your team to build the best version of that feature in the universe—not just maintain it
Don’t wait until you’re losing deals. Don’t wait until the board panics. Don’t wait until you’re bringing in outsource partners to rewrite the heart of your product. Invest now, invest always, and make your crown jewels so good that competitors don’t even bother trying to compete.
Your customers chose you for a reason. Make sure that reason is still the best thing in the market.