Keep Network: Node Operators Explorative User Study

For infrastructure operators in the Web3 space, running a node is rarely an automated "set-and-forget" task. Keep Network operators must balance two full-time jobs at once, maintaining continuous machine uptime while managing highly volatile locked collateral. If local infrastructure drops, or even if the random peers they are grouped with misbehave, operators face capital slashing and irreversible liquidations.

The internal team assumed it understood how operators discover the protocol, keep up with updates, and manage risk. This generative study was launched to probe those assumptions: to chart how independent node operators actually navigate the network, and to find where the protocol's cognitive and financial load was quietly pushing them away.

"It was like a test for myself. Spinning up a node and running it and getting to know all the facts around that. It was like a quiz for me, like a test, and I liked that. In the first month it wasn't about money, it was about this puzzle and putting it together."

Uncovering the true operational workload of node operators required moving past stakeholder assumptions to direct user evidence.

The Methodology

Starting from stakeholder interviews and an assumption-gathering phase, I designed and facilitated 60-minute remote qualitative interviews, captured in Miro, transcribed, and synthesized into empathy maps and a low-level report in Dovetail. Topics ran from crypto discovery and DeFi risk profiles through the full node-running arc: learning, setup, monitoring, and liquidation events, plus sentiment on the upcoming Keep × NuCypher merger.

The Participants

3 self-operating node runners, out of 8 scheduled; recruiting this niche, time-poor audience proved brutal and most dropped out. All three were sophisticated, entrepreneurial, tech-savvy users, exactly the profile the protocol demands, since operating a Keep node requires balancing heavy technical work against live financial risk.

The Benchmark

The sessions overturned a core stakeholder belief: that operators track protocol changes through GitHub. None of them did. The study established a different operational baseline, operators are motivated by curiosity and yield, but on mainnet they all defected to third-party staking providers, because testnet slashing is theoretical and mainnet slashing is real money.

The interviews exposed significant misalignments between the platform's technical architecture and the everyday realities of live node management.

The GitHub Communication Fallacy

The Expectation: Node operators are deeply embedded developers who natively monitor GitHub to catch network updates.

The Reality: No participant used GitHub for updates; most operators are not developers at all. They pieced together protocol changes from Discord threads, Medium and blog posts, a noisy, unreliable pipeline they themselves called "bad communication". Some had been slashed or lost funds purely because an update never reached them. The need was unambiguous: a single source of truth for releases and protocol changes.

The Active Desktop Prison

The Expectation: Protocol parameters give operators sufficient economic control over their collateral.

The Reality: Liquidation was the operators' single biggest fear, and they felt helpless against it. With no alerting tools, they had to sit at their desktops watching collateral ratios, on their toes the moment the ratio dropped toward the 125% threshold. Some over-collateralized just to buy peace of mind; others noted that fellow operators would have avoided liquidation entirely "if they were paying attention". What was sold as passive income was, in practice, an active job.

"If you have some service of automated liquidity, avoiding liquidation, redeeming an automatic deposit, a redeeming pool or whatever, this would make everything easier for all node operators."

The Peer-Slashing Helplessness

The Expectation: Decentralized multi-node verification groups ensure bulletproof network trust.

The Reality: Multi-operator dynamics induced real psychological stress. Good actors felt vulnerable because the protocol bundled their financial safety with strangers: if the other two operators in a signing group misbehaved or went offline, everyone was slashed. Being competent was not enough to be safe.

The Squad Wall: Social Proof as Due Diligence

The Expectation: Operators discover and vet protocols through documentation, audits, and whitepapers.

The Reality: Investment decisions ran through what I named the "squad wall", a small trusted circle that shares information, splits research effort, and often pools capital. Vouches from friends, direct access to the team on Discord, and the founders' perceived competence mattered more than any document.

"Most of the projects that I look at, they're referred to me through someone I trust."

"I spent quite a bit of time talking to the team, so I understood the white paper and what it actually did. As I got to know Matt, I was like, he's a pretty smart guy. You know, back people that are smarter than yourself and things tend to work out pretty well."

The Automation Preference

Across their wider DeFi lives, all participants gravitated to protocols that quietly do jobs for them, self-repaying loans on Alchemix, auto-rebalancing pools on Balancer, and they wanted their rewards accessible and convertible into BTC and ETH, the assets they actually trust. The insight for Keep: every micro-job a product lifts off the user's shoulders is a retention feature, and complexity erodes the relationship.

The Persona Set: From Devs to Wealth Collaborators

The research reframed node operators not as pure developers but as financial participants with distinct operating modes: the Inquisitive Solo Tech, who runs nodes for the intellectual puzzle first and the money second; the Out-Sourced Staker, who can set up a testnet node but delegates mainnet operations to professional staking providers to protect capital and sleep; and the Wealth Cluster Investor, who splits costs, risk, and maintenance duties across a trusted inner circle.

The findings converted structural operator anxieties into a concrete set of product directions, centered on reducing cognitive load and building automated fallbacks.

A Single Source of Truth

Replace the Discord-digging ritual with one dedicated place for updates, releases, and protocol changes, with redundant distribution across channels so every operator profile is reached.

Automated Collateral Safety

Pursue automated liquidation-avoidance mechanisms, alerting, easier monitoring, and top-up flows, so operators can defend their positions without living at their desks.

Accessible Rewards

Make rewards easy to claim and convert into the blue-chip assets operators actually hold, removing the friction between earning and storing value.

A Prioritized Research Pipeline

The study closed by mapping five candidate research directions, documentation, node operation process, user growth, developer experience, and liquidity pools, feeding directly into the team's ideation and prioritization workshops.

Research Outcomes

The Communication Fix: Exposing the GitHub fallacy, and the funds lost to it, gave the team hard evidence that update distribution was a product problem, not a community-management nuisance, and reset the documentation and communication strategy.

The Roadmap Pivot: The liquidation-helplessness findings pushed monitoring, alerting, and automation up the priority list, and the persona set gave the team a durable model of who actually operates nodes, informing both this protocol's next iteration and my later node-operator research (Case Study 01).

Retrospective

This study's honest constraint is its sample: 3 of 8 scheduled operators showed up. I treated the output as directional, strong for hypothesis generation, insufficient for quantified claims, and it earned its keep by killing false stakeholder assumptions early and setting the agenda for the larger studies that followed. It also taught me to over-recruit aggressively for niche technical audiences.