Robert Leshner built Compound to make lending and borrowing as simple as a savings account.
The mechanics got there. The interface still has not.
Compound was one of the first DeFi protocols to show health factors, utilization rates, and borrow APY in a single dashboard. It was also one of the first protocols where users got liquidated because they did not understand what those numbers meant.
The data was accurate. The dashboard worked. The problem was that showing a metric is not the same as explaining its consequence.
What Aave got right — eventually
Stani Kulechov has been consistent about what actually drives DeFi adoption: if interacting with a protocol feels like a savings app, the audience widens from thousands to millions.
Aave's early dashboard and its current one show the same underlying data. What changed is how it is framed:
Before — protocol state
Supply APY: 3.2%
Borrow APY: 5.8%
Health Factor: 1.34
Loan-to-Value: 74%
Liquidation Threshold: 77%
After — user consequence
You're earning ~$32/month on your $1,000 deposit
Your position is safe.
Collateral can drop 26% before liquidation triggers.
You can borrow up to $740 against this position.
Same numbers. Completely different job.
That shift — from state display to decision support — is the entire UX challenge in DeFi.
The counterargument worth taking seriously
Andre Cronje has argued the opposite: accessible DeFi dashboards create false confidence. If you are using a lending protocol, you should understand what a health factor is.
He is not entirely wrong.
A dashboard that shows You're safe ✓ when the health factor is 1.1 is worse than one showing the raw number. There is a version of "simplification" that removes critical information under the guise of clarity.
But Cronje's argument cuts in the opposite direction from where most DeFi teams actually land. The problem is not that dashboards are too simple.
Most DeFi dashboards are too raw. They show protocol state without user context. The choice is not between raw data and hidden data — it is between raw data and translated data.
Health factor: 1.4 is raw data.
Your position may be liquidated if ETH falls 18% is translated data.
Both are honest. One is useful.
Why dashboards are built for the wrong person
The structural reason DeFi dashboards are hard to use is straightforward: they are built by engineers who understand the protocol, for users encountering it for the first time.
What the engineer already knows
-
Health factor < 1.0= liquidation is imminent -
Utilization at 89%= borrow rates are about to spike -
Transaction failed= almost certainly a gas estimation error, not a contract failure -
Approval required= this is a two-step process before the real action
What the user sees
A wall of numbers with no explanation of what they mean or what to do about them.
None of the engineer's knowledge is on the screen — because it is too obvious to write down.
The fix is not more designers. It is having a non-technical person attempt the first five actions while someone watches. That 30-minute session will surface more than three months of internal design review.
What Uniswap's swap interface got right
Uniswap does something most DeFi dashboards do not: it answers "so what" inline.
Before you execute a swap, the interface shows:
You pay: 1.2 ETH
You receive: ≥ 2,832 USDC (minimum after slippage)
Exchange rate: 1 ETH = 2,847 USDC
Price impact: 0.12% ← green, safe
Slippage: 0.5% tolerance
Network fee: ~$3.20
This is not a general education section. It is specific to your transaction, shown before you commit.
The contrast with most DeFi dashboards: they show global protocol metrics and leave the user to calculate the implications for their specific position. Uniswap shows position-specific consequences and lets the user decide.
A 2% price impact on a $100 swap is a different decision than 2% on a $50,000 swap. A dashboard that shows "2% max slippage" globally is less useful than one that says "this specific trade has 1.4% expected slippage at current liquidity."
The "so what" test
For every major metric on a DeFi dashboard, apply one discipline: ask "so what?"
If the interface cannot answer, the metric is not ready for the primary UI.
APR is 14%.
At current rates, $10,000 earns ~$118/month. If rates drop to 8% next week, that drops to $67/month.
Utilization is 91%.
Borrow rates may increase. Withdrawals could face delays if utilization reaches 100%.
Health factor is 1.3.
Your position is vulnerable to a 12% drop in collateral value. At current ETH price, that means liquidation triggers below $2,992.
Gas is 45 gwei.
This transaction costs ~$8 right now. It cost $2 yesterday morning.
Lock period: 7 days.
You cannot withdraw before June 1.
If the dashboard cannot surface these translations, put the raw metric in a details panel. Keep the primary view for decisions.
What this means practically
Organize around decisions, not metrics. What can I do? What should I watch? What changed since last time?
Translate every major metric into consequence. Not
Health Factor— distance to liquidation in dollar terms.Make risk specific and contextual.
DeFi involves riskis useless.Your position can be liquidated if ETH falls below $1,840is actionable.Design the empty state as onboarding. A new user's first dashboard should explain what the product does — not show zeros and a connect button.
Design the post-transaction screen. After a deposit, the user has immediate questions: Did it work? Is it earning? Can I withdraw? Answer them before they ask.
Run the "so what" test before shipping every metric. If you cannot complete the sentence, the metric is not ready.
DeFi dashboards will not improve by showing more data. They will improve when teams stop treating the dashboard as a protocol readout and start treating it as a decision support tool.
The protocol can stay complex. The decision should not.
I work as a Web3 creative director helping DeFi teams design interfaces that make complex mechanics legible. somaryuu.xyz
