The give-back week reversed: volume +53% ($12.13M), unique traders +42% (26.3K), platform take +34%. More important than the bounce is the shape: the worst full day did $732K (prior week's worst: $325K), and the biggest day carried only 28% of the week (prior: 54%). That's the "rising floor, not taller spikes" pattern the last two decks demanded — and the #1 token by volume is now a utility coin (P0), not the founder meme. Two warning lights: take rate is still compressed at 0.46%, and Bonk overtook us on daily launch count.
From bags_trade_feed. This week (Jun 29–Jul 6) vs prior week (Jun 22–29). USD = trade notional, SOL @ $81.9.
The floor tripled. Daily volume: $2.16M → $732K → $1.85M → $1.58M → $3.35M (Jul 4) → $1.19M → $1.17M. Worst day $732K vs $325–500K last week; peak-day share of the week fell from 54% → 28%. Launches steady at 682 (695 prior) from 430 creators (+13%). Caution: 24h launch-count board now reads Pumpfun 29.4k · DBC 4.2k · Bonk 296 · Bags 116 — we slipped to #4. Per the 07-01 model review, count isn't the goal (graduation is) — but the optics matter on aggregator screenshots.
The public API confirms the recovery: 7d fees $119.4K (+46%), 7d revenue $59.7K (+46%). Lifetime fees crossed $61.55M. Internal platform take ($55.9K) reconciles with DefiLlama revenue ($59.7K) within ~6% — same tight match as the last three weeks.
| DefiLlama (public) | This week | Last week | |
|---|---|---|---|
| ● | Fees 7d 🕶️ | $119.4K | $81.9K |
| ● | Revenue 7d | $59.7K | $41.0K |
| ● | Fees 30d | $558.6K | $531K |
| ● | Fees 24h | $11.6K | $31.6K |
| ● | Lifetime fees | $61.55M | $61.4M |
Methodology unchanged: trading fees vs Bags DBC (pre-migration) + DAMMv2 (post-migration) pools, excl. Meteora protocol fee, LP fees, referral fees. Both internal & external say the same thing: the 06-29 give-back was a pause, not a break.
The bounce restored the revenue line but not the margin. This is the week's most important nuance.
Why take rate matters more than the bounce: at 0.8% take, this week's $12.13M volume would have produced ~$97K, not $55.9K. The compression costs us ~$41K/week — that IS the gap between a $2.9M and a $5M run-rate. The fix is mix, not luck: per the 07-01 model review, post-graduation DAMMv2 trades and the 96_LOCKED config (MC-decay fee schedule) carry structurally higher, more durable fees. Obj 2 stays: shift the mix, don't just grow the churn.
Net read: Last week's ask was "manufacture a runner." The market delivered one organically — P0 — and it's a utility token, which is the best possible answer to the casino critique. The job now is not to find the runner; it's to not fumble it (LP depth) and to systematize what made it work (proven-creator cohort, 96_LOCKED config, KPI visibility).
Our #1 token by volume ($1.39M/7d) has whales publicly complaining about LP depth. The immediate fix is a treasury/LP decision measured in hours. Then turn the mechanism into a SKU: graduated tokens hitting KPI thresholds get protocol-matched DAMMv2 depth. The 07-01 model review says deep pools are what retain buyers (96_LOCKED pools hold 7.5× more reserve and graduate 22× better).
Take rate compressed again (0.46% vs 0.8% peak) even as volume recovered. Ground truth from 8,714 config-labeled tokens: DEFAULT_96_LOCKED wins on graduation (22×), volume/token (2.25×), liquidity depth (7.5×) and post-grad fee durability (2× earners) — yet 50% of launches still use the worst-scoring DEFAULT. Make 96_LOCKED the default config; keep DEFAULT as opt-out. This is the single highest-leverage margin move we have.
The $PUMP unlock ($123.65M, +21% circ supply) lands Saturday. That's the counter-programming window — but only if we ship, not tease. Pick ONE ("SpaceX mode" staged float or the "time for things to change" reveal), give it a hard date this week, and kill or explicitly park the other. Two open teases + a live troll meme = follow-through debt compounding at 247 likes a pop.
P0 proving the utility thesis is only legible if the KPIs are visible: OpenRouter rank, inference volume, revenue, LP depth. A public per-token KPI page (bags.holdings style, live data) converts "trust me it's real" into a screenshot that travels. Also the natural home for the unclaimed-fees leaderboard ($ZACK playbook — data already in fee_share_user_claim_event).
The #1 performance predictor in the entire dataset: creators with a prior graduation graduate their next coin at 6.46% (5.5× average) with +57% volume/token. Launch count is a NEGATIVE signal. Build the "proven creator" fast lane: direct outreach, featuring, default 96_LOCKED, co-marketing. Bonk passing us on launch count is noise — this cohort is the signal.
Numbers are UP this week — lead with them, but keep the discipline: quality receipts, never name Pump, and land the product date before Saturday's unlock.
Cadence rule on an up week: receipts over victory laps. One anchor moment per day, product beats promises, and the unlock-day contrast only works if something real shipped by Thursday.
Goal reframed: $100M cumulative protocol revenue, and we start from what's already earned — $30.77M all-time (DefiLlama, protocol-revenue basis; $61.55M total fees incl. creator share). We're 30.8% of the way there. The remaining $69.2M is a throughput problem: volume × take rate. This week fixed the first factor (+53%); Obj 2 targets the second (0.46% → 0.8%).
| Milestone | Cumulative revenue | Increment | Status |
|---|---|---|---|
| Today (Jul 6) MEASURED | $30.8M | — | 30.8% of goal |
| End of Q3'26 TARGET | $36M | +$5M | needs ~$385K/wk |
| End of Q4'26 TARGET | $61M | +$25M | needs ~$1.9M/wk |
| $100M GOAL | $100M | +$39M | H1'27 at Q4 pace |
Basis: DefiLlama protocol-revenue (Bags' cut of trading fees, DBC + DAMMv2), all-time $30.77M — reconciles with internal fee_share_platform_accumulated weekly within ~6%. Current pace: $55.9K/wk platform take (≈$2.9M/yr). Honest gap: Q3's +$5M increment needs ~7× this week's take. The first $30M was earned in the launch supercycle — the next $69M has to come from structure: take-rate mix (Obj 2), durable runners (Obj 1/4), and the proven-creator funnel (Obj 5), not one hot month.
Quote-side notional, Bags pools. +53% WoW.
fee_share_platform_accumulated. +34% WoW.