MCC offers a real-time comic estimate (average latency 4-12 minutes) from multiple sources (eBay sold, Heritage, ComicConnect, WhatNot) with extensive VF/VO coverage (Lug, Semic, Panini, Urban). GoCollect ($89/year) remains the American CGC rating reference with official census but daily latency and VF catalog almost absent. MCC wins on freshness, French-speaking coverage and insurance use cases; GoCollect wins on depth census CGC vintage US.
Estimating a Strange 100 Lug 1978 at €85 or a Daredevil 181 CGC 9.6 at €1,400 without fresh and localized data leads to the wrong decision: overpaying for a purchase, underestimating an inheritance, declaring an inappropriate value for home insurance. Two approaches oppose each other in 2026 on the comics rating market:MCC (My Comics Collection), a French platform which calculates a multi-source real-time estimate with native VF/VO coverage, andGoCollect, an American veteran founded in 2008 which built its reputation on the official CGC census and the proprietary FMV (Fair Market Value) calculation updated daily. The choice of price source determines the quality of each insurance report, the relevance of purchase decisions and the accuracy of estimates before resale.
The positioning gap structures the comparison: MCC bets on freshness (average latency 4 to 12 minutes compared to 24 to 48 hours at GoCollect), multi-marketplace aggregation including WhatNot (a fast-growing channel ignored by GoCollect) and especially the VF catalog coverage which natively integrates Lug, Semic, Panini Comics and Urban Comics — segments where GoCollect remains almost empty. GoCollect maintains a decisive advantage on the official American CGC census and on the 5-year historical depth of premium data. For a French-speaking collector who is hesitating between the two sources before an €800 purchase, an insurance report for a €25,000 portfolio or an urgent sale, understanding the real methodology of each tool determines whether the decision will be right or wrong by 8 to 18%. This comparison details the methodology, price sources, update latency, catalog coverage and three quantified use cases on which the two tools clearly diverge.
Comics estimation problem in 2026: latency, VF coverage, multi-channels
The 2026 comics rating market suffers from three structural weaknesses which particularly penalize the French-speaking collector. First, latency: the majority of historical platforms (GoCollect, GPA, ComicHub) update their ratings once a day, often in UTC nightly batches. In a market where a record sale can move a modern comic by 15 to 30% in a few hours (film trailer effect, casting announcement, Marvel Studios rumor), working with a quote that is 18 to 36 hours old is equivalent to negotiating blindfolded. On modern speculative comics (House of
Second, VF catalog coverage remains anemic on American bases. A Strange 1 Lug 1970 (first Doctor Strange VF), a Spidey 23 Lug with central poster, a Titans 35 Lug, an X-Men Special Lug — so many pieces actively collected in France on which GoCollect does not offer any odds. The collector must then switch to ManquesetPertes.com, Bedetheque or the Le Bon Coin listings to reconstruct a manual estimate, a time-consuming and imprecise exercise. The post-2007 Panini Comics editions (Marvel Now, Marvel Knights, 100% Marvel ranges) and post-2012 Urban Comics (Batman, Justice League VF) remain treated as best efforts on a few scattered cards without a structured rating system.
Third, multi-channel aggregation is struggling to integrate WhatNot, a live-stream channel growing by 180% over 2024-2025 which now represents 14% of the global volume of CGC comics transactions according to 2026 estimates. WhatNot does not publish an official sold listings feed, which forces aggregators to scrap video replays or collect sales declarations, a costly methodology that GoCollect has not yet integrated. For a collector following the real market in 2026, ignoring WhatNot means losing 1 in 7 transactions.WhatNot fee structuremakes the channel particularly attractive on the seller side, which amplifies the volume migration effect. Beyond that, the European auction house market (Cornette de Saint Cyr, Tajan, Catawiki) remains completely off the GoCollect radar, even though it represents the No. 1 channel for premium vintage VF comics where Strange 1s reached €6,200 in the 2025 sale.
How MCC calculates its real-time rating: weighted multi-source architecture
MCC has built its rating engine around a weighted, continuously updated multi-source aggregation logic. Four main flows feed the calculation: eBay sold listings (50% of the weight for US comics, 35% for VF comics), Heritage Auctions (20% on premium vintage US comics), ComicConnect (12%) and WhatNot via video scraping replay (18% since integration October 2025). On VF comics, two additional flows enrich the calculation: ManquesetPertes archives + Bedetheque declared transactions (cumulative 22%), and French-speaking auction houses via Drouot Live and Catawiki feeds (cumulative 8% on average, up to 35% on rare vintage Lug pieces). This differentiated distribution by segment (US vs VF, vintage vs modern) makes it possible to avoid the bias of a single source in segments where it does not have real market dominance.
The average latency measured between the closing of a public sale and the impact on the MCC rating is 4 to 12 minutes depending on the source. Completed eBay sales are captured via API feed + push notification as soon as the auction is finalized, processed by an automatic classification module which validates the declared CGC grade, the exact title (series + number + variant), the certification code when present. Heritage and ComicConnect sales arrive via official webhooks with average latency 2-5 minutes. WhatNot remains the most latent source (video scraping replay, average latency 8-25 minutes depending on the length of the live). This continuous architecture contrasts with the GoCollect daily nightly batch, where a Heritage sale closed at 7 p.m. New York appears in the quote around 7 a.m. the next morning (delta 12 hours minimum).
The weighting engine applies three additional corrections. Temporal degression: decreasing weight over 90 days with exponential discount (today's sale weighted 1.0, 30-day sale ×0.6, 90-day sale ×0.15). Exclusion of outliers: sales outside the range ±2.5 standard deviations automatically excluded from the calculation. CGC inventory pressure: algorithmic discount on modern comics whose census increases by more than 10% in a rolling quarter, an anticipatory signal of reversal of odds. On VF comics where the CGC census does not exist, MCC uses a population estimation proxy based on historical publisher print runs crossed with active eBay/Catawiki listings. To check acollection value in PDF insurance report, this multi-source weighting becomes an argument of robustness against the insurer.
GoCollect methodology: official CGC census and proprietary daily batch FMV
GoCollect has structured its offer since 2014 around the CGC partnership which gives it daily access to official census data: number of copies graded by number, by grade (from 0.5 to 10), by variant, by signature. In the first quarter of 2026, the GoCollect database totals more than 2.8 million comic references with 1.1 million active CGC census records updated daily. This census integration remains the central differentiating force: knowing that an Amazing Spider-Man 252 CGC 9.8 exists in 4,280 copies worldwide radically changes the estimate compared to a grade where only 180 copies are listed. The rating directly integrates this parameter via the proprietary FMV calculation.
GoCollect price sources are based on four channels: eBay sold listings (approximately 60% of volume), Heritage Auctions (20%), ComicConnect (12%) and ComicLink (8%). WhatNot integration remains absent in 2026 despite recurring announcements since 2023, which creates a blind spot of 14 to 18% of real global volume. Transactions declared by Premium users (Submit Sale option) complete at the margin but remain statistically minor (less than 2% of the volume processed). The calculation methodology applies a 90-day weighted average with exclusion of outliers ±2 standard deviations, then census correction via a proprietary algorithm which depreciates the rating by 5 to 12% when the census increases by more than 15% in 6 rolling months.
The update latency remains the structural weak point of the GoCollect architecture: daily nightly batch at 3 a.m. UTC which consolidates the data collected over the previous 24 hours. An eBay sale closed at 4 p.m. Paris time will appear in the GoCollect rating the next morning around 4-5 a.m., an average delay of 12 to 14 hours. For the collector who practices arbitrage on speculative modern comics, this delay creates exploitable error windows: an Ultimate Spider-Man 1 (2024) CGC 9.8 saw its price rise from $230 to $380 between January 12 and 14, 2025 following the announcement of an adaptation project. GoCollect only displayed the $380 rating on the morning of the 15th, when MCC displayed it on the 13th at 6 p.m. after detecting the first five sales above $320. The information differential reached 48 hours on an asset where arbitrage opportunities rarely last more than 72 hours before the market rebalances.
VF/VO catalog cover: Lug, Semic, Panini, Urban vs native US catalog
The catalog cover opposes two philosophies. MCC has invested specifically in the VF segment since its creation with a catalog which includes 100% of the Lug series 1969-1992 (Strange, Spidey, Titans, Nova, (Marvel Now, Marvel Knights, 100% Marvel, Deluxe, Special Edition ranges) and 100% of post-2012 Urban Comics in the DC segment (Batman, Justice League, Wonder Woman, Watchmen reissues). In total, around 38,000 VF references at calculated rating, compared to less than 800 at GoCollect (mainly a few Hachette Marvel reissues with international distribution).
This integrated VF coverage enables a key use case: portfolio diversity. A typical French collector owns on average 64% of VF (childhood heritage + adult acquisitions on emotional Lug/Semic titles) and 36% of VO (targeted acquisitions key issues Marvel/DC). On GoCollect, this collector can only estimate 36% of his portfolio — the VF portion remains off the radar and must be processed manually. On MCC, the estimate covers 100% of the portfolio with a unified report and subtotals per publisher. For a cumulative heritage of €18,000 including €11,500 in VF (Strange 1, Spidey 2, Titans 1, Strange Special 1) and €6,500 in VO (Amazing Spider-Man 252, Hulk 340), the difference in operational usefulness between the two tools is of the order of 3 out of 10 vs 9 out of 10.
On the US catalog side, MCC covers around 1.9 million comic references (vs. 2.8 million at GoCollect), with priority on key issues, premium variants and speculative modern comics. The gap of 900,000 references mainly concerns low-liquid silver/bronze age back-issues (series B and C, romance comics, war comics, western Charlton) where the odds difference between the two tools remains statistically marginal because the annual transaction volume sometimes does not reach 5 sales per reference. For 99% of collector use cases (emotional comics + invested key issues + insurance portfolio), MCC coverage is functionally complete, with the added VF advantage. For the exhaustive researcher who invests in marginal series, GoCollect maintains an advantage of depth of documentary catalog.
Latency Update: MCC live vs GoCollect daily batch
Latency is the most measurable differentiator between the two tools. A test conducted over the period November 2025 - February 2026 measured, on 60 sample sales (40 US + 20 VF), the time between the public closing of the transaction and the impact on the displayed rating. Aggregated results: MCC has a median latency of 7 minutes (average 11 minutes, 90 percentile at 24 minutes), GoCollect has a median latency of 14 hours (average 16 hours, 90 percentile at 22 hours). The structural gap reaches 120 times in median and 80 times on average, a gap which changes the nature of use: MCC functions as a real-time market terminal, GoCollect functions as a documentary database updated daily.
The operational impact of this latency is measured on three types of events. First category: Marvel Studios / DC Studios announcements (trailers, castings, release dates). In 2025, six major announcements (Thunderbolts trailer, Avengers Doomsday casting, Brave New World date, Fantastic Four First Steps first extract, Daredevil Born Again episode 7, Blade reboot) triggered rating increases of 18 to 47% on the key issues concerned in less than 48 hours. MCC captured the increase on average 6 hours after the announcement; GoCollect published the new rating an average of 38 hours later. 32-hour purchase arbitrage window for those following MCC, zero window for those following only GoCollect.
Second category: CGC grading waves on modern comics. When 200 new CGC 9.8s enter the census on an Ultimate Black Panther 1, the pressure of supply causes the rating to drop by 15 to 25% in 7-10 days. MCC depreciates the price from the first wave of discounted sales (day D+2 to D+4), GoCollect waits for the official census to be updated (day D+10 to D+14) which creates a structural delay of 6 to 10 days on bearish reversals. Third category: French-speaking auction houses (Drouot, Catawiki Premium) where VF records are sold for €3,800 on a Strange 1, €4,200 on a Spidey 1, €6,200 on a Titans 1 – events completely invisible to GoCollect and captured by MCC in 12 to 36 minutes via Drouot Live feed and Catawiki notifications. Forsell quickly without losing money, this difference in information is literally worth hundreds of dollars on a single premium lot.
Use case 1: home insurance and asset valuation report
The home insurance report imposes specific requirements: valuation per room with estimate date, documented price source, traceable methodology, temporal validity of less than 12 months. The average French insurer (Maif, Maaf, AXA, Allianz, GMF, Groupama) accepts three types of supporting documents: original purchase invoice, paid certified expert report (typically €80-180 per item), or estimation report from a recognized rating platform with transparent methodology. For a comic collection heritage valued between €8,000 and €50,000, the certified expert route becomes economically absurd (€1,200 in appraisal fees to value €12,000 of collection), which pushes towards the rating platform solution.
MCC produces a time-stamped PDF insurance report which aggregates: exhaustive list of pieces with title, number, publisher (Marvel US, Lug, Panini), grade when applicable, estimated unit value with weighted source (eBay sold The native inclusion of VF coins (Lug, Semic, Panini, Urban) which make up 50 to 70% of a typical French portfolio makes this report usable without supplement. GoCollect also produces a Premium PDF report but limited to the referenced US coins, which forces the French collector to manually complete the 50-70% of VF by ManquesetPertes or Bedetheque estimates, losing methodological consistency in the eyes of the insurer.
On a real case documented in 2025 (total portfolio €24,800 including €16,200 VF + €8,600 VO), the unified MCC report allowed a declaration accepted at first instance by AXA without request for additional documents. The same collection processed via GoCollect + manual VF reconstruction required three round trips with requests for additional supporting documents, spread over 11 weeks, to achieve coverage deemed acceptable but underestimated by 14% compared to the actual value. The operational difference is quantified: on €24,800 declared in real value vs. €21,300 in underestimated value, the difference in annual premium remains minor (typical delta 4-12 €/year), but above all the difference in compensation in the event of a total loss reaches €3,500, or 41 times the annual MCC subscription cost. To manage the portfolio beyond the report, thetotal collection value in monthly monitoringremains the natural complementarity.
Use case 2: pre-sale estimation and pricing arbitration
The pre-sale estimate requires a simple question: at what price should the listing be published to maximize the margin without blocking rotation? Three critical variables: the current quote (live, not yesterday), the market depth (how many active buyers on this reference this month?), and the time window of opportunity (the market is going up or down?). MCC renders all three in real time on the same sheet: estimated rating updated by the minute, number of active cross-marketplace listings (eBay + WhatNot + Catawiki), 90-day evolution graph with directional trend. On a Daredevil 168 CGC 9.6, the MCC sheet in February 2026 displayed: rating €850, 4 active eBay listings between €780 and €920, trend +6% over 30 days, 2 recent WhatNot sales at €810 and €840. This density of information makes it possible to publish at €895 with acceptable trading at €840, i.e. maximum capture of the bullish window.
On the same reference, GoCollect would have returned an FMV rating of €820 based on the 30-day average, without visibility on active listings or WhatNot sales over the last few hours, which would have guided the seller towards a listing at €850 — underpricing of €45 on the real potential. Multiplied by 30 annual sales, GoCollect systematic underpricing reaches €1,350/year for an average reseller, or 16 times the MCC annual subscription cost. For modern speculative comics (Walking Dead 1, Saga 1, Something Is Killing The Children 1, House of X 1 variants), where bullish windows often last 5 to 15 days, the optimal pricing gap between MCC and GoCollect widens even further.
The opposite also works: on bearish reversals (grading wave which floods the market), MCC captures the fall in 48 to 96 hours, GoCollect in 7 to 14 days. Selling an Ultimate Black Panther 1 CGC 9.8 on day D+3 of a grading wave at €280 via MCC rating, vs selling it on day D+12 at €215 according to late GoCollect rating, represents a €65 difference per piece. The professional reseller who processes 8 to 15 speculative modern comics per month recovers 520 to 975 €/month of timing difference in favor of MCC. To calibrate the listing price, cross MCC with historical salesComicConnect vs. Heritage Auctionsremains recommended on premium pieces above €2,000.
Use case 3: monthly asset monitoring and portfolio reporting
Monthly asset monitoring responds to a different logic: it is less about arbitrating a single transaction than measuring the value trajectory of the portfolio over 12-36 months and informing reallocation decisions (selling a part to buy a premium piece, balancing VF/VO, securing latent capital gains before market reversal). MCC produces a monthly monitoring dashboard with total value recalculated at the end of each month, breakdown by publisher (Marvel US, Lug, Panini, Urban, DC US), breakdown by segment (silver age, bronze age, modern), top 10 contributors to value, top 5 winners and losers of the month, comparison vs month M-1 and year N-1.
The granularity of MCC reporting exceeds that of GoCollect on two key dimensions. First, the native VF/VO segmentation which does not exist at GoCollect: a collector with 60% VF can visualize the separate evolution of each sub-portfolio, identify that the 1969-1975 Lugs have increased by 14% while the 2012-2018 Paninis have stagnated at +1%, and arbitrate their next purchases accordingly. Secondly, the 36-month historical comparison depth available from the standard MCC offer, compared to 12 months on GoCollect Premium and 60 months only on GoCollect Pro (€188/year). For a monitoring horizon of 3 to 5 years, MCC delivers the equivalent of the GoCollect Pro offer at a standard rate.
The MCC monthly report is exported as a time-stamped PDF with an automatically generated "Trend Analysis" section: synthetic textual commentary on the 3 main movements of the month (e.g.: Strange 1 +8% on Catawiki record sale, Daredevil 181 -4% on wave grading, ASM 252 stable). This narrative analysis remains absent from GoCollect which only restores the raw figures without interpretation, which forces the collector to mentally reconstruct the causes of the variations. For the collector who devotes 30 minutes per month to his reporting, the time saving represents 6 hours per year, equivalent to the opportunity cost of the subscription. To manage over time, themonthly price tracker on portfolioremains the natural extension. Long-term monitoring is also aligned with thecomics investment strategy 2027.
Final comparison table and recommendation by profile
The table below summarizes the seven key criteria discussed in the previous sections, restoring the concrete advantage by dimension.
Criterion 1 — Updated latency:MCC 4-12 minutes (median 7 min), GoCollect 12-36 hours (median 14 h). MCC advantage, 120× difference in median.
Criterion 2 — Integrated price sources:MCC = eBay sold + Heritage + ComicConnect + WhatNot + ManquesetPertes + Catawiki/Drouot, GoCollect = eBay sold + Heritage + ComicConnect + ComicLink. MCC advantage on the French-speaking market and on WhatNot, GoCollect equivalent on US hotels.
Criterion 3 — VF coverage:MCC 38,000 references (Lug, Semic, Panini, Urban), GoCollect less than 800 references. MCC advantage without competition in this segment.
Criterion 4 — Original catalog cover:MCC 1.9 million references, GoCollect 2.8 million. GoCollect advantage on marginal silver/bronze back-issues depth, functional equivalence in 99% of use cases.
Criterion 5 — Official CGC Census:MCC = proxy estimated via publisher population + active listings, GoCollect = daily official CGC census. GoCollect advantage, review on the analysis of premium vintage US speculative modern comics.
Criterion 6 — Unified VF/VO insurance report:MCC = native unified report accepted in first instance, GoCollect = US limited report + manual VF reconstruction. MCC advantage on the French-speaking insurance use case.
Criterion 7 — Historical asset reporting:MCC = 36 months standard, GoCollect = 12 months Premium / 60 months Pro. Functional equivalence for a 1-3 year horizon, GoCollect Pro advantage for a 4-5 year horizon.
Profile recommendation 1: mixed French-speaking collector VF/VO (portfolio 50-300 pieces).MCC as the main tool, possibly supplemented by a free GoCollect account for occasional census CGC consultation on the 5-10 premium US coins in the portfolio. This stack maximizes catalog coverage and odds freshness, while maintaining free access to the official CGC census for premium arbitrages.
Profile 2 recommendation: 100% US vintage CGC collector (silver/bronze age premium wallet).GoCollect Premium as the main tool (CGC census + historical depth), supplemented by MCC for occasional consultation on Heritage and WhatNot sales in the last hours before arbitration. The census remains the decisive argument in this segment.
Profile 3 recommendation: multi-segment professional reseller (volume 100+ transactions/year).Combination of the two tools in a complementary stack. MCC for live pricing and timing arbitrage on speculative modern comics and VF. GoCollect Pro for advanced grade-by-grade reporting on premium US coins and access to in-depth CGC census. The ROI is profitable from the first season on the pricing gap captured.
Profile 4 recommendation: beginner with wallet less than 30 coins.MCC in a single tool. The native VF coverage covers 100% of the typical French beginner portfolio (Lug childhood acquisitions + 5-10 key issues Marvel purchased as adults), the unified insurance report is sufficient for home declarations, and the monthly educational reporting helps to understand market dynamics. The absence of an official CGC census does not weigh on this profile as long as the portfolio remains outside silver age US premium. To prepare for what follows, thecollector's guide France pillargives the fundamentals. To get started without commitment, thefree estimateallows you to test the coverage on 3 rooms before subscription. To explore the referenced catalog, the sectioncomicsoffers the main navigation. To then approach resale, theguide to selling resale pillarstructures the entire process.
FAQ — MCC Estimation vs GoCollect
MCC or GoCollect: which one to choose for a French mixed VF/VO collector?
MCC systematically remains more relevant for a French mixed VF/VO collector because its catalog cover natively integrates 38,000 VF references (Lug, Semic, Panini Comics, Urban Comics) compared to less than 800 at GoCollect. A typical French portfolio composed of 60-70% VF cannot be estimated coherently on GoCollect without time-consuming and imprecise manual reconstruction. MCC real-time latency (median 7 minutes vs. 14 hours at GoCollect) adds a decisive operational advantage for live arbitrage, and the unified assurance report accepts VF + VO coins on the same time-stamped document. The free GoCollect account remains useful as a complement for the official CGC census consultation on the 5-10 premium US coins in the portfolio, but as a main tool it structurally penalizes the French-speaking collector. The difference in operational usefulness between the two tools on this profile reaches 9 out of 10 for MCC compared to 3 out of 10 for GoCollect.
Is MCC real-time latency really 120 times better than GoCollect daily batch?
On modern speculative comics (House of X 1, Ultimate Spider-Man 1, Walking Dead 1, Saga 1) where arbitration windows often last 24 to 96 hours, yes without reservation. The timing difference between MCC (quotes updated in 4-12 minutes) and GoCollect (quotes updated in 12-36 hours) makes it possible to capture or lose 6 to 25% margin on buying or selling depending on the direction of movement. On vintage silver age comics with very low liquidity (less than 10 documented annual sales), the advantage of latency is reduced because the odds change little month to month. Latency nevertheless remains useful for capturing records signed in French-speaking auction houses (Drouot, Catawiki Premium) which are completely invisible on GoCollect. On a weighted average over a typical mixed portfolio, the operational advantage of MCC latency amounts to 8 to 15% annual margin gain for an active reseller, a ratio well above the subscription gap.
Can the official GoCollect CGC census be replaced by an MCC estimation proxy?
Partially. The MCC proxy based on the historical publisher population crossed with active eBay/Catawiki listings provides an acceptable approximation for 80% of common use cases, with a measured median deviation of 12-18% compared to the official CGC census on a sample of 50 modern references. On modern post-2015 comics where the grading ratio remains high and the dynamic census critical, the official GoCollect census maintains a clear advantage in anticipating waves of grading and their impact on the rating. On vintage silver age US comics where the census evolves slowly (less than 3% per year), the MCC proxy is more than sufficient. For the mixed French collector, the free GoCollect account used for occasional consultation on 5-10 premium US coins provides access to the official census without additional subscription costs, the most efficient strategy.
Which product tool is the most relevant home insurance report for a mixed VF/VO collection?
MCC delivers a unified insurance report that covers 100% of a mixed VF/VO portfolio on a single time-stamped PDF document, with weighting methodology in the appendix, documented price source and estimate date. This unification facilitates acceptance at first instance by French insurers (Maif, Maaf, AXA, Allianz, GMF, Groupama) without requesting additional documents. GoCollect produces a Premium report limited to referenced US coins, which forces the French collector to manually complete the 50-70% of VF by ManquesetPertes or Bedetheque estimates, a time-consuming exercise which creates methodological inconsistencies which are penalizing for the insurer. On a real documented case 2025 (portfolio €24,800), the unified MCC report was accepted at first instance, while the GoCollect solution + manual reconstruction required three round trips over 11 weeks to achieve an underestimated coverage of 14%. For the French-speaking insurance use case, MCC remains without functional equivalent.
Should we combine the two tools in a stack and at what annual budget?
The combined stack is justified for two main profiles. First profile: the multi-segment professional reseller with an annual volume greater than 100 transactions, which combines MCC for live pricing and timing arbitrage, plus GoCollect Pro (€188/year) for advanced grade-by-grade reports and in-depth CGC census on premium US coins. The ROI is profitable on the pricing difference captured from the first season. Second profile: the French vintage US premium collector with more than 30 CGC silver/bronze age references, who combines MCC for the VF coverage of its secondary portfolio and GoCollect Premium (€84/year) for census precision on the premium US portfolio. For beginners or light collectors under 30 pieces, MCC as a single tool is more than enough and the combined stack represents an unjustified expense. The free GoCollect account completes at no additional cost for one-off census consultations, which already covers 90% of premium arbitrage use cases without Premium subscription.