Marketplace pricing and feedback
•Download as PPTX, PDF•
11 likes•6,488 views
Lessons on feedback and pricing systems. 5 star feedback systems are flawed.
Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Similar to Marketplace pricing and feedback
Similar to Marketplace pricing and feedback (20)
Recently uploaded
Recently uploaded (20)
Marketplace pricing and feedback
- 1. MARKETPLACE INSIGHTS ON PRICING AND FEEDBACK
- 2. Jackson Square Ventures | Confidential | 2 Hello, My Name is Josh
- 3. Jackson Square Ventures | Confidential | 3 Lessons from Three $B+ GMV Marketplaces acrowdedspace.com Jackson Square Ventures | Confidential | 3
- 4. Jackson Square Ventures | Confidential | 4Jackson Square Ventures | Confidential | 4 On Feedback Observation Solution 5 StarFeedback Systems areBad Grade Inflation Lots ofUnused Signals Meet ArpadElo
- 5. Jackson Square Ventures | Confidential | 5Jackson Square Ventures | Confidential | 5 Virtuous Cycles People spend lots of time here Butit’sineffective Attract OptimizeScreen People don’t spend much time here Butit’svery effective Retain Expel Quality is never static (will increase or decrease) Don’t waste too much time screening; Invest in O.R.E
- 6. Jackson Square Ventures | Confidential | 6Jackson Square Ventures | Confidential | 6 Grade Inflation Observation Solution Don’t bea Jerk Relativevs Absolute Lots ofUnused Signals Meet ArpadElo
- 7. Jackson Square Ventures | Confidential | 7Jackson Square Ventures | Confidential | 7 Feedback and False Signals 0% 20% 40% 60% 80% 100% Yelp oDesk Amazon (Books) Amazon (Toys) 5 4 3 2 1 5 star systems do not provide granularity Ask for private feedback, 3 point scales, and use behavioral data
- 8. Jackson Square Ventures | Confidential | 8Jackson Square Ventures | Confidential | 8 Real World vs 5 Stars 5 star systems do not reflect reality Consider a different approach RealWorld 5 StarDistribution 1 2 3 4 5
- 9. Jackson Square Ventures | Confidential | 9Jackson Square Ventures | Confidential | 9 Feedback for Bad 5 stars can separate the bad sellers But do very little to identify the best 5 stars separatebad. 5 StarDistribution 1 2 3 4 5
- 10. Jackson Square Ventures | Confidential | 10Jackson Square Ventures | Confidential | 10 Feedback for Good Arpad Elo developed a system for scoring 2-person competitions It results in a more real world distribution ELO Ratingsseparate good.
- 11. Jackson Square Ventures | Confidential | 11Jackson Square Ventures | Confidential | 11 Relative vs Absolute Humans are bad at assessing absolute value But we’re extremely good at assessing relative value How good isthis football team? Which one of these teams is better?
- 12. Jackson Square Ventures | Confidential | 12Jackson Square Ventures | Confidential | 12 Finding Signals You won’t be able to separate the good by user feedback alone Need to use behavior and marketplace actions If I choose Jeannie, she should move upand Romany andJamieshould move down.
- 13. Jackson Square Ventures | Confidential | 13Jackson Square Ventures | Confidential | 13 Signals of Progression Look for meaningful signals User feedback is often too subjective Rapidcareeror rateprogressionis a great indicator ofquality!
- 14. Jackson Square Ventures | Confidential | 14Jackson Square Ventures | Confidential | 14 Signals of Repeat Usage Actions speak louder than words Ask your doctor what she would do rather than what she recommends Repeat purchase rateor durationis a strong indicator! “The psychologist Gerd Gigerenzer has a simple heuristic. Never ask the doctor what you should do. Ask him what he would do if he were in your place. You would be surprised at the difference” ― Nassim Nicholas Taleb, Antifragile: Things That Gain from Disorder
- 15. Jackson Square Ventures | Confidential | 15Jackson Square Ventures | Confidential | 15 User Authority Some users feedback is more valuable than others But we treat it all the same
- 16. Jackson Square Ventures | Confidential | 16Jackson Square Ventures | Confidential | 16 Feedback Incentive Users have little incentive to rate accurately Unless you tie it back to the user experience Rate this provider: Do youwant to workwiththis provideragain? Yes No Helps themarketplace Helps thebuyer
- 17. Jackson Square Ventures | Confidential | 17Jackson Square Ventures | Confidential | 17 Feedback Conclusion Takeaways: • Bewarefalsesignals • Look forbehavioral signals • Ditch the 5 stars • Develop your normal curve
- 18. Jackson Square Ventures | Confidential | 18Jackson Square Ventures | Confidential | 18 On Pricing Goal of everymarketplaceshould betoincreaseefficiency inthe world. Fees extracted<< efficiencygained.
- 19. Jackson Square Ventures | Confidential | 19Jackson Square Ventures | Confidential | 19 Pricing Three primary models for marketplaces Each has their own pros and cons Freemium SaaS Ads / Lead Gen Transactional • OftenB2B • Workswell whenyouhave mixedusage (casualvs professional) • Worksif accessisessential to business • OftenB2C • Workswell if big imbalance in supply/ demand • But,isfatallyflawedw/ misalignedincentives • Generally worksthe bestand hasthe mostaligned incentives • Butcan’talwaysdoit becauseofdisintermediation orvery high prices
- 20. Jackson Square Ventures | Confidential | 20Jackson Square Ventures | Confidential | 20 Match Types Upwork,Craigslist Airbnb,Zillow Uber,Rev SupplierPicks Commodity Increasing Fill Ratio AND Rake Buyer Buyer Buyer Supplier Supplier Supplier Buyer Picks Variety Double Commit Complex, custom stuff Marketplaces evolve to increasing fill ratio Supplier picks is easiest way to approach 100%
- 21. Jackson Square Ventures | Confidential | 21Jackson Square Ventures | Confidential | 21 Frequency of Purchase Size of Transaction Plumbing Vertical (HighFrequency) Vertical (HighPrice) Horizontal Locksmith Sporting Goods Furniture Flights Hotels Taxi Meals Homes Cars Clothing Horizontal B2B Horizontal vs. Vertical Verticalize (buyer picks for high price) / (supplier picks for high frequency) Or group categories in horizontal; Or build a B2B platform Supplier Picks High Transaction Rake B2B FreemiumSaaS BuyerPicks Lead Genor Low Rake Double Commit Low Rake
- 22. Jackson Square Ventures | Confidential | 22Jackson Square Ventures | Confidential | 22 50% 0% 20% High Rakes --Managed marketplace, monopoly (iOS), opacity Tiny – Long term, regularwork,users meet in person, paymentoff-platform Med – Quality control, insurance,compliance, scheduling and availability Low – Transparentfees, buyerbeware, high price goods / services Rake Factors High rake not necessarily better, just different More marketplace management – more rake
- 23. Jackson Square Ventures | Confidential | 23Jackson Square Ventures | Confidential | 23 Disintermediation and Dynamic Pricing Time Perceived Value Total PerceivedValue FlatFee Disintermediationhappens here Users seek the best financial outcome Either increase value or decrease fee over time SteppedDiscounts
- 24. Jackson Square Ventures | Confidential | 24Jackson Square Ventures | Confidential | 24 Payment Timing Paying sellers fast is consistently the best thing for seller happiness Pay fast. Pay reliably. Assume some risk and guarantee payment. Pay Fast High NPS. High Loyalty. High Retention.
- 25. Jackson Square Ventures | Confidential | 25Jackson Square Ventures | Confidential | 25 Optics in Pricing People often ask the question, “who do you charge?” While it’s a stupid question, it does matter Buyer Seller Pays $100 Receives $80 Rakes $20 Who paidthe fee?
- 26. Jackson Square Ventures | Confidential | 26Jackson Square Ventures | Confidential | 26 Optics in Pricing These are all equivalent But create different levels of happiness Buyer Seller Pays$100 Receives$80Rake$20 Invoice Services $80 MarketplaceFee $20 Earnings Services $80 Invoice Services $100 Earnings Services $80 MarketplaceFee $20 Invoice Services $90 Access Fee $10 Earnings Services $80 MarketplaceFee $10
- 27. Jackson Square Ventures | Confidential | 27Jackson Square Ventures | Confidential | 27 Pricing Conclusion Takeaways: • Pay people fast • Choose the rightmatchtype andmodel • Consider more dynamic pricing • Optics matter
Editor's Notes
- Option 1
- Grade inflation Causes mismatching distribution curves Causes an inability to find high quality Means we need to use unused signals
- Grade inflation Causes mismatching distribution curves Causes an inability to find high quality Means we need to use unused signals
- Option 2