Lemonade, Inc. · NYSE:LMND
Lemonade — Investor Day (November 2022)
125 slides from Lemonade's investor day presentation, Nov 15, 2022
Slide 1: Investor Day November 15 2022
Investor Day November 15 2022
Slide 2: Safe Harbor
Safe Harbor This presentation contains forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. All statements contained in this presentation that do not relate to matters of historical fact should be considered forward-looking statements, including stat
Slide 3: Today, we hope to demonstrate that LMND has
Today, we hope to demonstrate that LMND has: A structural technological advantage that will manifest as a superior loss-cost and expense-load An inbuilt edge in acquiring & delighting new entrants, whose full value we’ll unlock as we grow with them An advantaged long-term business model, and a clear
Slide 4: Today's agenda
Today’s agenda: Daniel Schreiber Co-CEO & Cofounder Strategy Maya Prosor Chief Business Officer Products Tim Bixby Chief Financial Officer Financial Q&A
Slide 5: Daniel Schreiber
Daniel Schreiber November 15 2022
Slide 6: Alan Turing quote on machines
“It is customary to o!er a grain of comfort, in the form of a statement that some peculiarly human characteristic could never be imitated Alan Turing, 1951 by a machine. I cannot o!er such comfort!”
Slide 7: History of AI timeline
GenerativeiAI Self-Supervised Learning (Foundation Models: GPT3, DALL-E, Stable Diffusion…) Birth of AI Big Data + Machine & Deep Learning Symbolic AI (image & speech recognition, AlphaGo, autonomous driving) (DARPA, Checkers) Big Computers + 2022 2021 2020 2019 2018 2017 2016 2015 2014 2013 2012 20
Slide 8: Shakespearean sonnet about AI and insurance
>> Write a Shakespearean sonnet about AI and insurance GPT3
Slide 9: GPT3 sonnet output
When I do think ofiinsurance, ’tis but To ponder howiAI doth make it better. With nimble speediand keenest eye, The claimsiand coverages do it spy; And inia trice, with little fuss The benefitsiand the best rates thus Are foundiand given with but aiclick; So, fairestcinsurance, I do thee praise For
Slide 10: Impressionist painting prompt
>> Draw an impressionist painting an insurance adjuster denying coverage after a fire DALL·E
Slide 11: DALL·E impressionist painting
An impressionist painting of an insurance adjuster denying coverage after a fire DALL·E
Slide 12: Insurance agent made of salad prompt
>> Render a photo of an insurance agent made out of a salad on a table DALL·E
Slide 13: DALL·E salad insurance agent
An insurance agent made out of a salad on a table DALL·E
Slide 14: 1960s insurance broker photo prompt
>> Create a 1960s photo of a despondent insurance broker in a red dress sitting at a desk full of paperwork, with florescent lighting, taken using a 50mm lens, expired CineStill 800t film, and an oversaturated filter DALL·E
Slide 15: DALL·E 1960s broker photo
Photo of a despondent insurance broker in a red dress in the 1960s DALL·E
Slide 16: First of two pictures from the IPO prospectus
The stunning progress in AI brings us to the first of two pictures drawn by Lemonade for our IPO prospectus
Slide 17: Insurance as a neural network
>> Draw a schematic of insurance as a neural network
Slide 18: Customer Cortex
Customer Cortex insurance as a neural network
Slide 19: Customer Cortex: data and machine learning models
We’ve used hundreds of millions of customer interactions, and 160TB of textured data, to train about 50 machine learning models. Customer Cortex
Slide 20: Customer Cortex: interconnectedness
But the power of our neural network derives less from the sheer tonnage of our data, than from their interconnectedness. Customer Cortex
Slide 21: Slide 21
Slide 22: Pie] press 1-Q419
2000 1000 1000 =2000 Pie] press 1-Q419 ' ‘ 1 1 ' 1 UV1-Qe19 uv2- 92-20 V3- 94-20 UV 4 gaat UV 5-94-21, LTV 6 - Q2-22
Slide 23: Exponential Progress
Exponential Progress Flywheel spinning fast, and getting faster 3B 110m Features Predictions Called Made 250m 8.5m 2020 2022 2020 2022 2020 2022 2020 2022
Slide 24: The upshot: More for Less
Theiupshot: More for Less
Slide 25: More
More
Slide 26: More Delight
More Delight “The cheerless customer service representative at Liberty Mutual had about half the personality of Lemonade’s chatbot”
Slide 27: More Delight: Net Promoter Score
More Delight Net Promotor Score familiar from Apple & Tesla
Slide 28: More Delight: Awards
More Delight Award winning insurance America’s Best Insurance 2023 “Using AI to approve “Lemonade is No. 1 in our coverage in 90 seconds Best Homeowners Insurance and pay claims in as little Companies of 2022” as three minutes”
Slide 29: For Less
For Less
Slide 30: Expense Ratio and Loss Ratio
G&A Marketing Support Fraud LAE Claims Expense Ratio Loss Ratio
Slide 31: Expense Ratio: Cooper
Expense Ratio Cooper monitors servers, handles tasks, instantiates virtual environments, pushes >12,000 software builds into production yearly, saving ~10,000 (human) dev hours.
Slide 32: Expense Ratio: AI Maya
Expense Ratio AI Maya and our APIs sell 98% of Lemonade’s policies, morphing between millions of onboarding flows. As a point of reference: ~95% of incumbents’ homeowners policies are sold by humans, taking ~15% of premiums, in perpetuity, for their efforts.
Slide 33: Expense Ratio: marketing models
Expense Ratio Machine-learning models generate LTV predictions for every customer, every campaign, every product and every geo. These models govern 86% of our marketing spend.
Slide 34: Expense Ratio: CX.ai
Expense Ratio CX.ai handles about ⅓ of all customer requests, including scheduling items, adding spouses, moving home and much more, with a CSAT score of 4.5
Slide 35: Loss Ratio: AI Jim
Loss Ratio AI Jim takes first notice of loss (FNOL) for 98% of claims, and handles 45% end- to-end, with a level of accuracy that surpasses humans. Excluding Metromile
Slide 36: Loss Ratio: fraud algorithms
Loss Ratio 1,000,000,000% 9.2m 9,189,851 7,823,060 750,000,000% 7,157,463 6,822,014 6,477,804 6,181,291 5,938,847 Fraud algorithms flagged over >$100m in 5,660,259 5,410,634 5,197,013 5,022,546 500,000,000% 4,873,521 fraudulent claims, and detected doctored 4,554,145 documents in $12m worth of claim
Slide 37: Loss Ratio: Watchtower
Loss Ratio Watchtower tracks wildfires and weather, automatically halting marketing, and instituting a delayed-start-date for affected areas.
Slide 38: Loss Ratio: Telematics
Loss Ratio Telematics, continuously used by >90% of customers, lowers LAE, lowers adverse selection, and enables pricing at a level of precision unavailable to legacy systems.
Slide 39: Loss Ratio: computer vision and NLP
Loss Ratio Computer vision and Natural Language Processing (NLP) detect risky homes for underwriting review, with a hit rate of ~80%.
Slide 40: Loss Ratio: machine learning pricing
Loss Ratio Machine Learning and Deep Learning match rate and risk with unrivaled precision, forming the basis of our newest rate filings.
Slide 41: But everyone's doing AI. Right?
“But everyone’s doing AI. Right?”
Slide 42: John Peters, formerly Liberty Mutual
John Peters, formerly EVP & Chief Underwriting O!cer Liberty Mutual
Slide 43: So why aren't we seeing it in the numbers?
“So why aren’t we seeing it in the numbers?”
Slide 44: We Are Seeing It In The Numbers
We Are Seeing It In The Numbers Predicted Cohort Lifetime Loss Ratio 87% 85.8% 84.3% 79.1% 77.75% 76.7% 76.3% 68.4% 68.5% 60.6% 59.25% Q1 21 Q2 21 Q3 21 Q4 21 Q1 22 Q2 22 Q3 22
Slide 45: Front Loaded: Near-term Burdens Become Long-term Assets
Front Loaded Near-term Burdens Become Long-term Assets o d u c t s n g P r Selli a n t a g e A d v Data s s o d u c t s e d B u s i n e di ng P r Season ion-Sales Buil d va n t a g e o - Co m m i s s t a D i s a a l ty N Da i n e s s P e n s t B u s i o n Co New r A c q u i s i t t o m e Cus
Slide 46: Front Loaded: Near-term Burdens Become Long-term Assets
Front Loaded Near-term Burdens Become Long-term Assets Buildin g Produ Data D cts isadvan New B tage usiness Selling Custom er Acqu Penalty Produc isition Data A ts Cost Season d vantage No-Co ed Busin mmissi ess on-Sale s
Slide 47: Built for Scale
Built for Scale A Growing Denominator Shrinks the Expense Ratio
Slide 48: Built for Scale: Adjusted Expense Ratio
Built for Scale Since Q3 21, Adjusted Expense Ratio Dropped 24% As IFP Grew 76% $609m 110% 61,000% 103% 100% 53,250% 96% 95% $458m 90% 45,500% $419m $380m 81% 80% 37,750% 80% $347m Q3 21 Q4 21 Q1 22 Q2 22 Q3 22 Adjusted Expense Ratio is calculated as Gross Earned Premium divided by all operational e
Slide 49: Second picture in Lemonade's IPO prospectus
Which brings us to the second picture in Lemonade’s IPO prospectus
Slide 50: Grow with Customers
>> Draw Grow with Customers
Slide 51: Renters, Condo, Homeowners
Renters Condo Homeowners
Slide 52: The Gateway Policy
n=1,500, 9/22 The Gateway Policy (+2.1%/-1.9%) First time buyers of Renters Insurance across USA in past 12 months 24.1% 14.8% 14.1% 10.7% 9.5% 26.9% State Farm Lemonade GEICO Progressive Liberty All Others
Slide 53: The Gateway Policy: under 35
n=340, 9/22 The Gateway Policy (+4.1%/-3.5%) First time buyers of Renters Insurance in past 12 months under 35yr olds 22.1% 21.8% 16.5% 11.5% 11.5% 16.8% Lemonade State Farm GEICO Progressive Liberty All Others
Slide 54: At IPO
At IPO Renters Homeowners 19% 20% 21% 18% 18% 17% 15% 14% 11% 12% 8% 9% 9% 10% 6% 7% Percent of Condo customers that 6% 3% 4% 2% 2% 3% graduated from Renters (at IPO) 2% Q1-17 Q2-17 Q3-17 Q4-17 Q1-18 Q2-18 Q3-18 Q4-18 Q1-19 Q2-19 Q3-19 Q4-19 Q1-20 Q2-20 Q3-20 Q4-20 Q1-21 Q2-21 Q3-21 Q4-21 Q1-22 Q2-2
Slide 55: At IPO and Since IPO: cross-sold
At IPO Since IPO Renters Homeowners Pet Term Life Car Cross sold:j ~⅓ ~⅓ ~⅓ ~⅓ Premium jump:j 6x 4x 5x 8x
Slide 56: It's Early Days...
It’s Early Days...
Slide 57: ~60% of incumbents' customers are multi-line
~60% of incumbents’ customers are multi-line customers
Slide 58: <4% of Lemonade's customers are multi-line
<4% of Lemonade’s customers are multi-line customers
Slide 59: We're on our way: multi-line customers
We’re on our way Multi-Line Customers Illinois: 6% 5.9% 4.5% USA: 3.7% 3% 1.5% 0% May 20 July 20 Sep 20 Nov 20 Jan 21 Mar 21 May 21 Jul 21 Sep 21 Nov 21 Jan 22 Mar 22 May 22 Jul 22 Sep 22
Slide 60: In Conclusion...
In Conclusion...
Slide 61: Conclusion: neural network insurance and growing with customers
Build insurance as a neural network Capture tomorrow’s most profitable = more love + less costs customers today, and grow with them
Slide 62: Play The Movie Forward
Play The Movie Forward >150x >100x >50x
Slide 63: Maya Prosor
Semonade Maya Prosor November 15 2022
Slide 64: Portfolio of products for risk management and healthy growth
Portfolio of products that supports Risk Management and Healthy Growth
Slide 65: The Perfect Onramp: Lemonade Renters
The Perfect Onramp Lemonade Renters
Slide 66: Renters Q3 Results
Q3 Results $223M 1.4M X3.15 IFP Customers LTV/CAC 65% 51% 50% Q3 Loss Ratio Predicted LR Conversion Q3 cohort rate
Slide 67: Cross Sales coming from just Renters
Cross Sales coming from just Renters IFP ($,M)
Slide 68: 5X Renters with 10X Heart: Lemonade Pet
5X Renters with 10X Heart Lemonade Pet
Slide 69: The US Pet Insurance Market
The US Market $2.6B at ~25% CAGR Households with Pets with Pet insurance* >80M 2.5% *2021 data 70
Slide 70: And we are growing way faster
And we are growing way faster From standstill to $110M in 2 years Coming Spring 2023
Slide 71: Pet Q3 Results
Q3 Results $110M X3.8 15% IFP LTV/CAC Quote Rate from other products ~60% 47% 450K Predicted LR Claims Number of Q3 cohort handled by AI claims As of Q3-22
Slide 72: High Frequency, Low Severity
High Frequency, Low Severity Total number of claims handled at Lemonade >450K pet claims handled to date All other product claims
Slide 73: Our focus on efficiency is paying off
Our focus on efficiency is paying off Loss Adjustment Expenses (LAE) as percentage of premiums
Slide 74: Rounds out the Family: Lemonade Life
Rounds out the Family Lemonade Life
Slide 75: Life bundled churn rate vs. Renters
35% 1/2 Bundled with other The churn rate vs. products single line Renters
Slide 76: Made For Bundling: Lemonade Home
Made For Bundling Lemonade Home
Slide 77: Home Q3 Results
Q3 Results >$120B $149M X1.6 TAM IFP LTV/CAC 119% 81% 40% Q3 Loss Ratio Predicted LR Auto Handled October UW Reviews
Slide 78: On the Right Path: 81% Predicted Loss Ratio
On the Right Path 81% Predicted Loss Ratio for new customer cohorts Predicted loss ratio by month
Slide 79: Time Lag: rate changes this year
Time Lag Rate changes we’ve taken this year 118% 85% 4.5% Filed Approved Earned rate as % of rate as % of rate as % of total book total book total book LTV
Slide 80: Geo Profitability Management
Geo Profitability Management Reducing California New Business Share
Slide 81: Getting the Right Risk
Getting the Right Risk
Slide 82: Slide 82
Slide 83: Real-Estate Property Description
TRIGGERS Real-Estate Property Description Property Overview Great starter pool home with loads of potential. On a corner lot featuring 3 bedrooms 2 bathrooms with copper plumbing, solar panels, new water heater and updated electrical. The home needs TLC inside but perfect for someone that doesn’t mi
Slide 84: Inspection Reports Auto-Review
TRIGGERS Inspection Reports Auto-Review
Slide 85: Aerial Imagery Capabilities
Aerial Imagery Capabilities Roofs segmentation, mobile homes and more…
Slide 86: Pricing: Wash, Rinse, Repeat
Pricing Wash, Rinse, Repeat File best assessment of Rates Forms & Underwriting Map learnings onto Rates, Sell critical mass of such policies Forms & Underwriting (to derive conversion, churn…) Pay critical mass of claims (to derive risk predictions, pricing)
Slide 87: Sophistication in Pricing and UW: Territory Factors
Sophistication in Pricing and UW Territory Factors
Slide 88: Sophistication in Pricing and UW: Address Level CAT
Sophistication in Pricing and UW Address Level CAT
Slide 89: Sophistication in Pricing and UW: ML Loss Model
Sophistication in Pricing and UW ML Loss Model
Slide 90: ML Model for Pricing in TX
ML Model for Pricing in TX ! Predicted loss ratios for new customer cohorts Conversion Rates Slightly Increased
Slide 91: On the Right Path: 81% Predicted Loss Ratio
On the Right Path 81% Predicted Loss Ratio for new customer cohorts
Slide 92: The Highway: Lemonade Car
The Highway Lemonade Car
Slide 93: Pent Up Demand and a Huge Market
Pent Up Demand and a Huge Market “Cheaper insurance. I am loyal customer and eagerly waiting for “Smooth experience “Lemonade stole the show ininsurance “Car home in Texas so just like the “My primary hope is that New York lemonade car insurance but renter insurance, insurance, compared to any Iothe
Slide 94: Crunching Time: >90% Telematics
Crunching Time aka bridging the data gap >90% Telematics
Slide 95: Power of Telematics
Power of Telematics Without Telematics 20 Year Old Male Medium Credit Score No Prior Accidents Top 10% highest risk
Slide 96: Power of Telematics: low-risk driver example
Power of Telematics Without With Telematics 20 Year Old Male Medium Credit Score No Prior Accidents 90% Day Driving 95% Within Speed Limit 0.3% Hard Brakes 5% High Traffic Driving 90% Main Roads 500 miles / month 0.2% Sharp Turns In reality a low risk driver, And we can price him as such
Slide 97: Touchless First Notice Of Loss
Touchless First Notice Of Loss Digitally capturing initial claim info, allowing advocates to handle claims faster
Slide 98: Affordable and Amazing
w& & & & & SZ United states Affordable and Amazing | just wanted to pop here and give a 5 star review because this morning | was taking my nephew to school and a car in front of me brake checked because a car pulled out in front of it so | also had to slam my breaks, and no more than asking if | was
Slide 99: Telematics Powered AI: Hidden drivers detection
Telematics Powered AI Hidden drivers detection
Slide 100: Telematics Powered AI: Rideshare detection
Telematics Powered AI Rideshare detection
Slide 101: Pricing: Wash, Rinse, Repeat
Pricing Wash, Rinse, Repeat File best assessment of Rates Forms & Underwriting Map learnings onto Rates, Sell critical mass of such policies Forms & Underwriting (to derive conversion, churn…) Pay critical mass of claims (to derive risk predictions, pricing)
Slide 102: Ratemaking Machine
Ratemaking Machine 2 Months of Coding Zero Coding
Slide 103: Our Strategy
Our Strategy Delight © Consumers \ Machine Grow Learns Fast + \ Predictive / Data
Slide 104: Tim Bixby
Semonade Tim Bixby November 15 2022
Slide 105: Let's do the numbers
Let’s do the numbers…….
Slide 106: 5 year IFP CAGR, Gross Loss Ratio, multi-line rate
20% 70% ~25% 5 year IFP Gross Loss Multi-line CAGR Ratio (2027) customer rate by 2027
Slide 107: Cash/Investments
Cash/Investments ~$1B Unrestricted Restricted 2022E 2023 2024 2025 2026 2027 Capex & Cash Invest. Income, net EBITDA *Net working capital assumed to be zero
Slide 108: 5 year IFP CAGR, Gross Loss Ratio, multi-line rate
20% 70% ~25% 5 year IFP Gross Loss Multi-line CAGR Ratio (2027) customer rate by 2027
Slide 109: Base - Profitable within 5 years
Base - Profitable within 5 years IFP 20% CAGR 5% 25% ~$100M Minimum Unrestricted Cash Gross 70% Loss Ratio Mid '26 60% 80% Turning profitable Multi-line 25% customer 10% 40%
Slide 110: Better - Earlier Profitability
Better - Earlier Profitability IFP 20% CAGR 5% 25% ~$175M Minimum Unrestricted Cash Gross 65% Loss Ratio 2025 60% 80% Turning profitable Multi-line 35% customer 10% 40%
Slide 111: Worse - Later Profitability
Worse - Later Profitability IFP 20% CAGR 5% 25% Gross 78% ~$125M Loss Ratio Unrestricted 60% 80% Cash Deficit In 2027 Multi-line 20% customer 10% 40%
Slide 112: Adjusted Growth - Profitability in 5 years
Adjusted Growth - Profitability in 5 years IFP 12% CAGR 5% 25% ~$100M Minimum Unrestricted Cash Gross 78% Loss Ratio 2026 60% 80% Turning profitable Multi-line 20% customer 10% 40%
Slide 113: Metrics
Metrics
Slide 114: Consistent, Steady Topline Growth
Consistent, Steady Topline Growth IFP development Ukraine Invasion Bear Market Inflation! IPO COVID Q1-18 Q3-22
Slide 115: 10 consecutive quarters meeting or beating guidance
10/10 For 10 consecutive quarters, we have met or beat our financial guidance
Slide 116: Loss Ratio by Product
Loss Ratio by Product Home Pet Renters Q1-22 Q4E-22
Slide 117: Larger Investments Behind Us
Larger Investments Behind Us EBITDA/Gross Earned Premium, 5 years Q1-18 Q4E-22
Slide 118: Continued Improvement
Continued Improvement EBITDA/Gross Earned Premium, 3 years MILE Pet Car IPO Q1-20 Q4E-22
Slide 119: Improving Efficiency & Automation
Improving Efficiency & Automation IFP/headcount ($ 000’s) 2 5% + Q4 -’21 Q4E -’22
Slide 120: Consistent Premium Growth
Consistent Premium Growth Annual premium per customer Q3-18 Q3-19 Q3-20 Q3-21 Q3-22
Slide 121: Vast Potential to Grow With Customers
Vast Potential to Grow With Customers Annual premium per customer High Net Worth Top LMND customer LMND LMND 2022 2018 US Average
Slide 122: Surplus Strategy
Surplus Strategy
Slide 123: Evolving Surplus Strategy
Evolving Surplus Strategy ~3:1 If 100% retained ~5:1 ~6:1 If 75% ceded to Overall captive ~8:1 If 75% ceded to reinsurance
Slide 124: What if...?
What if…..? Today Future #2 Future #1 1.8m 7.2m users 3.6m users customers 100% of avg 50% of avg PPC $343 PPC PPC IFP: ~$6B IFP: ~$0.6B IFP: ~$23B Today Future #1 Future #2
Slide 125: Investor Day November 15 2022
Investor Day November 15 2022