2024

Tech
Trends
Make smarter
tech decisions.
Our platform will help you discover,
understand, and manage your response to
emerging tech. Make confident tech decisions
that drive growth, improve operations, and
beat the competition with CB Insights.

▮▮▮▮IIII
Try
Sign
our
upplatform
for a free
– trial
free!

2
Your always
on AI-Driven
Research Analyst
Do less searching, get more answers

▮▮▮▮IIII
Join the waitlist for
early access

3
Tech trends
AI Healthcare & life sciences
GPU shortage forces companies to be smarter 05 The great AI drug race heats up 71
Multimodal AI will rise on corporates’ wish lists 12 Digital therapeutics (DTx) & wellness consolidates 78
Synthetic data bonanza 17 The brain becomes a fierce tech battleground 84

Enterprise Retail & consumer

DIY software development upends engineering orgs 22 AI sales agents flood the e-commerce landscape 89
Global employment & payroll market shakes out 28 Retailers tackle shrink with AI loss prevention 96
Quantum computing advances hint at faster commercialization 33 AI sends a shockwave through gaming 102

Cybersecurity Industrials
Cyber chaos drives security consolidation 38 Humanoid robots come for manufacturing 109
AI vs. AI dogfights redefine data security 46
Venture
Financial services & insurance Corporate venture refocuses on strategic fit 115
Banks get AI FOMO 52 Unicorns need a new playbook to survive 120
Blockchain’s uphill finserv battle 59
Extreme weather is an opportunity for insurtech 66

4
AI

GPU shortage
forces companies
to be smarter

5
AI

Companies are desperate for GPUs

H100s sell at a massive mark-up

$97,835
“The GPUs at this point
are considerably harder
to get than drugs.”
$45,000
Elon Musk at the WSJ
$30,000 CEO Council Summit
$3,320
Cost of making h100 Average sale price of H100 cost on ebay H100 cost on CDW,
Average
H100sale Listed on eBay, Listed on CDW,
Jan'24

Estimated cost of price as of Jan ’24 as of Jan ’24

making an H100

Source: Estimates from Raymond James on cost of making an H100. GPUs = graphics processing units, 6
which are used to run intensive AI applications.
AI

Startups link up with big tech to access their AI chips

and compute power
Generative AI companies with two or more big tech investors (as of 1/30/2024)
Big tech investors
Indicates where
big tech invested
Company

Source: CB Insights financing data. Includes investments from M12, GV (Google Ventures), and NVentures. *AWS. 7
AI

While also looking outside of big tech for cheaper cloud compute
Nvidia is giving priority access to its GPUs to smaller specialty cloud
providers CoreWeave and Lambda Labs, which are offering cloud GPU access at
lower costs to enterprises

Cloud provider GPU model Price per hour

A100 40 GB
p4d.24xlarge $4.10

A100 (80 GB)

NC24ads A100 v4 $3.67

A100 40GB $3.48

A100 40GB NVLINK $2.06

NVIDIA A100 PCIe 40GB $1.29

A100 40GB NVLINK $1.29

$0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 $3.50 $4.00 $4.50

Source: Cloud provider websites. 8

AI

But as demand continues to outstrip supply, companies aim

to optimize their AI models to reduce compute time

“For the deployment side of things, we found that the performance of our training process
was quite slow, especially when it gets into these large language models and when you
train from scratch. MosaicML offers what's called programmatic optimization, which is
not so much on the hardware side of things, but rather on the algorithmic side. Can you
find ways of optimizing the time it takes to get to a certain performance bar?... We did
experiments where we used MosaicML programmatic optimization versus not, and we did
see a 12% compute hour reduction for some of the models.”

MosaicML customer
Senior Manager, Data Science at $1B+ valuation technology company

Source: CB Insights buyer transcript – MosaicML. 9

AI

Startups building software to help AI models run efficiently on

available hardware gain traction

Company Headquarters Total funding Latest funding round Application

Seed VC Identifies bottlenecks & underutilized

Canada $31M
(10/25/2023) hardware to optimize model deployment

Seed VC Enables AI models to run on cheaper CPU

US $6M
(8/13/2021) hardware

Series C Optimizes GPU resources between the

Israel $118M
(3/15/2022) cloud and on-prem

Acquired by Nvidia Shrinks models so they can run on edge

US $10M
(2/1/23) devices

Series B
Israel $55M Offers hardware-aware AI optimization
(7/13/2022)

Source: CB Insights company profiles – CentML, ThirdAI, Run:AI, OmniML, Deci. 10

AI

GPU shortage also opens the door for experimenting with novel
processor approaches
Approach Company Headquarters Latest funding round
Uses light instead of electrons to transfer data,
Photonic United States Series C (12/19/2023)
with the aim of increasing speed and bandwidth
while using less energy
United States Series B (4/13/2023)

United States Series A (3/3/2022)

Processes information using analog signals Acquired by Snap

Neuromorphic passed between artificial “neurons” and
France
(10/1/2023)
“synapses” to reduce energy consumption
Israel Unattributed (12/22/2022)

Netherlands Accelerator (1/1/2021)

In-memory Brings processors and memory together onto a

single chip to increase efficiency of data transfers United States Series A (8/1/2023)
computing
and reduce energy consumption
Canada Series B (7/20/2021)

China Seed (4/18/2022)

Highly integrated chips are stacked vertically to

3D chips increase processing power and save energy
United Kingdom Series E (12/29/2020)

Source: CB Insights — Company profiles. 11

AI

Multimodal AI will
rise on corporates’
wish lists

12
AI

Multimodal AI is in its infancy, but models are quickly evolving

Google introduces first commercial model taking diverse inputs in December 2023

2021 & 2022 2023

1 2

Image-text models* LLMs understand Models tackle Natively

proliferate images more sensory multimodal
data models emerge

2021 2022
• Salesforce BLIP 2 • Google PALM-E • Google Gemini
• OpenAI CLIP • Google LiT
• Google ALIGN • Google COCA • Microsoft KOSMOS • Meta ImageBind
• Salesforce ALBEF • DeepMind FLAMINGO • OpenAI GPT-4
• Salesforce BLIP • LLaVA
• Adept FUYU-8B

Source: Published research; press releases. *E.g., image generation based on text, image description generation. 13
Note: Multimodal AI refers to models that can understand and respond in multiple data modalities (text, video, image, etc.).
AI

More LLM developers will follow in Google’s footsteps to remain

competitive
Company Total funding Mode Multimodal features

GPT-4V Limited modalities

$14B Text, image à text commercially available

Claude-2 In the roadmap or not

$3.4B Text only announced

Inflection-2
$1.5B Text à text, image

Magma
$642M Image à text

$469M Text only

$317M Text only

Grok
$135M Text only

Source: CB Insights Market Report — Large language model (LLM) developers. 14

AI

Models that can accept diverse inputs open up new opportunities

for AI in industries that are highly multimodal
Google will bring Gemini-based models into its MedLM family of foundation
models fine-tuned for healthcare use cases in coming months

“Medicine is an inherently multimodal discipline. When providing care, clinicians routinely

interpret data from a wide range of modalities including medical images, clinical notes, lab tests,
electronic health records, genomics, and more. Over the last decade or so, AI systems have
achieved expert-level performance on specific tasks within specific modalities…But how do we
bring these capabilities together to build medical AI systems that can leverage information
from all these sources?”

Greg Corrado, Head of Health AI, Google Research

Yossi Matias, VP, Engineering and Research, Google Research

Source: Google. 15
AI

Watch for commercial applications across healthcare, automotive,

retail, and manufacturing
Sector Company Multimodal AI application

ArteraAI’s prostate test analyzes both biopsy images and patients’ clinical
data to predict patient outcomes
Healthcare diagnostics
Raised $90M in Q1’23 from investors including Coatue and J&J; raised $20M in Q1’24

Ghost Autonomy explores using multimodal LLMs in autonomous driving

Autonomous vehicles
Raised $5M from the OpenAI Startup Fund in Q4’23

Retail Meta is bringing multimodal AI to consumers with its Ray-Ban smart glasses

Google’s multimodal model PaLM-E targets robotics, combining language

Robotics with robotic sensory data

Source: CB Insights company profiles – Artera, Ghost Autonomy; Google, Meta reports. 16
AI

Synthetic data bonanza

17
AI

We are running out of high-quality data to train AI models

Researchers estimate that, by

2026, we will exhaust high-
quality text data* for training
LLMs — a trend that can slow down
AI progress.

Source: Villalobos et al., “Will we run out of data? An analysis of the limits of scaling datasets in Machine Learning.” 18
*Data from sources such as news articles, Wikipedia, filtered web content.
AI

Scraping proprietary sources is getting harder

2023

January March July December

1 2 3 4

Getty Images X imposes a Reddit starts NY Times sues

sues Stability AI $42K bulk access fee charging $0.24/1K OpenAI and Microsoft
for 50M tweets API calls for copyright
infringement

“Scraping proprietary data to train generative AI models for commercial

application is now universally recognized as an unsustainable business tactic.
It's fraught with legal, financial, and reputational challenges.“
Shutterstock CEO Paul Hennessy, Q4’23 earnings call

Source: CB Insights — Earnings Transcripts Analytics (date reflects quarter call occurred); media reports. 19
AI

Cost and scarcity drive model developers to experiment with synthetic data
Company Synthetic data initiatives
Raised $3.5M in seed funding in Q3’23 from Spark Capital, Village Global, and Replit CEO Amjad Masad
to train custom LLMs based on a synthetic data platform
Released glaive-coder-7b, a coding model trained with 140K+ synthetic programming problems and
solutions
Trained LLMs (OpenHermes-2, Nous Hermes) on 900K to 1M data entries generated by GPT-4
Raised $5.2M in seed funding in Q4’21 from DistributedG, OSS Capital, and others

Trained small language models phi-1 and 2 on GPT-generated synthetic textbooks

Released Stable Beluga 1 and 2 language models fine-tuned on synthetic data

Released AlphaGeometry model in Q1’24 capable of solving Olympiad-level geometry problems — the
model was trained on a synthetic data pool of 100M unique examples

CEO Sam Altman was quoted as saying all data will soon be synthetic

Source: CB Insights company profiles – Glaive, Nous, OpenAI, Stability AI, DeepMind; company reports. 20
AI

Synthetic data providers will next target niche applications

Stage: Series A Stage: Series E Stage: Series B

“For banks, where fraud

“Synthetic data applied to Launched Scale Synthetic in
typically constitutes less than
healthcare is not common, so it February 2022 with early
100th of a percent of total
was important to fit our needs customers including the US
data, Gretel’s software can
and protect sensitive data. Department of Defense and
generate thousands of edge
Aindo's technology for privacy autonomous trucking
case scenarios on fraud and
enhancement was crucial...” company Kodiak Robotics.
train [AI] models with it.” – FT
– Customer transcript

Source: CB Insights company profiles – Gretel, Scale, Aindo; Financial Times. 21

Enterprise

DIY software
development upends
engineering orgs

22
Enterprise

Skilled software developers are in high demand

Percent change in employment, projected 2022-2032

Software
Software
developers 26%
developers

Total, all
Total, all
occupations
occupations
3%

0% 5% 10% 15% 20% 25% 30%

Source: US Bureau of Labor Statistics. 23

Enterprise

A massive number of software development tools have emerged

to assist developers
CB Insights tracks over 70 tech markets related to enterprise
software development

Source: CB Insights Market Reports. 24

Enterprise

Skills gap has given rise to “no-code” and “low-code” platforms, but more
flexible AI coding tools could surpass this market in the long term
Low-code player OutSystems Other low-code platforms conduct layoffs
faces valuation compression

$9.5B

$4.3B

Series E Private Equity - II

Feb 2021 Oct 2022

Source: CB Insights company profiles – OutSystems, Retool, Appsmith. 25

Enterprise

AI copilots supercharge developer productivity

A flurry of code generation tools has emerged
Company/product Status

Released by Microsoft (10/29/2021)

Released by OpenAI (11/30/2022)

Released by Meta (8/24/2023)

StableCode by Released (8/8/2023)

Released by Hugging Face & ServiceNow (5/4/2023)

Released (4/13/2023)

AI Released widely (10/9/2023)

Private (founded 2017)

Private (founded 2020)

Source: CB Insights company profiles; company reports. 26

Enterprise

Enterprises will be on guard against confidential code leakage,

security vulnerabilities in AI-written code
Percentage of coders submitting secure answers to coding questions

Using AI Not using AI

50%
43%
36%
29%

21% 21% 21%

14%
3% 12%

Company A &
Encryption Company
Signing B
a message Company C
Sandboxed Company
SQL D StringsE
Company
decryption directory

Source: Perry et al., “Do Users Write More Insecure Code with AI Assistants?” 27
Enterprise

Global employment
& payroll market
shakes out

28
 Enterprise

Funding for global employment startups went to zero in 2023

10 leading startups* raised over $2B in equity funding from 2021 to 2022
$1,600M

$1,400M
$1,494M

$1,285M
$1,200M

$1,000M

$800M

$600M

$400M

$200M
$306M
$45M $0M
$0M

2019 2020 2021 2022 2023

Source: CB Insights — financing data. *Atlas, Deel, Globalization Partners, Multiplier, Omnipresent, Oyster, 29
Papaya Global, Remofirst, Remote, and Velocity Global.
Enterprise

Demand for remote talent jumpstarted the market, but slowdown

in tech hiring will cramp players

“Given the current context, though, I don't know what their traction looks like,
but in the startup world, hiring is slowing down in the US and Europe. And
my guess is that it's probably impacting them at some point. On our side, we
have frozen hiring in most of our countries. So we have not been opening
new Remote.com contracts since probably April 2022. So it will be important
for them to remain competitive on pricing.”

Remote customer
Chief of Staff, $1B+ valuation online marketplace

Source: CB Insights buyer transcript – Remote. 30

Enterprise

Meanwhile competition continues to heat up as HR startups & incumbents

expand their international offerings
Company Valuation Description Development

Expanded into global employment & payroll (October

$11.3B Employee management
2022); business expected to generate $40M in ARR in
(March 2023) platform
2023

$97B (market cap as HR management ADP GlobalView payroll solution expanded into 5 new
of January 2024) software markets (July 2023)

$10B Payroll & HR solutions Launched Gusto Global powered by Remote (June
(August 2021) platform 2023)

$10.6B (market cap HR management Ceridian Dayforce available across most of the Middle
as of January 2024) software East and Africa (August 2023)

HR & workforce
N/A Acquired global payroll provider Immedis (June 2023)
management solutions

Source: CB Insights company profiles – Rippling, Gusto, ADP, Ceridian, UKG; media reports. 31
Enterprise

With capital tight and larger players looming, players will have to look for
scale and efficiencies
We predict that consolidation is approaching the workforce globalization space

Source: CB Insights Market Reports – Workforce globalization, Global payroll. 32

Enterprise

Quantum computing
advances hint at faster
commercialization

33
 Enterprise

Quantum computing funding surged to a record high in 2023,

bucking the broader venture downturn for 2 years in a row
Disclosed equity funding to quantum computing companies

$1.4B
Funding 80
$1.3B
70
$1.2B
Deals
$1.0B
54 64
60

47
43 50
$0.8B

40
$0.6B
31
30

$0.4B
20

$0.2B
10

$0.4B $0.4B $0.9B $1.0B

$0.0B 0

2019 2020 2021 2022 2023

Source: CB Insights Expert Collection — Quantum Tech. 34

Enterprise

Neutral atom quantum computers are having a breakout moment,

hinting at accelerated timelines for commercialization
Atom Computing passed 1,000 qubits and QuEra announced an error correction milestone

Source: CB Insights research — 102 companies racing to commercialize quantum computing. 35

Enterprise

More enterprises are running pilots with quantum startups

“We worked together on “We received a solution for “It had experts in finance and
developing and deploying tests benchmarking a specific was interested in quantum-
for the algorithm they coded [optimization] problem class, inspired techniques, which we
for us. We are currently waiting which is relevant to the see great potential in for the
for testing and scaling up the problems we have in our
short term […] We are still
proof of concept to bring the company for benchmarking
quantum algorithms. This analyzing the impact of the
algorithm from the test result of this work, but we
gave us a working code base
environment to the exercise expect that it will decrease the
that we could use.”
environment.” time and cost associated with
Entropica Labs customer solving certain problems.”
Qilimanjaro customer Analyst, Fortune Global 500 company
Manager, Global transport company Multiverse Computing customer
Senior Analyst, Fortune Global 500 company

Source: CB Insights buyer transcripts — Qilimanjaro, Entropica Labs, Multiverse Computing. 36

Enterprise

Interest in the intersection between quantum computing and AI is heating

up as quantum advances tease more powerful AI models down the road
Number of media articles mentioning both quantum computing and AI

Source: CB Insights — News Mentions Analytics. 37

Cybersecurity

Cyber chaos drives

security consolidation

38
Cybersecurity

Cyber crimes have massive financial consequences

The largest threats drive losses of $100M+, while the average cost of a data
breach has grown 15% in the last 3 years

MGM Resorts
ransomware attack $100.0M
(2023)

Caesars Entertainment
social engineering $15.0M
attack (2023)

Average data breach $4.5M

23andMe data breach

(2023) $1.5M

$0.0M $20. 0M $40. 0M $60. 0M $80. 0M $100.0M

Source: Media reports. Cost of average data breach comes from a 2023 IBM study. 23andMe reported that the 2023 data breach would likely cost 39
between $1M and $2M.
Cybersecurity

And bad actors are getting smarter — and busier — by the day

"From an adversarial perspective,

we expect that the adoption of
LLMs will lower the barrier of
Attack attempts more entry for malicious actors to
create sophisticated cyber
than doubled in 2023 attacks. Generative AI is expected
to make it even easier for less

104%
advanced attackers to crack
nation-state-caliber campaigns.”

CEO George Kurtz, Q2'23 earnings call

Source: CB Insights — Earnings Transcripts Analytics (date reflects quarter call occurred); Armis. 40
 Cybersecurity

To keep up with an evolving attack surface, cybersecurity leaders

snap up competitors
Cybersecurity sector sees steady consolidation, with a surge in M&A deals in 2023
350

300
M&A deals
260
250
269
248
200
205
150
154
100

50

2019 2020 2021 2022 2023

Source: CB Insights Expert Collection — Cybersecurity. Note: Excludes financial M&A such as private equity transactions. 41
Cybersecurity

Cybersecurity leaders have been on an M&A tear

Top strategic acquirers by # of cybersecurity acquisitions since 2022

Acquirer Cybersecurity acquisitions Select acquisition targets

1 7

2 4

3 3

3 3

3 3

3 3

3 3

3 3

Source: CB Insights Expert Collection — Cybersecurity. Note: Includes pending acquisitions, such as Cisco’s acquisition of Splunk. Data as of 42
1/30/2024.
Cybersecurity

Who’s next to get acquired?

Strong early-stage cyber startups with incumbent backing could get scooped up

Mosaic score*
Company (out of 1,000) Latest funding round Application Select cybersecurity investors

$33M Series A Identity & access

780
(December 2023) management

$23M Series A
760 Cloud detection & response
(September 2023)

$13M seed
710 Cyber risk visibility
(August 2023)

Undisclosed seed Identity & access

710
(October 2023) management

Source: CB Insights company profiles — Oleria, Gem Security, Silk Security, Aembit. *Mosaic refers to CB Insights’ proprietary model that scores 43
tech company health and potential. Data as of 2/6/2024.
Cybersecurity

But do customers like the all-in-one approach?

“…we are a little bit concerned “…even though Palo has been “We considered Mimecast's
about the acquisition of Splunk on the acquisition spree, the CyberGraph, but the
by Cisco and what that will mean integration piece natively in integration with the core
for the future of the product. the back-end has not fully Mimecast platform…was weak
We're uncertain if Cisco will completed yet. It has gotten
due to it being an acquisition.
invest in research and better in the last few years,
but I wouldn't say it is It wasn't a cohesive product…
development to ensure that it We ultimately chose to look
seamless as of yet. The
continues to be the best elsewhere and found
synergy piece…is still not
solution in the market.” there.” Abnormal.”
Splunk customer Palo Alto Networks customer Abnormal customer
VP, Publicly traded automobile manufacturer CISO, Global engineering firm Director, Personal injury law firm

Source: CB Insights buyer transcripts — Splunk, Palo Alto Networks, Abnormal. 44

Cybersecurity

In some cases, verticalization wins out

“As a financial services organization, our company had specific requirements for a
product or solution that can integrate with multiple tech stacks, including SAP,
Salesforce, Microsoft, and other vendors…
The reason why CrowdStrike stood out was that it has more strength in the banking
and financial services industry than Palo Alto, Cisco, or Juniper.”

CrowdStrike customer
VP, Financial services company

Source: CB Insights buyer transcript – CrowdStrike. 45

Cybersecurity

AI vs. AI dogfights
redefine data security

46
Cybersecurity

Cyber attacks continue to rise, with AI a driving force

Global average weekly cyber attacks per organization hit an all-time peak in Q2’23

1,400

1,258
1,200

1,000

800

600

400

200

0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2
2021 2022 2023

Source: CheckPoint. 47
Cybersecurity

ChatGPT drives an explosion of email phishing attacks

One email security vendor recorded a 1,256% increase in phishing emails from Q4’22 to Q3’23

Malicious phishing emails

Source: SlashNext. 48
Cybersecurity

Strategic and financial (PE) acquirers go after email

security startups…
Recent M&A exits and take-private transactions in the email security market

Source: CB Insights Market Report — Email security. 49

 Cybersecurity

…While investors pile into the breach & attack simulation market
The largely early-stage space has seen rising investor interest, as more startups use
generative AI to emulate threats

$1M 20

$1M 18
Equity deals Select early-stage deals in 2023
16
14
$1M
Amount
14
Company
$1M Round
11 12
$1M
10 $5M
10 Pre-seed
$1M
8
$0M 8
$1M
$0M 6 Seed

$0M 4

$0M 2
Undisclosed
Seed
$0M 0

2020 2021 2022 2023

Source: CB Insights company profiles — Jericho Security, Brightside AI, ASants; Breach & attack simulation Market Report. 50
Cybersecurity

GenAI could be a major boon for cyber attack prevention

“SlashNext is working on a generative AI-based solution, which means it will

generate its own kind of phishing and malware and it will train the software to
automatically be aware of any new kind of threats arriving in the market…
That is a differentiator from the technology perspective.”

SlashNext customer
Senior Design Engineer, Fortune 500 company

Source: CB Insights buyer transcript — SlashNext. 51

Financial services & insurance

Banks get AI FOMO

52
Financial services & insurance

AI is the talk of the town in banking

Just 1 month into Q1’24, mentions of AI and related terms on earnings
calls hit record high among 10 largest US banks
21

Who talks about AI the most?

Mentions of
Bank
AI since 2019

44

36

29

25

Source: CB Insights — Earnings Transcripts Analytics (date reflects quarter call occurred). 10 largest US banks by consolidated assets (per the 53
Federal Reserve) are JPM Chase, Bank of America, Wells Fargo, Citibank, U.S. Bank, PNC, Goldman Sachs, Truist, Capital One, and TD Bank.
Financial services & insurance

With many viewing genAI as an operational silver bullet

How valuable would implementing this
use case be for your organization?

Financial document search and synthesis with genAI

"Consumers interacted over 20
Extremely valuable.
36%
Extremely
42% Fairly million times last year with Fargo,
valuable valuable
Fairly valuable. our AI-powered virtual assistant.”
Enhanced virtual assistants with genAI Somewhat valuable. CEO Charles Scharf, Q1’24 earnings call
45% 35% Not very valuable.

Not valuable at all.

Capital markets research with genAI
N/A or “I don’t know”.
34% 38%
Regulatory code change consultant with genAI "...to the extent that we can deploy
[AI] wisely, … we can always hire a
25% 41% little less if we see the efficiencies
coming through and redeploy the
Personalized financial recommendations with genAI people we have.”

36% 40% CEO Brian Moynihan, Q1’24 earnings call

Source: Google Cloud Gen AI Benchmarking Study, July 2023, Financial services respondents n=200; CB Insights — Earnings Transcripts Analytics 54
(date reflects quarter call occurred).
Financial services & insurance

But some banks are more prepared for the AI boom than others
The top 3 players in the CBI Retail Bank AI Readiness Index are building moats via
internal R&D, IP, and key AI hires

Hired Abhijit Bose, a head of

engineering at Facebook AI
Research, to lead its enterprise
machine learning department.

Has applied for 9 patents since

2021 pertaining to federated
learning — a privacy-preserving
and decentralized AI training
method.

Launched Turing text-to-SQL, a

tool that converts natural
language queries into SQL code,
via its Borealis AI institute.

Source: CB Insights. The Retail Bank AI Readiness Index assesses how prepared top banks are to adopt and respond to rapidly evolving AI
technologies. Our analysis looked at the 50 largest retail banks in the Americas and Europe by market cap. 55
Financial services & insurance

Banks will have to overcome significant challenges

to realize AI’s potential
The four greatest obstacles to reaching financial services orgs’ AI goals

Data issues: privacy,

sovereignty, and disparate 38%
locations

Recruiting and retaining AI

experts and data scientists 32%

Lack of budget 28%

Sufficient data sizes for

model training and 27%
accuracy

0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%

Source: Nvidia survey on the State of AI in Financial Services: 2024 Trends. 56

Financial services & insurance

To get a foot in the door, some are linking up with

AI chatbot developers
Select AI chatbot developers with investors and clients from the banking sector

AI chatbot developer Banking investors Banking clients

Source: CB Insights Market Report — AI-powered chatbots. 57

Financial services & insurance

While others double down on internal R&D and shop around for
outside talent & capabilities
Among the 10 largest US banks, the top 3 AI
patent filers by # of applications since 2021

Key AI acquisitions in 2023

967
Bank Acquisition target Application

AI investment
insights &
640 document
extraction

AI personal
274 assistant

0 100 200 300 400 500 600 700 800 900 1000

Source: CB Insights — patent analytics and M&A data. 58

Financial services & insurance

Blockchain’s uphill
finserv battle

59
Financial services & insurance

Funding to blockchain startups is in the gutter

The sector has yet to rebound following crypto blow-ups

$35. 0B 2,357 2,500

$30. 0B
2,000

$25. 0B 1,602 Deals

1,451
1,500
$20. 0B

$15. 0B
797 Funding 1,000

$10. 0B $8.2B
500
$5.0B

$3.5B $27.3B $29.4B

$0.0B 0

2020 2021 2022 2023

Source: CB Insights Expert Collection — Blockchain. 60

Financial services & insurance

Despite bitcoin’s recent ascent, the underlying tech has yet to see
significant commercial traction…
The price of BTC trending up, driven in part by recent spot bitcoin ETF approval

Source: Google Finance. Data as of February 1, 2024. Note: Spot bitcoin exchange-traded funds (ETFs) are open-end funds that allow investors to 61
gain exposure to bitcoin without holding it, making it more accessible to everyday investors.
 Financial services & insurance

…as blockchain has slipped off executives’ radar, paling in

comparison to genAI hype
3,000
Earnings call mentions of blockchain vs. generative AI

2,546
2,500

2,000 Generative AI
1,629 1,841 mentions
1,500

1,000

444
500
215 235 229 229
111 158 164 157
100 121 68 Blockchain
27
0
47 mentions
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2021 2022 2023

Source: CB Insights — Earnings Transcripts Analytics. 62

Financial services & insurance

Early blockchain experiments in banking have hit roadblocks

Major blockchain trade finance networks shut down

Entered insolvency 6/6/2022 Entered insolvency 2/24/2023 Announced shutdown 10/27/2023

Joint venture 30+ member banks including 9 shareholder banks including

with 10+ banks including

Source: Media reports. 63

Financial services & insurance

Financial services leaders haven't written blockchain off,

but they aren’t racing to deploy it either
Crypto complications continue to weigh on the sector

“In terms of…blockchain and crypto, it's, as far as “The first [weakness of this solution] is not really a
we see, it's still early days, still emerging weakness of ADDX, but rather a lack of understanding in
technologies. But as we've said in other venues, the market and among clients about the difference
we see the potential for blockchain. We see the between blockchain and crypto. Crypto is volatile and
potential for stablecoins…And we'll see how they risky, whereas blockchain is simply a technology to
play out in the market. And if there's things to facilitate contracts and tokenize assets. When I tell
scale, we want to be part of that.” customers that ADDX uses blockchain technology, they
are sometimes confused and I have to explain further.
CEO Ryan McInerney, Q2’23 earnings call
Some are convinced, but still worried that it may be
related to crypto.”
ADDX customer
Senior Vice President, Publicly traded financial services company

Source: CB Insights — Earnings Transcripts Analytics; CB Insights buyer transcript — ADDX. 64

Financial services & insurance

The future of finserv incumbents’ limited blockchain experiments

will hinge on adoption & revenue
Select blockchain projects & recent announcements

Digital token JPM Coin Launches Citi Token Successfully pilots Launches platform that
processes $1B worth of Services, which converts transferring tokenized tokenizes gold bars to
transactions a day, though client deposits into digital assets across public and simplify tracking & trading
still a fraction of the $10T tokens to enable instant private blockchains with (11/1/2023)
processed daily by JPM cross-border payments & Chainlink (8/31/2023)
(10/26/2023) automated trade finance
solutions around the
Onyx launches JPM Coin-
clock (9/18/2023)
based programmable
payments, scales back NFT,
metaverse work (11/10/2023)

Source: Company & media reports. 65

Financial services & insurance

Extreme weather
is an opportunity
for insurtech

66
Financial services & insurance

The US saw a record number of billion-dollar

weather disasters in 2023

Source: National Oceanic and Atmospheric Administration via Bloomberg (data as of September 2023). 67
Financial services & insurance

Extreme weather and climate change have gripped public attention

Number of related media articles published per month

Source: CB Insights – News Mentions Analytics. 68

Financial services & insurance

Unpredictable and costly weather disasters see insurers exit

markets, reduce coverage
Natural disaster losses amounted to $250B in 2023 — less than half insured ($95B)
$200.0B
Insurer Select exit
$180.0B Stopped applications for new business
and personal lines of P&C insurance in
$160.0B California (2023)

$140.0B Paused new homeowners,

commercial insurance, and condo
policies in California (2022)
$120.0B

$100.0B
Discontinued new auto, home, and
umbrella policies in Florida (2023)
$80. 0B

$60. 0B

$40. 0B

$20. 0B
$40B $37B $58B $174B $96B $63B $96B $133B $124B $95B
$0.0B

2014 2015 2016 2017 2018 2019 2020 2021 2022 2023

Source: Munich Re (adjusted to 2023 values based on local CPI) via Bloomberg; media reports. 69
 Financial services & insurance

Insurers turn to solutions for predicting and managing

extreme weather risk
Funding to these solutions remains steady while dealmaking ticks up

$500M 23 25

$450M
Deals Key early-stage
19
$400M
17 18 20 deals in 2023
$350M Amount
Company
Round
$300M 13 Funding 15

$17M
$250M
$210M Series A
$200M 10

$150M $14M
Series A
$100M 5

$50M
$243M $147M $442M $237M $5M
Seed
$0M 0

2019 2020 2021 2022 2023

Source: CB Insights Market Reports — Catastrophe modeling, Climate and weather financial risk modeling, Weather risk intelligence, Wildfire risk 70
intelligence, and Flood intelligence.
Healthcare & life sciences

The great AI drug

race heats up

71
Healthcare & life sciences

Small-molecule drugs are central to the pharmaceutical industry

From 2013 to 2022, nearly 2 in 3 drugs approved by the FDA were small-molecule
2013 89% of total 2014 66% of total 2015 71% of total 2016 50% of total 2017 63% of total

27 approvals 22 approvals

Small-molecule
41 approvals drug approvals
45 approvals 46 approvals
Non-small-molecule
2018 64% of total 2019 67% of total 2020 66% of total 2021 62% of total 2022 46% of total drug approvals

37 approvals

53 approvals 50 approvals
48 approvals
59 approvals
Source: US Food and Drug Administration via Chemical & Engineering News. Note: Small-molecule drugs are organic compounds with low 72
molecular weights. They engage biological targets for therapeutic purposes, and most can be administered orally.
Healthcare & life sciences

AI is disrupting the small-molecule drug development landscape

Source: CB Insights research — Pharma AI Market Map. Note: This market map shows technology companies that use AI organized by market. It is 73
not intended to be exhaustive of companies in the space, and categories are not mutually exclusive.
Healthcare & life sciences

AI is boosting the speed and precision of small-molecule drug discovery

Select win reasons for leading players in the space

Integrates machine learning Utilizes a leading biopharma Integrates AI for both target
with complex biology to AI platform to accelerate identification and chemistry,
enhance drug discovery drug target selection and enhancing drug discovery
efficiency and accuracy. development in complex precision.
diseases like cancer.

Source: CB Insights Market Scorecard — AI small molecule drug discovery platforms. 74

Healthcare & life sciences

The race for the first AI-discovered drug is picking up rapidly, with
candidates approaching late-stage trials
All eyes were on Insilico Medicine as it entered the first fully AI-generated
small-molecule drug into Phase II clinical trials in June 2023

Source: CB Insights company profile — Insilico Medicine. 75

Healthcare & life sciences

Venture activity in AI small molecule drug discovery holds strong

heading into 2024
Equity funding ticked up while deal count held steady in 2023, bucking the broader
venture slowdown
16
16
$2000M

14

12
$1500M
10
Deals 10

$1000M
7 7 8

Funding 6

$500M
$456M 4

2
$991M $1.9B $374M
$0M 0

2020 2021 2022 2023

Source: CB Insights Market Report — AI small molecule drug discovery platforms. 76

Healthcare & life sciences

Pharma incumbents will continue to work with these companies to get ahead
The top 3 players in the CBI Pharma AI Readiness Index have all linked up with AI-
focused drug discovery startups

Partnered with AI
precision medicine startup
Exscientia to design pre-
clinical drug candidates.

Led a Series D round for

Recursion, which develops AI
models for drug discovery.
Recursion later went public and
partnered with Bayer to develop
new treatments for fibrotic
diseases.

Entered into a multi-target

discovery agreement with
genAI-focused biotech Insilico
Medicine.

Source: CB Insights business relationships data and research. The Pharma AI Readiness Index assesses how prepared top pharma companies are 77
to adopt and respond to rapidly evolving AI technologies. Select AI partnerships shown for top 3 players. Partnerships and licensing agreements
with AI companies factor into Execution scores.
Healthcare & life sciences

Digital therapeutics
(DTx) & wellness
consolidates

78
 Healthcare & life sciences

Investors turn away from DTx & wellness amid lack of payer
support and reimbursement challenges
Disclosed equity funding to DTx & wellness companies
$7.0B 600

$6.0B 481
500

$5.0B
400

$4.0B
292 308
261 Deals 300
$3.0B
197
200
$2.0B
Funding
$1.0B $0.7B 100

$1.2B $1.6B $6.0B $1.8B

$0.0B 0

2019 2020 2021 2022 2023

Source: CB Insights Expert Collection — Digital Health. Note: Digital therapeutics (DTx) & wellness refers to the use of digital technologies, such as 79
mobile apps, software programs, and other digital platforms, to deliver therapeutic interventions and promote overall well-being.
Healthcare & life sciences

The industry was rocked by bankruptcy of DTx trailblazer Pear

Therapeutics in 2023
Reimbursement challenges weighed heavily on Pear, which was sold for parts in May ‘23

Source: CB Insights valuation data – Pear Therapeutics. Note: chart shows valuation by round. 80
Healthcare & life sciences

Amid funding downturn, DTx & wellness M&A is gaining

momentum
8
9
8 M&A deals
8

6
5
5
4 4
4
3 3
3
2
2

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2022 2023

Source: CB Insights Expert Collection — Digital Health. 81

Healthcare & life sciences

Well-funded DTx & wellness companies are being snapped

up by peers amid consolidation
Select M&A deals in 2023

Total equity funding: $284M Total equity funding: $55M Total equity funding: $27M

Assets sold to Click Acquired by Acquired by

Therapeutics, Harvest Bio, Beacon Biosignals Big Health
Nox Health, and Welt

Source: CB Insights M&A transaction data — Pear Therapeutics, Dreem, Limbix. 82

Healthcare & life sciences

Insurance coverage will remain a critical challenge

for DTx & wellness in 2024
Looking ahead this year, therapeutics & wellness buyers identify coverage
as key area for growth

“Luminopia may have room for growth by getting “Oshi Health has room for growth in expanding their
approval with insurance companies. However, the coverage to all 50 states and increasing their reach
insurance companies are hesitant to cover the cost of into health plans. Currently, Oshi Health is
dichoptic therapy, which is at least $150 a month. integrated with Marpai, Quantum Health, UHC, and
Luminopia may need to work with the insurance Aetna and serves 75% to 80% of their clients' needs.
companies and find success that even doctors have As demand for their product offering increases, [my
failed to achieve…The price point of $45 per patient with client] expects to see continued growth within the
the Medicaid population may not be sustainable.” health plans and across the United States.”
Luminopia customer Oshi Health customer
President, Pediatric opthalmology company (December 2023) Director, Publicly traded insurance company (December 2023)

Source: CB Insights buyer transcripts — Luminopia, Oshi Health. 83

Healthcare & life sciences

The brain becomes a

fierce tech battleground

84
Healthcare & life sciences

Elon Musk’s Neuralink just announced its first human brain

implant — and its workforce is rising quickly as it revs up

Source: CB Insights company profile — Neuralink. 85

Healthcare & life sciences

Investors are betting on brain-computer interfaces to initially help

motor-impaired patients, and human trials are kicking off
Company Headquarters Total equity funding Latest funding round Recent milestone

Series D
US $680M Human trial in progress
(08/07/2023)

Series C
US $125M Human trial in progress
(12/15/2022)

Series A Received FDA Breakthrough

US $96M
(11/01/2023) Device designation

Series B Received FDA Breakthrough

US $53M
(12/19/2022) Device designation

Source: CB Insights — deals data for select brain-computer interface companies. 86

Healthcare & life sciences

Meanwhile, brain activity data can now be coupled with AI to ‘read

minds’ with eerie accuracy
Original images a participant was viewing compared to those reconstructed from their
fMRI data and Stable Diffusion

Original images

Reconstructed images

Source: Yu Takagi, Shinji Nishimoto, “High-resolution image reconstruction with latent diffusion models from human brain activity.” 87
Healthcare & life sciences

As the brain-computer interface landscape matures, privacy

debates will reach a new fever pitch
Data privacy media mentions were at an all-time high in 2023 — but tech that can
capture even superficial information on inner thoughts will intensify scrutiny

Source: CB Insights — News Mentions Analytics. 88

Retail & consumer

AI sales agents
flood the e-commerce
landscape

89
 Retail & consumer

As e-commerce grows, retailers don’t want

to be caught on the back foot
E-commerce sales are poised to increase, making up a greater share of total retail sales
$12. 0T 35.0%

$10. 0T Share of total 30.0%

retail sales
22.2% 23.0% 25.0%
$8.0T
20.3% 21.2%
18.9% 18.9% 19.5% E-commerce 20.0%
$7.5T retail sales
$6.0T $6.9T
$6.3T $8.0T
15.0%
$5.8T
$5.0T $5.3T
$4.0T
10.0%

$2.0T
5.0%

$0.0T 0.0%

2021 2022 2023 2024 2025 2026 2027

Source: Insider Intelligence | eMarketer. Years 2023 to 2027 are projections. 90

 Retail & consumer

But finding new customers is cutting deeper into profits over time
E-commerce brands & retailers lose $29 for each new customer acquired — over
200% more than in 2013
$40

$35

$29
$30

$25

$20
+222%
Since 2013
$15

$9
$10

$5

$0

2013 2022

Source: SimplicityDX. 91
Retail & consumer

Retailers turn to startups to find new ways to engage

and convert online shoppers…

Source: CB Insights research — Retail Tech 100. 92

Retail & consumer

…As advances in NLP allow for AI bots that can haggle and upsell
Early-stage companies in the space include Retail Tech 100 winners Nibble and Zowie

Company Latest round Select investors

$3.3M seed
Venrex
(March 2022)

$14M Series A Tiger Global Management,

(May 2022) Gradient Ventures, Jack Altman

$0.04M accelerator
Fashion Technology Accelerator
(March 2023)

$1.5M pre-seed
Anthemis, Wischoff Ventures
(October 2023)

Source: CB Insights company profiles — Nibble Technology, Zowie, BATNA, ergo. 93

Retail & consumer

AI price negotiation boosts customer engagement

“India is a very price-sensitive market, and haggling is in our DNA. Customers

are used to shopping and bargaining in an offline setting, so when it was
presented to them in an online setting, it was something very new,
something worth trying, and a lot of them just did it for the fun of it.”

Nibble Technology customer

Founder, Retail company

Source: CB Insights buyer transcript – Nibble Technology. 94

Retail & consumer

Expect to see more AI sales agents built into

e-commerce & social channels
AI sales chatbot developers flock to platforms like Shopify, Instagram

Commerce platform Select partners offering AI sales chatbot integrations

Source: CB Insights business relationships data; company websites. 95

Retail & consumer

Retailers tackle shrink

with AI loss prevention

96
 Retail & consumer

Inventory loss costs retailers $100B/year & it’s not getting better
% of US retail sales lost to shrink has remained stubbornly high

2. 0%

1.6% 1.6% 1.6%

1. 5%
1.4% 1.4% 1.4% 1.4%
1.3%

1. 0%

0. 5%

0. 0%

2015 2016 2017 2018 2019 2020 2021 2022

Source: National Retail Federation. Note: Shrink refers to inventory loss due to theft or error. 97
 Retail & consumer

Executives express growing concern around shrink

Mentions of retail shrink on earnings calls
140

120 102
mentions
100

80

60
50 50 49
43
40

20

2019 2020 2021 2022 2023

Source: CB Insights — Earnings Transcripts Analytics. 98

Retail & consumer

Incumbents look to tech to prevent inventory loss

“…we are making investments in “...shrink was in line with last year's results,
technology to improve store despite industry-wide challenges, driven by
productivity, lower digital fulfillment our proactive customer service, tech-driven
solutions, industry-leading asset protection
costs, and reduce waste and shrink.
program, and our penetration of rural
These improvements will drive an stores. We're developing...RFID technology
incremental $1B of cost savings in embedded in power tools to prevent theft.”
2023.”
EVP of Stores Joe McFarland, Q3’23 earnings call
CFO Gary Millerchip, Q1’23 earnings call

Source: CB Insights — Earnings Transcripts Analytics (date reflects quarter call occurred). 99
Retail & consumer

Investors funnel money toward AI-powered loss prevention

Loss prevention tech companies raise massive funding rounds in 2023

Company HQ Latest funding round Select customers Round investors

Alkeon Capital Management (Lead),

$100M Series D
US First Round Capital, Lightspeed Venture
(October 2023)
Partners, next47

$71M Series A
Ireland Crosspoint Capital Partners (Lead)
(May 2023)

Goldman Sachs Asset Management

$60M Series C
Canada (Lead), BDC Venture Capital, OMERS
(July 2023)
Ventures

Source: CB Insights company profiles — Verkada, Everseen, Solink. Note: Examples of AI applications 100
for in-store loss prevention include real-time camera feed analysis and self-checkout behavior scanning.
Retail & consumer

AI loss prevention can also enable other in-store tech,

like self-checkout, to drive efficiency

“The Everseen product identifies and pauses transactions when merchandise is

present at the self-checkout but not scanned…to prevent loss. Additionally, it uses
AI to identify merchandise and ensure that customers are scanning the correct item.
… Throughout our organization, from store level to senior management, there is a lot
of support for Everseen and the value that it adds. It has become a necessary part
of our business and the protection of our assets.”

Everseen customer
Director, Fortune 500 company

Source: CB Insights buyer transcript – Everseen. 101

Retail & consumer

AI sends a shockwave
through gaming

102
 Retail & consumer

Video games represent a massive — and growing — opportunity

With $190B+ in annual revenues, gaming is 5x bigger than the movie industry
$250.0B

$227.6B
$200.0B
$211.7B
$201.6B $195.6B $195.1B
$182.0B
$150.0B

$142.5B
$128.2B
$100.0B
$113.8B

$50. 0B

$0.0B

2017 2018 2019 2020 2021 2022 2023 2024 2025

Source: IDC. Years 2023 – 2025 are projections. 103

Retail & consumer

But making games isn’t cheap

AAA games can require hundreds of developers and 9-figure budgets

Star Citizen
(in development) $600M+

Cyberpunk 2077
(2020) $316M

Starfield
(2023) $200M

The Witcher 3:
Wild Hunt $81M
(2015)
$0M $100M $200M $300M $400M $500M $600M

Source: Media reports. 104

Retail & consumer

Developers rush to adopt genAI even as they worry about its misuse

Com pan yA
Com pan yA

64% 84%
of developers use of developers are
(or want to use) genAI concerned about the
or have colleagues ethics* of using genAI
who use it for development

Source: Game Developers Conference. *Developers express concern about the possibility 105
of copyright infringement, worsening labor conditions for human artists, and more.
 Retail & consumer

AI could cut dev costs while making gameplay more immersive

The cost of outsourcing development for a
single in-game character vs. using AI to do it

$800

$700
$675

$600
“The average game time
increased by 5% due to
$500
-94% the interaction with AI
$400 with AI [non-player characters].”
$300

Ryan Yang, founder & CEO of MageCosmos,

$200
in a case study on Inworld’s AI NPCs

$100 $40
$0

Outsourced labor AI

Source: Crypko via Bloomberg; Inworld. 106

Retail & consumer

And the market for AI-generated content is heating up

Most synthetic gaming companies — which use AI for lifelike non-player characters,
object animations, & more — ended 2023 with strong YoY headcount growth

Source: CB Insights headcount data — Synthetic gaming content Market Report. 107
Retail & consumer

Incumbents will lead the shift with developer- and creator-focused tools
Recent genAI releases from major gaming companies and their partners

Gaming incumbent Partners (if applicable) Release description

Unity releases two creator-focused AI engine platforms — Sentis and

Muse — and launches a marketplace for 3rd-party AI tools with
integrations to Inworld, modl.ai, and others (June 2023)

Roblox launches Assistant, an AI copilot that helps creators to code,

N/A build worlds using natural language commands, and more (September
2023)

Microsoft’s Xbox announces multi-year partnership with Inworld to build

AI-powered game design and narrative tools (November 2023)

Epic’s Unreal Engine links up with 3D model creator Cybever to give

developers access to Cybever’s 3D assets (January 2024)

Source: CB Insights business relationships data; company and media reports. 108
Industrials

Humanoid robots come

for manufacturing

109
Industrials

Humanoid robots have been in development for decades

Select advances since 1960 showcase progression of humanoid capabilities

1960s–70s 1980s–90s 2000s 2010s 2020s

1 2 3 4 5

• Zero-Moment Point • P-Series robots (by • QRIO (by Sony): advanced • Atlas (Boston Dynamics): • Optimus (Tesla): leverages
(ZMP) stability theory Honda): improved motion, face and voice search and rescue self-driving AI for
is introduced walking, stair climbing recognition capabilities navigation
• Pepper (Softbank): • Ameca (Engineered Arts):
• WABOT-1 (by Waseda • Robonaut 1 (by NASA): • ASIMO (by Honda): walking, emotion recognition and genAI-powered robot
University): one of the space exploration running, human interaction formation capable of holding
earliest full-scale capabilities. R2 goes to
• REEM-A (by Pal Robotics): • Sophia (Hanson Robotics): conversations in dozens
humanoid robots space in 2011.
AI-powered humanoid of languages, mimicking
first fully autonomous
European biped humanoid capable of conversation human facial expressions/
and facial expressions body language
robot

• Wakamaru (Mitsubishi
Heavy Industries): in-home
assistance, companionship

Source: Media reports, press releases, journal articles, and company websites. Note: This timeline is not exhaustive of all notable events or 110
capabilities for featured models.
Industrials

Leading public players in humanoid robotics face a growing

network of innovative startups
Boston Dynamics and Tesla currently lead in market and execution strength

Source: CB Insights ESP Ranking — Industrial humanoid robots. Note: This company ranking is not exhaustive of players in the market. 111
Industrials

Humanoid robotics startups gain momentum despite broader

venture slowdown
Dealmaking in the space ramped up in 2023

$700M Deals
11 12
$600M

9 10
$500M
8 8
8
$400M

Funding
6
$300M $262M
4
$200M

$100M 2

$308M $582M $388M

$0M 0

2020 2021 2022 2023

Source: CB Insights Market Report — Industrial humanoid robots. 112

Industrials

As the market develops, competition for market share gets heated

Tesla CEO Elon Musk discusses competitor Boston Dynamics on Q3’23 earnings call

“…If you no longer have a constraint on people, effectively,

you’ve got a humanoid robot that can do as much as you’d
like, your economy is wisely infinite or infinite for all
intents and purposes. So, I don’t think anyone is going to
do it better than Tesla, not by a long shot. Boston
Dynamics is impressive, but their robot lacks a brain or
like the Wizard of Oz, whatever. Yes, lacks a brain.”

Elon Musk, Tesla Co-founder and CEO

Source: CB Insights — Earnings Transcripts Analytics (date reflects quarter call occurred). 113
Industrials

Humanoid robotics companies want to disrupt manufacturing

Manufacturing mentioned as a focus area 4x as much as construction

Defense and disaster response, 7%

Construction, 7%
Manufacturing
28%
Healthcare
18%

Retail Logistics

18% 22%
Source: CB Insights Analyst Briefings, company websites, press releases, and media reports. 114
Venture

Corporate venture
refocuses on
strategic fit

115
 Venture

Corporate VC dealmaking is in retreat in the face of

macroeconomic challenges
CVC-backed equity deals hit a 4-year low in 2023 amid corporate belt-
tightening and subdued returns from startup exits
$200.0B 7,000

$180.0B
6,000
$160.0B 5,142 5,197
$140.0B 5,000
Deals
$120.0B
3,526 3,641 3,545 4,000

$100.0B

$80.0B Funding 3,000

$60.0B
$55.1B
2,000

$40.0B
1,000
$20.0B
$58.2B $71.1B $174.4B $102.4B
$0.0 B 0

2019 2020 2021 2022 2023

Source: CB Insights financing data. 116

 Venture

One exception: The most active CVCs are now based in Japan
Japanese CVC activity has held steady with Top CVCs by company count in Q4'23
strong government backing and cash stockpiles
Company
Investor count Country
800
697 1 Mitsubishi UFJ Capital 22
700
608 Deals
572
2 SMBC Venture Capital 18
600
569 553 548
3 Mizuho Capital 15
500
4 Google Ventures 14
400 353
5 KB Investment 11
300
5 Samsung NEXT 11
200
7 SBI Investment 10
100 Mitsui Sumitomo Insurance
8 8
Venture Capital
0

2017 2018 2019 2020 2021 2022 2023 8 SDIC Venture Capital 8

Source: CB Insights financing data. Chart shows equity deals backed by Japan-based corporate VCs. 117
 Venture

Amid the subdued venture investing environment, we expect CVC

formation broadly will continue to slow
Number of new corporate venture arms by year
350

300

295
250 274
246 244
200
230 162
new CVCs
150
163
100

85
50
63
0

2015 2016 2017 2018 2019 2020 2021 2022 2023

Source: CB Insights. Note: CVC formation date based on when investor made first disclosed investment. 118
Venture

Existing CVCs are refocusing on strategic fit in

sectors they know well
CVC Parent Development

M12 refocuses investment strategy to align with Microsoft's priorities

Microsoft
(January 2023)

Verizon brings venture arm under its corporate strategy group

Verizon
(January 2023)

Total Energies winds down its climate-focused venture arm, pivots to

TotalEnergies TotalEnergies On accelerator focused on early-stage energy companies
(May 2022)

Salesforce Ventures' portfolio is a key part of Salesforce's M&A pipeline,

Salesforce with two recent acquisitions (Airkit, September 2023; Spiff, December
2023) preceded by investment

Source: Media reports. 119

Venture

Unicorns need a new

playbook to survive

120
 Venture

Startups compete for shrinking funding pool

Global equity dollars and deals hit lowest levels in 6+ years in Q4’23

$200.0B 14,000
12,631
$180.0B
11,134 11,030 12,000
$160.0B
10,639
9,881 9,691 9,943
$140.0B 8,856 8,688 -24% 10,000

$120.0B 7,419
7,027 QoQ 8,000
$100.0B
Deals
6,000
$80. 0B
Funding 6,169
$60. 0B $51.0B 4,000

$40. 0B

2,000
$20. 0B

$142.5B $157.7B $168.1B $180.3B $152.9B $118.6B $84.4B $70.3B $72.0B $58.7B $66.8B
$0.0B 0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2021 2022 2023

Source: CB Insights research — State of Venture 2023 Report. 121

 Venture

Startup mortality on the rise as macroeconomic headwinds persist

Number of publicly disclosed startup deaths
450

400

309
350
Deaths
300

250

200

150

100

50

402 253 92 182

0

2019 2020 2021 2022 2023

Source: CB Insights financing data. Note: Excludes exited companies and bankruptcies that did not result in dissolution. 122
 Venture

Failed startups flush billions of dollars down the toilet

Startups that shut down in 2023 raised $4.1B in equity from 2019 to 2022
$1.8B

$1.6B

$1.4B

$1.2B

$1.0B

$0.8B
$0.7B
$0.6B

$0.4B

$0.2B

$0.8B $1.1B $1.6B

$0.0B

2019 2020 2021 2022

Source: CB Insights financing data. Note: Excludes exited companies and bankruptcies that did not result in dissolution. 123
Venture

With tech boom in the rearview mirror, some unicorns can’t sustain
valuations and are forced to close shop
Top unicorn deaths in 2023 by latest disclosed valuation prior to shut-down

Valuation: $4B Valuation: $3.8B Valuation: $2.3B

Total equity funding: Total equity funding: Total equity funding:
$850M $836M $446M

Source: CB Insights company profiles — Olive, Convoy, Zume. 124

Venture

Remaining unicorns could be short on cash and face valuation

loss, M&A, or closure in 2024
Top-valued unicorns that haven’t raised in 2+ years

Source: CB Insights Expert Collection — Unicorns. 125

Note: Graphic shows select unicorns in the US and Europe that haven’t disclosed a funding round since 2021.
Our most popular
research
Research
Generative AI predictions for 2024
The 2024 Tech IPO Pipeline
Generative AI Bible
State of AI 2023 Report
The future of big tech in 10 charts

Market reports
Large language model developers
ESG reporting, monitoring, & data analysis
AI development platforms
Real-world data

126
Expert Collections
CB Insights customers can find and
track companies using our analyst-
curated Expert Collections.

Filter by technology, funding, valuation,

business model, and more.

Track companies and monitor targets

with real-time data.

CB Insights helps the world’s leading companies understand everything they

need to know about disruptive technologies — find out more about why our 127
customers love us here.
Buyer Transcripts
Buyer profile
Understand the buyer — their title, their
organization, and their experience

Purchase intelligence
How much did they spend and when

Consistent structure
Moderated & structured conversations
make transcripts easy to scan

Deep insights
Details on everything from evaluation
criteria to competitors to satisfaction

CB Insights helps the world’s leading companies understand everything they

need to know about disruptive technologies — find out more about why our 128
customers love us here.
▮▮▮▮▮IIII
Try our platform for free

129