p perplexity
docs Consumer AI Template
Investor Deck — [REPLACE: Round]

The answer engine for the internet.

Where you go when you need a direct, cited answer — not ten blue links or a chatbot hallucination. Perplexity is replacing search as the default way people ask the web a question.

Founders[REPLACE: Founder names]
Raising$[REPLACE: amount] · [REPLACE: round]
Date[REPLACE: Month Year]
perplexity
02 · Problem
The problem

Search is broken. Chatbots lie. Neither gives you an answer you can trust.

Search returns links, not answers

You type a question. You get 10 SEO-optimized results. You click, skim, cross-reference, and synthesize. The work happens in your head.

8.6 minAvg time to answer a research question via Google

Chatbots hallucinate confidently

LLMs produce fluent text, but without citations you can't tell what's real. You have to verify everything, which defeats the speed.

17%Of LLM answers contain at least one fabricated fact

Both waste knowledge workers' time

Every analyst, student, lawyer, and operator runs the same cycle 30–50 times a day: question → search → read → synthesize. A productivity tax paid in minutes.

2.3 hrs/dayKnowledge worker time on information retrieval
perplexity
03 · Why Now
Why now

The three shifts that make answer engines inevitable.

01 / 03

LLMs can finally synthesize

GPT-4 class models crossed the threshold where multi-source synthesis is reliably better than human-eyeballing. Citations + reasoning = trust.

Synthesis quality ↑ 5.2× since 2022
02 / 03

Search habit is breaking

Gen Z doesn't Google — they ask. TikTok, Reddit, ChatGPT. The "default answer source" is up for grabs for the first time since 1998.

40% of 18–24s don't use Google as default
03 / 03

The economics finally work

Inference cost dropped ~90% in 24 months. A cited, synthesized answer now costs cents, not dollars. Advertising + subscription both viable.

GPT-4-class inference: $30/M → $3/M tokens
perplexity
04 · Solution
Our solution

Ask a question. Get an answer. See the sources.

Perplexity is the answer engine built for how people actually want to use the internet. You type a question in natural language. We retrieve the most relevant, recent sources in real time. An LLM synthesizes them into a direct answer with inline citations.

Every answer is grounded, cited, and editable. You can ask follow-ups that inherit context. You can browse the sources. You can share the thread. It's the workflow millions of people hacked together with ChatGPT + Google — rebuilt natively.

Our conviction: the default question-asking surface on the internet is about to change. We're building the thing it changes to.

Your question → How did inflation affect Q4 housing? nytimes.com bloomberg.com fed.gov Synthesize + cite Mortgage rates rose 0.6pp in Q4, pulling existing home sales down 7% YoY 1 2 with inline citations · browse sources
perplexity
05 · Product
What we ship

One engine. Three surfaces. Every kind of question.

Perplexity Search

The web-first answer engine. Real-time retrieval, inline citations, follow-up threads, copilot mode for deeper queries.

  • Inline citations per claim
  • Follow-up threading
  • Image + chart rendering

Pro

The subscription tier with frontier models, unlimited Pro Searches, file & image uploads, and Spaces for research workflows.

  • GPT-4 / Claude / Llama backends
  • File + image context
  • Shareable Spaces

Enterprise & API

SSO, audit, internal knowledge sources. API gives developers the answer engine as a backend primitive for any product.

  • Internal + web retrieval
  • SSO, SCIM, audit logs
  • Sonar API for developers
perplexity
06 · Demo
The product

Ask. Answer. Cited.

perplexity.ai · Pro Search
Answer
Enterprise AI spending jumped 43% YoY in 20261, driven by three converging factors. First, F500 CIO surveys show AI moved from "experimental" to "core budget" for 71% of large enterprises2. Second, inference costs dropped ~90% over 24 months, making previously-blocked use cases economical3. Third, Microsoft and Google both tied AI capabilities to existing productivity suite renewals, accelerating line-item adoption1.
[1]
Q4 Enterprise AI Spend Report — up 43% YoY
morganstanley.com
[2]
CIO Survey: 71% put AI in core tech budget
gartner.com
[3]
GPT-4-class inference pricing trajectory 2024–26
semianalysis.com
perplexity
07 · Market
Market

We're taking share from the most valuable market on the internet.

$42B SOM · CAPTURE SAM · $180B Answer-replaceable queries TAM · $680B Global search + productivity
TAM
$680B
Global search advertising + knowledge-work productivity software — the two pools any answer engine can pull revenue from.
SAM
$180B
The 40% of queries that are answer-replaceable — not navigation, not pure commerce. The queries worth a cited synthesized answer.
SOM · 5-year
$42B
Assumes 5% share of SAM in consumer + 8% share in enterprise knowledge work. Conservative vs. current growth curve.
perplexity
08 · Business Model
How we make money

Three revenue streams. All software margins.

Free
Full answer engine with daily Pro Search quota. Primary acquisition + awareness surface.
$0
Top of funnel
Pro
Unlimited Pro Searches, frontier models, Spaces, file context. Core consumer revenue.
$20 / mo
86% gross margin
Enterprise
SSO, internal sources, audit. Annual contracts to F500 + mid-market knowledge-heavy orgs.
$40–$80 / seat / mo
NRR 145%
Sonar API
Answer-engine-as-a-service. Developers build retrieval-backed AI products on our stack.
Usage-based
74% gross margin
145%
NRR — Enterprise cohort
$0.81
Gross margin per Pro query
7.8×
LTV / CAC — Pro
3.2 mo
Payback — Enterprise
perplexity
09 · Traction
What's working

Weekly query volume — and the curve is accelerating.

Weekly queries (millions)
last 16 weeks
W1W8W16
780M queries/mo
Run rate as of latest reporting period. Up from 15M/mo two years ago.
42% QoQ growth · retention curve flattening (good sign)
3.4M Pro subs
Up 3.8× YoY. Cohort retention stable at 72% month 12.
DAU/MAU 51% · top-decile for consumer AI
420 enterprise logos
Including 38 of the F100. Avg ACV $94K; largest $2.1M.
NRR 145% · land → expand median 4.2 months
perplexity
10 · Competition
Competition

Different goals, different primitives, different winners.

Capability
Perplexity
Google
ChatGPT
Bing/Copilot
You.com
Inline citations per claim
Answer-first UX
Real-time retrieval
Follow-up threading
Enterprise-ready (SSO, audit)
perplexity
11 · Moat
Why we win

The answer-engine stack compounds. We own every layer.

01
Base
Foundation models
We route across frontier providers (OpenAI, Anthropic, Meta). Commodity — but our routing layer is differentiated.
02
Retrieval
Real-time web index
Proprietary crawl, freshness-optimized ranking, cited-source scoring. Hard to replicate without web-scale infrastructure.
03
Synthesis
Answer-grade generation
Citation-grounded output, hallucination detection, trust scoring. Every product feature routes here.
04
Data flywheel
Query + feedback loop
780M queries/month feed ranking, hallucination detection, answer quality. Every competitor starts from zero.
05
Surface
Default answer engine
Becoming the reflexive place people ask questions. Once that habit locks in, the economics of the next decade flow through us.
perplexity
12 · Go-to-Market
How we grow

Four motions. Each with a clear signal.

Organic consumer
PLG

Free tier is the acquisition engine. Shareable answer threads carry the brand into every knowledge workflow. Word of mouth among students, analysts, writers.

63% organic
$6 blended CAC
Pro subscription
Self-serve

Free → Pro conversion via quota, model selection, Spaces. Credit-card-only, no sales. Runs on product signal (query depth, return rate).

9.1% free→pro
$20 ARPU
Enterprise sales
Direct

Named accounts in F500 research-heavy orgs (finance, law, consulting, gov). Pro usage signals drive outbound. Team-landing → expansion in quarters not years.

420 logos
145% NRR
Sonar API
Developer

Answer-engine-as-a-service for teams building AI products. Every customer built on Sonar is a distribution partner for the consumer brand.

12K+ API devs
74% GM
perplexity
13 · Team
Who we are

Research-depth founders with shipping experience.

AS

Aravind Srinivas

CEO · Co-founder

Former research scientist at OpenAI (GPT models) and DeepMind. PhD Berkeley. Published on retrieval-augmented generation — the technique Perplexity is built on.

DY

Denis Yarats

CTO · Co-founder

Former research scientist at Meta AI. PhD NYU under Yann LeCun. Core contributor to the retrieval-ranking stack that differentiates Perplexity's answer quality.

Team
[REPLACE: N people] — [REPLACE: breakdown, e.g., 60% research + eng, 20% product, 20% GTM]
Advisors
[REPLACE: Advisor names]
HQ
San Francisco · remote-flex
perplexity
14 · Ask
The ask

We're raising $[REPLACE: amount] to become the default answer engine on the internet.

$[REPLACE]

Led by [REPLACE: lead]. [REPLACE: round structure]. Existing investors committing to pro-rata.

40%
Inference & infrastructure — scale to 2B+ monthly queries
Capacity
30%
Product & research — ranking, synthesis, agentic answers
R&D
20%
Enterprise GTM — named accounts in research-heavy verticals
Commercial
10%
Consumer brand — make "just ask Perplexity" a default
Brand
01 / 14