Analyse a funding market

A market analysis in five requests. This example measures venture funding into European AI companies: how much capital the market absorbs each year, whether deal count is rising or falling with it, what a typical round now looks like, and who is deploying. Two taxonomy lookups resolve the market, then three aggregates measure it. Together they answer a question that no single endpoint answers on its own, which is the point: the analytics endpoints are composable, so you assemble a view rather than fetching a prebuilt one.

To analyse a different market, change the two lookups in step 2. Every later request reuses the IDs they return, so nothing else needs editing. Pass type on both lookups: an untyped search matches every taxonomy row containing your term, which is slower and returns rows you did not ask for.

Prerequisites

Complete the Quickstart to create an API key and make your first request.
DEALROOM_CLIENT_ID and DEALROOM_CLIENT_SECRET exported as environment variables.
jq for the cURL and Command Prompt snippets. The PowerShell, Node.js, and Python variants don’t require it.

The recipe

# 1. Exchange credentials for a Bearer token
ACCESS_TOKEN=$(curl -s -X POST "https://accounts.beta.dealroom.co/oauth/token" \
  -H "Content-Type: application/json" \
  -d "{
    \"client_id\": \"$DEALROOM_CLIENT_ID\",
    \"client_secret\": \"$DEALROOM_CLIENT_SECRET\",
    \"audience\": \"https://api-next.beta.dealroom.co\",
    \"grant_type\": \"client_credentials\"
  }" | jq -r '.access_token')

dr() {
  curl -s -G "https://api.beta.dealroom.app$1" \
    -H "Authorization: Bearer $ACCESS_TOKEN" \
    -H "X-Client-Id: $DEALROOM_CLIENT_ID" "${@:2}"
}

# 2. Resolve the market definition to IDs. Never hardcode taxonomy IDs.
#    A lookup can return several matches, so select by exact name rather than
#    taking the first row, and stop if there is no match.
TAG_ID=$(dr /reference/filters/tag_id/values \
  --data-urlencode "q=artificial intelligence" \
  --data-urlencode "type=technology" \
  | jq -r '.data[] | select((.name | ascii_downcase) == "artificial intelligence") | .id' \
  | head -n 1)
test -n "$TAG_ID" || { echo "Artificial Intelligence tag not found" >&2; exit 1; }

GEO_ID=$(dr /reference/filters/location/values \
  --data-urlencode "q=Europe" \
  --data-urlencode "type=continent" \
  | jq -r '.data[] | select((.name | ascii_downcase) == "europe") | .id' \
  | head -n 1)
test -n "$GEO_ID" || { echo "Europe continent not found" >&2; exit 1; }

MARKET="tag_id[in_any]:$TAG_ID,hq_location[eq]:$GEO_ID,is_vc_round[eq]:true"

# 3. Size and momentum: capital raised and deal count per year
echo "Year  Rounds  Raised"
dr /analytics/aggregate/funding-rounds \
  --data-urlencode "metric=count,sum:amount" \
  --data-urlencode "group_by=year" \
  --data-urlencode "filter=and($MARKET,year[gte]:2021)" \
  --data-urlencode "sort=dimension" \
  --data-urlencode "currency=USD" \
  | jq -r '.data[] | "\(.dimension)  \(.count)  $\(.sum_amount / 1000000000 * 10 | round / 10)B"'

# 4. Deal shape: what a typical round looks like.
#    amount[gte]:1 drops rounds with no disclosed amount.
echo "Year  Rounds  p25  Median  p75"
dr /analytics/aggregate/funding-rounds \
  --data-urlencode "metric=count,p25:amount,median:amount,p75:amount" \
  --data-urlencode "group_by=year" \
  --data-urlencode "filter=and($MARKET,amount[gte]:1,year[gte]:2021)" \
  --data-urlencode "sort=dimension" \
  --data-urlencode "currency=USD" \
  | jq -r '.data[] | "\(.dimension)  \(.count)  $\(.p25_amount / 1000000 * 10 | round / 10)M  $\(.median_amount / 1000000 * 10 | round / 10)M  $\(.p75_amount / 1000000 * 10 | round / 10)M"'

# 5. Who is deploying into this market
echo "Investor  Rounds  Participated in"
dr /analytics/aggregate/funding-rounds \
  --data-urlencode "metric=count,sum:amount" \
  --data-urlencode "group_by=investor_name" \
  --data-urlencode "filter=and($MARKET,year[gte]:2024)" \
  --data-urlencode "sort=-count" \
  --data-urlencode "limit=10" \
  --data-urlencode "currency=USD" \
  | jq -r '.data[] | "\(.dimension)  \(.count)  $\(.sum_amount / 1000000 | round)M"'

REM 1. Exchange credentials for a Bearer token
curl -s -X POST "https://accounts.beta.dealroom.co/oauth/token" ^
  -H "Content-Type: application/json" ^
  -d "{\"client_id\": \"%DEALROOM_CLIENT_ID%\", \"client_secret\": \"%DEALROOM_CLIENT_SECRET%\", \"audience\": \"https://api-next.beta.dealroom.co\", \"grant_type\": \"client_credentials\"}" > token.json
for /f "delims=" %A in ('jq -r .access_token token.json') do set ACCESS_TOKEN=%A
del token.json

REM 2. Resolve the market definition to IDs. Never hardcode taxonomy IDs.
REM    A lookup can return several matches, so select by exact name rather than
REM    taking the first row, and stop if there is no match.
curl -s -G "https://api.beta.dealroom.app/reference/filters/tag_id/values" ^
  -H "Authorization: Bearer %ACCESS_TOKEN%" ^
  -H "X-Client-Id: %DEALROOM_CLIENT_ID%" ^
  --data-urlencode "q=artificial intelligence" ^
  --data-urlencode "type=technology" > tags.json

set "TAG_ID="
for /f "delims=" %A in ('jq -r "[.data[] | select((.name | ascii_downcase) == \"artificial intelligence\") | .id][0] // empty" tags.json') do set TAG_ID=%A
del tags.json
if not defined TAG_ID (
  echo Artificial Intelligence tag not found
  exit /b 1
)

curl -s -G "https://api.beta.dealroom.app/reference/filters/location/values" ^
  -H "Authorization: Bearer %ACCESS_TOKEN%" ^
  -H "X-Client-Id: %DEALROOM_CLIENT_ID%" ^
  --data-urlencode "q=Europe" ^
  --data-urlencode "type=continent" > locations.json

set "GEO_ID="
for /f "delims=" %A in ('jq -r "[.data[] | select((.name | ascii_downcase) == \"europe\") | .id][0] // empty" locations.json') do set GEO_ID=%A
del locations.json
if not defined GEO_ID (
  echo Europe continent not found
  exit /b 1
)

set "MARKET=tag_id[in_any]:%TAG_ID%,hq_location[eq]:%GEO_ID%,is_vc_round[eq]:true"

REM 3. Size and momentum: capital raised and deal count per year
echo Year  Rounds  Raised
curl -s -G "https://api.beta.dealroom.app/analytics/aggregate/funding-rounds" ^
  -H "Authorization: Bearer %ACCESS_TOKEN%" ^
  -H "X-Client-Id: %DEALROOM_CLIENT_ID%" ^
  --data-urlencode "metric=count,sum:amount" ^
  --data-urlencode "group_by=year" ^
  --data-urlencode "filter=and(%MARKET%,year[gte]:2021)" ^
  --data-urlencode "sort=dimension" ^
  --data-urlencode "currency=USD" ^
  | jq -r ".data[] | \"\(.dimension)  \(.count)  $\(.sum_amount / 1000000000 * 10 | round / 10)B\""

REM 4. Deal shape: what a typical round looks like.
REM    amount[gte]:1 drops rounds with no disclosed amount.
echo Year  Rounds  p25  Median  p75
curl -s -G "https://api.beta.dealroom.app/analytics/aggregate/funding-rounds" ^
  -H "Authorization: Bearer %ACCESS_TOKEN%" ^
  -H "X-Client-Id: %DEALROOM_CLIENT_ID%" ^
  --data-urlencode "metric=count,p25:amount,median:amount,p75:amount" ^
  --data-urlencode "group_by=year" ^
  --data-urlencode "filter=and(%MARKET%,amount[gte]:1,year[gte]:2021)" ^
  --data-urlencode "sort=dimension" ^
  --data-urlencode "currency=USD" ^
  | jq -r ".data[] | \"\(.dimension)  \(.count)  $\(.p25_amount / 1000000 * 10 | round / 10)M  $\(.median_amount / 1000000 * 10 | round / 10)M  $\(.p75_amount / 1000000 * 10 | round / 10)M\""

REM 5. Who is deploying into this market
echo Investor  Rounds  Participated in
curl -s -G "https://api.beta.dealroom.app/analytics/aggregate/funding-rounds" ^
  -H "Authorization: Bearer %ACCESS_TOKEN%" ^
  -H "X-Client-Id: %DEALROOM_CLIENT_ID%" ^
  --data-urlencode "metric=count,sum:amount" ^
  --data-urlencode "group_by=investor_name" ^
  --data-urlencode "filter=and(%MARKET%,year[gte]:2024)" ^
  --data-urlencode "sort=-count" ^
  --data-urlencode "limit=10" ^
  --data-urlencode "currency=USD" ^
  | jq -r ".data[] | \"\(.dimension)  \(.count)  $\(.sum_amount / 1000000 | round)M\""

# 1. Exchange credentials for a Bearer token
$body = @{
  client_id     = $env:DEALROOM_CLIENT_ID
  client_secret = $env:DEALROOM_CLIENT_SECRET
  audience      = "https://api-next.beta.dealroom.co"
  grant_type    = "client_credentials"
} | ConvertTo-Json

$accessToken = (Invoke-RestMethod -Method Post `
  -Uri "https://accounts.beta.dealroom.co/oauth/token" `
  -ContentType "application/json" `
  -Body $body).access_token

$headers = @{
  Authorization = "Bearer $accessToken"
  "X-Client-Id" = $env:DEALROOM_CLIENT_ID
}

function Get-Dealroom($Path, $Query) {
  Invoke-RestMethod -Method Get -Uri "https://api.beta.dealroom.app$Path" -Headers $headers -Body $Query
}

# 2. Resolve the market definition to IDs. Never hardcode taxonomy IDs.
#    A lookup can return several matches, so select by exact name rather than
#    taking the first row, and stop if there is no match.
$tags = Get-Dealroom "/reference/filters/tag_id/values" @{
  q    = "artificial intelligence"
  type = "technology"
}
$tag = $tags.data | Where-Object { $_.name -ieq "artificial intelligence" } | Select-Object -First 1
if (-not $tag) { throw "Artificial Intelligence tag not found" }

$locations = Get-Dealroom "/reference/filters/location/values" @{
  q    = "Europe"
  type = "continent"
}
$geo = $locations.data | Where-Object { $_.name -ieq "europe" } | Select-Object -First 1
if (-not $geo) { throw "Europe continent not found" }

$market = "tag_id[in_any]:$($tag.id),hq_location[eq]:$($geo.id),is_vc_round[eq]:true"

# 3. Size and momentum: capital raised and deal count per year
$momentum = Get-Dealroom "/analytics/aggregate/funding-rounds" @{
  metric   = "count,sum:amount"
  group_by = "year"
  filter   = "and($market,year[gte]:2021)"
  sort     = "dimension"
  currency = "USD"
}
"Year  Rounds  Raised"
foreach ($row in $momentum.data) {
  '{0}  {1}  ${2:N1}B' -f $row.dimension, $row.count, ($row.sum_amount / 1e9)
}

# 4. Deal shape: what a typical round looks like.
#    amount[gte]:1 drops rounds with no disclosed amount.
$shape = Get-Dealroom "/analytics/aggregate/funding-rounds" @{
  metric   = "count,p25:amount,median:amount,p75:amount"
  group_by = "year"
  filter   = "and($market,amount[gte]:1,year[gte]:2021)"
  sort     = "dimension"
  currency = "USD"
}
"Year  Rounds  p25  Median  p75"
foreach ($row in $shape.data) {
  '{0}  {1}  ${2:N1}M  ${3:N1}M  ${4:N1}M' -f $row.dimension, $row.count,
    ($row.p25_amount / 1e6), ($row.median_amount / 1e6), ($row.p75_amount / 1e6)
}

# 5. Who is deploying into this market
$investors = Get-Dealroom "/analytics/aggregate/funding-rounds" @{
  metric   = "count,sum:amount"
  group_by = "investor_name"
  filter   = "and($market,year[gte]:2024)"
  sort     = "-count"
  limit    = 10
  currency = "USD"
}
"Investor  Rounds  Participated in"
foreach ($row in $investors.data) {
  '{0}  {1}  ${2:N0}M' -f $row.dimension, $row.count, ($row.sum_amount / 1e6)
}

import { ClientCredentials } from "simple-oauth2";

const CLIENT_ID = process.env.DEALROOM_CLIENT_ID!;
const CLIENT_SECRET = process.env.DEALROOM_CLIENT_SECRET!;

const oauth = new ClientCredentials({
  client: { id: CLIENT_ID, secret: CLIENT_SECRET },
  auth: { tokenHost: "https://accounts.beta.dealroom.co", tokenPath: "/oauth/token" },
});
const token = await oauth.getToken({ audience: "https://api-next.beta.dealroom.co" });

const headers = {
  Authorization: `Bearer ${token.token.access_token}`,
  "X-Client-Id": CLIENT_ID,
};

async function get(path: string, params: Record<string, string>) {
  const response = await fetch(
    `https://api.beta.dealroom.app${path}?${new URLSearchParams(params)}`,
    { headers },
  );
  if (!response.ok) throw new Error(`${response.status} on ${path}`);
  return await response.json();
}

// A lookup can return several matches, so select by exact name rather than
// taking the first row, and stop if there is no match.
function pickByName(rows: Array<{ id: number; name: string }>, name: string) {
  const match = rows.find(row => row.name.toLowerCase() === name.toLowerCase());
  if (!match) throw new Error(`No exact match for "${name}"`);
  return match;
}

// Resolve the market definition to IDs. Never hardcode taxonomy IDs.
const tags = await get("/reference/filters/tag_id/values", {
  q: "artificial intelligence",
  type: "technology",
});
const locations = await get("/reference/filters/location/values", {
  q: "Europe",
  type: "continent",
});

const tag = pickByName(tags.data, "artificial intelligence");
const location = pickByName(locations.data, "Europe");

const market = [
  `tag_id[in_any]:${tag.id}`,
  `hq_location[eq]:${location.id}`,
  "is_vc_round[eq]:true",
].join(",");

const billions = (n: number) => `$${(n / 1e9).toFixed(1)}B`;
const millions = (n: number) => `$${(n / 1e6).toFixed(1)}M`;

// Size and momentum: capital raised and deal count per year
const momentum = await get("/analytics/aggregate/funding-rounds", {
  metric: "count,sum:amount",
  group_by: "year",
  filter: `and(${market},year[gte]:2021)`,
  sort: "dimension",
  currency: "USD",
});
console.log("\nYear  Rounds  Raised");
for (const row of momentum.data) {
  console.log(`${row.dimension}  ${row.count}  ${billions(row.sum_amount)}`);
}

// Deal shape: what a typical round looks like.
// amount[gte]:1 drops rounds with no disclosed amount.
const shape = await get("/analytics/aggregate/funding-rounds", {
  metric: "count,p25:amount,median:amount,p75:amount",
  group_by: "year",
  filter: `and(${market},amount[gte]:1,year[gte]:2021)`,
  sort: "dimension",
  currency: "USD",
});
console.log("\nYear  Rounds  p25  Median  p75");
for (const row of shape.data) {
  console.log(
    `${row.dimension}  ${row.count}  ${millions(row.p25_amount)}  ` +
      `${millions(row.median_amount)}  ${millions(row.p75_amount)}`,
  );
}

// Who is deploying into this market
const investors = await get("/analytics/aggregate/funding-rounds", {
  metric: "count,sum:amount",
  group_by: "investor_name",
  filter: `and(${market},year[gte]:2024)`,
  sort: "-count",
  limit: "10",
  currency: "USD",
});
console.log("\nInvestor  Rounds  Participated in");
for (const row of investors.data) {
  console.log(`${row.dimension}  ${row.count}  ${millions(row.sum_amount)}`);
}

import os
from authlib.integrations.requests_client import OAuth2Session

CLIENT_ID = os.environ["DEALROOM_CLIENT_ID"]
CLIENT_SECRET = os.environ["DEALROOM_CLIENT_SECRET"]

session = OAuth2Session(client_id=CLIENT_ID, client_secret=CLIENT_SECRET)
session.headers.update({"X-Client-Id": CLIENT_ID})
session.fetch_token(
    url="https://accounts.beta.dealroom.co/oauth/token",
    grant_type="client_credentials",
    audience="https://api-next.beta.dealroom.co",
)

BASE = "https://api.beta.dealroom.app"
AGGREGATE = f"{BASE}/analytics/aggregate/funding-rounds"


def get(path, **params):
    response = session.get(path, params=params)
    response.raise_for_status()
    return response.json()


# A lookup can return several matches, so select by exact name rather than
# taking the first row, and stop if there is no match.
def pick_by_name(rows, name):
    for row in rows:
        if row["name"].lower() == name.lower():
            return row
    raise LookupError(f'No exact match for "{name}"')


# Resolve the market definition to IDs. Never hardcode taxonomy IDs.
tags = get(
    f"{BASE}/reference/filters/tag_id/values",
    q="artificial intelligence",
    type="technology",
)
locations = get(
    f"{BASE}/reference/filters/location/values",
    q="Europe",
    type="continent",
)

tag = pick_by_name(tags["data"], "artificial intelligence")
location = pick_by_name(locations["data"], "Europe")

market = ",".join([
    f"tag_id[in_any]:{tag['id']}",
    f"hq_location[eq]:{location['id']}",
    "is_vc_round[eq]:true",
])

# Size and momentum: capital raised and deal count per year
momentum = get(
    AGGREGATE,
    metric="count,sum:amount",
    group_by="year",
    filter=f"and({market},year[gte]:2021)",
    sort="dimension",
    currency="USD",
)
print("\nYear  Rounds  Raised")
for row in momentum["data"]:
    print(f"{row['dimension']}  {row['count']}  ${row['sum_amount'] / 1e9:.1f}B")

# Deal shape: what a typical round looks like.
# amount[gte]:1 drops rounds with no disclosed amount.
shape = get(
    AGGREGATE,
    metric="count,p25:amount,median:amount,p75:amount",
    group_by="year",
    filter=f"and({market},amount[gte]:1,year[gte]:2021)",
    sort="dimension",
    currency="USD",
)
print("\nYear  Rounds  p25  Median  p75")
for row in shape["data"]:
    print(
        f"{row['dimension']}  {row['count']}  ${row['p25_amount'] / 1e6:.1f}M  "
        f"${row['median_amount'] / 1e6:.1f}M  ${row['p75_amount'] / 1e6:.1f}M"
    )

# Who is deploying into this market
investors = get(
    AGGREGATE,
    metric="count,sum:amount",
    group_by="investor_name",
    filter=f"and({market},year[gte]:2024)",
    sort="-count",
    limit=10,
    currency="USD",
)
print("\nInvestor  Rounds  Participated in")
for row in investors["data"]:
    print(f"{row['dimension']}  {row['count']}  ${row['sum_amount'] / 1e6:.0f}M")

How the requests fit together

Every request after the lookups reuses one filter fragment, MARKET, which is the market definition: a technology tag, a location, and is_vc_round[eq]:true to exclude grants, debt, and other non-venture events. Holding that fragment constant is what makes the results comparable.

Step	What it measures	Key parameters
2	The market definition, as taxonomy IDs	`q`, `type`
3	Capital and deal count over time	`metric=count,sum:amount`, `group_by=year`
4	The distribution of round sizes over time	`metric=count,p25:amount,median:amount,p75:amount`
5	The most active investors in the market	`metric=count,sum:amount`, `group_by=investor_name`

Steps 3 and 4 are the pair that carries the analysis. Step 3 tells you whether money is flowing in. Step 4 tells you whether it is spread across more companies or concentrated into larger rounds. Reading them together is what separates a growing market from a concentrating one.

Common pitfalls

Reporting lag. Rounds keep being recorded for weeks after they close, so the most recent period always understates. Exclude the current year from a trend, or label it as partial.
Undisclosed amounts. amount[gte]:1 in step 4 excludes rounds with no reported amount, which is why its count is lower than step 3’s. Leave the filter out and those rounds enter the distribution as zeroes.
Investor amounts double count. A round contributes its full size to every investor in it, so step 5’s amounts rank investors but should not be totalled. Deal count also favours investors who write many small cheques, so use sort=-sum_amount when you want capital deployed instead.
Small groups. Narrow the market far enough and a percentile is drawn from a handful of rounds. Check count before reading a median.

Where to go from here

Break the same market down by geography instead of time with group_by=hq_country, or by stage with group_by=standardized_round. The filter fragment stays identical. See Analytics and aggregates for every supported source, metric, and dimension.
Add per-metric filters and percentages with /analytics/aggregate/{source}/multi-metric, which can calculate shares such as “what proportion of this market is unicorns” in one request.
Request every monetary metric and threshold in another currency with currency=EUR. See Currencies.
Move from the aggregate to the underlying records with /data/transactions, using the same filter syntax. See Filtering and the Filters and sorting reference.
For a shorter first query against a /data endpoint, see Top fintech startups.

Getting Started

Examples

Agents

Concepts

References

Analyse a funding market

Prerequisites

The recipe

How the requests fit together

Common pitfalls

Where to go from here

​Prerequisites

​The recipe

​How the requests fit together

​Common pitfalls

​Where to go from here

Prerequisites

The recipe

How the requests fit together

Common pitfalls

Where to go from here