Recipe: Channel analysis

Run a complete channel ROI report by composing the two read-only channel analyses: per-event attribution (who joined the channel because of which event) and audience overlap (how much each pair of channels shares). The recipe wires analysis channel attribution and analysis channel audience together with jq for downstream charting.

Prerequisites: authenticated CLI and at least two channels with event registrations. Audience overlap requires between 2 and 5 channel IDs.

Steps

Discover the channels you want to analyze

talkvalue path channel list --json \
  | jq '.data[] | {id, name, personCount}'

Returns every channel with its current member count. Pick two to five IDs for the rest of the script.

Discover the events for attribution scope

talkvalue path event list --json \
  | jq '.data[] | {id, name, startAt}'

Use this when you want to pin attribution to a specific event subset rather than every event the channel touched.

Per-channel attribution across selected events

CHANNEL_ID=7
talkvalue path analysis channel attribution "$CHANNEL_ID" \
  --event-id 18 --event-id 22 \
  --json \
  | jq '{
      channel: .data.channel.name,
      channelSize: .data.metrics.channelSize,
      eventParticipationRate: .data.metrics.eventParticipationRate,
      events: (.data.events | map({name, total, joinedSinceLastEvent, acquisitionRate}))
    }'

Returns the channel’s headline metrics (channelSize, membersEverRegistered, eventParticipationRate) plus a per-event breakdown showing how many new members the channel acquired from each event (joinedSinceLastEvent, with acquisitionRate) versus people already on the channel (alreadyInChannel).Omit --event-id entirely to roll up every event the channel ever appeared on.

Audience overlap across the channel group

talkvalue path analysis channel audience \
  --channel-id 7 --channel-id 12 --channel-id 19 \
  --json \
  | jq '{
      channels: (.data.channels | map({name, personCount})),
      intersections: .data.intersections,
      multiChannelRate: .data.metrics.multiChannelRate
    }'

Returns each channel’s size, every n-way intersection (pairs, triples, and so on), and the aggregate multiChannelRate: the fraction of unique people who sit in more than one channel.

Combine into a single report

CHANNEL_IDS=(7 12 19)
OUT="reports/channel-analysis-$(date +%Y-%m-%d).json"
mkdir -p reports

OVERLAP=$(
  talkvalue path analysis channel audience \
    $(printf -- '--channel-id %s ' "${CHANNEL_IDS[@]}") \
    --json
)

ATTRIBUTION="["
for id in "${CHANNEL_IDS[@]}"; do
  row=$(talkvalue path analysis channel attribution "$id" --json | jq '.data')
  ATTRIBUTION+="$row,"
done
ATTRIBUTION="${ATTRIBUTION%,}]"

jq -n --argjson overlap "$OVERLAP" --argjson attribution "$ATTRIBUTION" \
  '{generatedAt: now | todate, overlap: $overlap.data, attribution: $attribution}' \
  > "$OUT"
echo "Wrote $OUT"

Iterates the channel list, captures each channel’s full attribution snapshot, runs one overlap call, and writes a single timestamped JSON file. Drop the loop body into a cron job for a daily archive.

Variants

Scope every event analysis to a single tag

talkvalue path analysis channel attribution 7 --tag-id 4 --json \
  | jq '.data.events | map({name, total, joinedSinceLastEvent, acquisitionRate})'

--tag-id limits the attribution to events that carry the given tag. Useful for “webinars only” or “Q2 campaign only” reports. See Tags and tag attach for tagging events.

Spot the largest audience overlap in a trio

talkvalue path analysis channel audience \
  --channel-id 7 --channel-id 12 --channel-id 19 --json \
  | jq '.data.intersections | sort_by(-.count) | .[0]'

Returns the single most overlapped n-way slice across the trio. The full intersection list includes every 2-way pair plus the 3-way intersection.

Track multi-channel rate over time

talkvalue path analysis channel audience \
  --channel-id 7 --channel-id 12 --channel-id 19 --channel-id 25 --json \
  | jq -r '[.data.metrics.multiChannelRate, (now | strftime("%Y-%m-%d"))] | @csv' \
  >> reports/multi-channel-rate.csv

Appends a daily row to a long-form CSV. Useful for monitoring how concentrated your audience is becoming over time.

Tips

acquisitionRate on an event is joinedSinceLastEvent / total: the share of the event’s audience that the channel acquired specifically because of that event.
eventParticipationRate on a channel is membersEverRegistered / channelSize: the share of channel members who registered for at least one event.
Audience overlap requires 2 ≤ channelCount ≤ 5. Outside that range, the command exits with a usage error (exit 2).
For the dashboard view of the same data, see Channel attribution and Audience overlap.

Get Started

Global

Commands

Recipes

AI Agents

Reference

Steps

Variants

Scope every event analysis to a single tag

Spot the largest audience overlap in a trio

Track multi-channel rate over time

Tips

See also

​Steps

​Variants

​Scope every event analysis to a single tag

​Spot the largest audience overlap in a trio

​Track multi-channel rate over time

​Tips

​See also

Steps

Variants

Scope every event analysis to a single tag

Spot the largest audience overlap in a trio

Track multi-channel rate over time

Tips

See also