CT · Ledger Room Insights
Blended — CT + Pipedrive
LONDON 14:32— queried accounts · clients · pipedrive.deals ·669ms·100 of 1,284 rows

1,284 clients arrived via Facebook in 2026 — the largest single acquisition source this year.

Read the headline as a sample. The figures below describe the 100-row sample shown, not all 1,284 rows — raise the cap or export for the full set before quoting totals.

Assumed Facebook = paid social (utm_medium) · since 1 Jan 2026 · revenue = realised · “signed up” = account created.

Clients (2026)1,284▲ 38% vs 2025
Realised rev£487k▲ £121k
Activated41%▼ 4pt
Avg FTT6d▲ 1.2d faster
Paid-social share100%of this cohort

FTT = first-trade time, the days from signup to first executed trade.

Merged result flagged 1,180 of 1,284 joined · 91.9% CT accounts · 1,284 Pipedrive deals · 1,180 104 unmatched · 0 null-key dropped — counts are per-source, not summed across sources.
Chart advisor: kept the deterministic bar selection — a monthly time series reads clearly here; a log axis was not needed.
Facebook · paid_social — new client signups monthly, 2026 (Jan–Jun)
Jan
142
Feb
168
Mar
205
Apr
231
May
264
Jun
274
Cumulative 1,284 clients · facebook / paid_social.
FTTfirst-trade time — days, signup → first executed trade paid_socialutm_medium value on the CT account
All rows share stream = Inbound utm_source = facebook utm_medium = paid_social
Sampled cohort — first 10 of 1,284 Facebook-sourced 2026 signups.
Client Type user_id Created Realised
    What this query does

    Counts CurrencyTransfer accounts created since the start of 2026 whose acquisition source is Facebook paid social, joins each to its client profile and any matching Pipedrive deal, and returns them newest-first.

    • Reads accounts, left-joins clients and pipedrive.deals.
    • Filters utm_source = 'facebook' and utm_medium = 'paid_social'.
    • Restricts to created_at >= 2026-01-01.
    • Orders by created_at DESC; display capped at 100 rows.
    SQL — generated from your question read-only · lore_ct_ro 
    1-- Blended: CT accounts + clients, joined to Pipedrive deals2SELECT a.user_id,3       c.display_name        AS client,4       a.client_type,5       'Inbound'             AS stream,6       a.utm_source, a.utm_medium7FROM   accounts a8LEFT JOIN clients c        ON c.account_id = a.id9LEFT JOIN pipedrive.deals d ON d.ct_user_id = a.user_id10WHERE  a.utm_source = 'facebook'11  AND  a.utm_medium = 'paid_social'12  AND  a.created_at >= DATE '2026-01-01'13ORDER BY a.created_at DESC;
    validated read-only · 0 writes · executed in 669ms
    utm_source breakdown
    facebook
    1,284
    google
    0
    organic
    0
    referral
    0
    What this query does

    Counts 2026 Facebook paid-social accounts, joined to clients and Pipedrive deals, newest-first.

    • Filters utm_source = 'facebook' + utm_medium = 'paid_social'.
    • Restricts to created_at >= 2026-01-01.
    SQL — generated from your question read-only · lore_ct_ro 
    1-- Blended: CT accounts + clients, joined to Pipedrive deals2SELECT a.user_id, c.display_name AS client,3       a.client_type, a.utm_source, a.utm_medium4FROM   accounts a5LEFT JOIN clients c ON c.account_id = a.id6WHERE  a.utm_source = 'facebook' AND a.utm_medium = 'paid_social'7  AND  a.created_at >= DATE '2026-01-01'8ORDER BY a.created_at DESC;
    validated read-only · 0 writes · executed in 669ms
    Lore made these reads to answer your question.read-only
    1
    “Facebook” read as paid social
    utm_source = 'facebook' AND utm_medium = 'paid_social'
    2
    Timeframe: since 1 Jan 2026
    accounts.created_at >= DATE '2026-01-01'
    3
    Revenue: realised
    sum over trade_bookings where status = 'settled'
    4
    “Signed up” = account created
    accounts.created_at — not first deposit or first trade
    Analyst notes
    41% activation lags the book. Volume is up 38% but activation slipped 4pt; a wider top is diluting quality.
    FTT 6d — fast. First trade lands 1.2d quicker than 2025; the funnel converts, the gap is downstream.
    ~12% identity gaps. Rows resolve to client_type = Client with no name — a Pipedrive deal exists but the CT profile is still thin.
    Next
    The Ledger Room · CT

    Ask a question. Get a finding, not a table.

    Plain English in, a verdict out — with the evidence, SQL and confidence one tap away. Start from a department below, or just type in the desk.

    Acquisition
    Revenue
    Affiliates
    Trading
    Clarify before run
    Pin down the source-tagged terms before Lore queries.
    Reading FacebookCT? 2026CT✓ proportionCT? first-month tradePD✓
    CT sourceacquisition channel1 / 2
    Facebook Ads (paid) or Facebook.com referral (organic)?
    CT sourcedenominator scope2 / 2
    A “% of the month” of what — the cohort’s own monthly total, or the whole book’s?
    2 to answer · resolved terms carry over from earlier turns.
    Low confidence — confirm before I run this low
    • “first month” is ambiguous — measured from account_created, not first deposit.
    • Proportion denominator resolved to within (cohort’s own monthly total).
    Show generated SQL 
    1SELECT date_trunc('month', a.created_at) m,2       count(DISTINCT t.account_id)::numeric3       / count(DISTINCT a.id) AS first_month_rate4FROM accounts a LEFT JOIN trade_bookings t …
    LONDON 14:32— querying accounts · clients·0ms
    The verdict lands first, then the metric strip and chart cascade.
    read-only verified Read-only · verified. Every query runs as lore_ct_ro — SELECT only. Writes are rejected at the connection, not just the UI. Enter to ask · Shift+Enter for a new line