For DTC brands £500k–£5m

Every part of your business.
Every week. With a £ tag.

Operator Intelligence is the accountability layer your DTC brand has been missing. Not another attribution debate — synthesis across paid, site, email, creative, inventory and affiliate, with the £ at stake on every recommendation.

Built on real DTC data, not slideware · 54.5/62 on combined platform + edge-case eval · From £99/month
operatorintelligence.com/workspace/frkl
Oi
Operator Intelligence
frkl workspace
Home Intelligence Channels Commerce Audience
Monday · 27 May · operator weekly

£30.7k/mo in play this week

Concentrated in active leakage to plug. The three highest-impact actions are queued below.

£20.6k
Leaking now
£6.9k
At risk
£3.2k
Opportunity
£30.7k
Total identified
What changed since last run
5
new findings
1
strengthened
0
resolved
Do this week
1
Fix cart-checkout JS error preventing ~18% of mobile sessions from completing
P1 Pulse £9.5k/mo
2
Tighten FIRSTCHARM code — driving 33% of discount load on 12% of orders
P1 Atlas £4.3k/mo
3
Refresh DPA creative pool — 35% of spend on Below-avg conv-rank ads
P2 Frame £2.2k/mo
£30.7k
/mo identified on £16k/mo brand
189%
of monthly revenue in play
8
business functions, one screen
7
algorithm engines running per brand
~10 min
from connection to first insight
The wedge

Your dashboard tells you what happened.
It doesn't tell you what to do.

Triple Whale, Northbeam, Polar — every "all-in-one" platform competes on attribution accuracy. That's the wrong job.

What attribution platforms answer

Was my Meta ROAS 2.8× or 3.4× last week? Which channel touched the order? Should the credit go to the prospecting campaign or the retargeting one?

Useful for a media buyer on a Wednesday. Not what an operator opens on Monday morning.

What operators actually ask

I have 18 tabs open — which 3 things matter this week? What's leaking money right now and I don't know it? Why is my MER trending down and what do I do about it? I'm running out of stock on bestsellers while £290k sits in dead inventory.

No attribution platform answers any of these.

"Triple Whale tells you where your last sale came from.
We tell you why next month's looks shaky."

— Operator Intelligence positioning

How we win

Two pillars Triple Whale can't bolt on

Neither dimension can be added later. They have to be baked in from day one — which is why we are.

Pillar 01

Holistic coverage

Eight business functions, one screen, one weekly action list. Every recommendation comes with a £/month figure attached.

Paid intelligence (Meta + Google, creative-level)
Site + CRO (GA4 + Clarity, £-quantified friction)
Commercial (SKU economics, bundles, inventory cover)
Creative review (vision-AI scoring per asset)
Email + CRM (true Klaviyo attribution, flow gap audit)
Affiliate + creator (code tracking + discovery)
Competitive (live monitoring)
Cross-channel synthesis (the thing nobody does properly)
Pillar 02

Intellectual honesty

Most analytics tools look impressive in a demo and fail you when the data gets messy. Ours does the opposite — it tells you when your dashboard is wrong.

Refuses to forecast without your planning inputs
Excludes BFCM + brand-sale days from baselines
Brand-age banner: 'directional only' below 60 days
Connection-health strip on every tab
Divergence engine flags decoupling pairs automatically
Action conflict surfacing (scale vs tighten = pick one)
P(hit) calibration with sample-size badges
Diagnostic recipes that explain why a finding fired

The layer Triple Whale will never build

It would force them to admit how often their own numbers are wrong.

Failure mode in other tools
What Operator Intelligence does instead
Silently builds trend lines on 14 days of data
Brand-Age Banner — "<60d of history — directional only, forecasts not meaningful." Engines refuse to fire below min-sample thresholds.
Treats Black Friday data as baseline for 23 days afterwards
Seasonal Events table — BFCM, Christmas, Mother's Day, brand sales explicitly excluded from baselines. Affected findings annotated.
Shows broken connections as smooth metric decay
Connection Health Strip on every tab — colour-coded per source, explicit "X days stale" prompt when critical.
Reports "ROAS up" without noticing "revenue down" alongside
Divergence engine monitors 11 operator-critical metric pairs. Auto-flags decoupling and sign-flips.
Lets you set "scale spend" AND "tighten ROAS" without flagging the conflict
Action Conflict Banner detects mutually-opposing open actions and forces an explicit pick.
Reports gross Klaviyo orders as "email revenue"
Pulls Klaviyo AttributedFlow / AttributedCampaign reports and shows true-attributed vs gross-tracked side by side.
Proof

£30.7k/mo found in one weekend

Real receipts on a real brand. frkl is a £16k/mo DTC jewellery brand. Operator Intelligence was deployed on a Friday — by Monday the team had a new action list.

Design partner frkl · DTC jewellery · £16k/mo

184% of monthly revenue in play — and three of four "headline crises" turned out to be data artefacts.

£30.7k
/mo of identified £-at-stake
£295k
working capital locked in overstock
12% → 36%
discount rate, corrected from "crisis" to healthy
~10 min
from connection to first £-quantified insight
!
The discount-rate correction. Every other tool reported 36–47% discount load. Real DTC rate was 12% — the rest was the gifting channel showing up as 100%-discount orders. Two of frkl's "headline crises" disappeared.
£
The inventory finding. 4 SKUs critically low (£2.5k/mo of lost sales until reorders land) while £290k of working capital sits in 82 overstocked SKUs. 18 months of DTC revenue tied up in stock that doesn't sell.
The seasonal correction. When the seasonal-awareness module switched on, a trend stopped firing because its baseline had been contaminated by brand sale days. For Q4 brands, this is the difference between credible and confidently-wrong.
The creative vision finding. Claude vision scored 7 ad creatives. The 'how your email finds me' meme reel scored 4.8/5 — gold-standard formula to replicate. The catalogue DPA scored 1.8/5 — restrict to retargeting only. The layer Triple Whale categorically doesn't have.
What it refused to do. Divergence engine returned 0 findings — channels are tightly coupled, no decoupling to surface. Forecast tab refuses without explicit spend plans. It told the truth, including the truth that it didn't know.

Read the full case study including the Klaviyo attribution correction, the affiliate concentration finding, and the five auto-diagnoses surfaced by the recipe engine. Read the full case study →

Eval scorecard

Tested on two axes most analytics tools fail one of

A platform that scores 28/35 on main capability but 9/27 on edge cases is more dangerous than one that scores 20 across the board — because it performs in demos and fails when it matters.

Main platform capability
28 / 35
"Genuinely valuable team member"
Edge-case stress test
26.5 / 27
"Intellectually honest. Safe to trust."
Combined
54.5 / 62
Trustworthy team member, not a demo-impressive dashboard
Pricing

Honest pricing. Honest product.

Triple Whale starts at £450/mo for attribution alone. We start at £99/mo for everything else that actually matters.

Starter
£99/month
Pre-£500k DTC, single operator
Full dashboard, all 8 functions
Monday brief delivered to inbox
Trend + anomaly engines
Single brand workspace
Ask Data (100 questions/month)
Start free trial
Scale
£599/month
£3m+ DTC
Everything in Growth
Live competitive monitoring
Affiliate discovery (Heepsy)
Slack delivery of weekly brief
Custom diagnostic recipes
Monthly office hour
Start free trial
Operator (Managed)
£1,750+/month
Brand wants the dashboard + a human on top
Everything in Scale
Managed onboarding
Weekly human review call
Custom agent calibration
Strategic call cadence
Book a call

All plans include a 7-day free trial. No credit card required to start. Cancel any time.

FAQ

The questions a sceptical DTC founder asks

How is this different from Triple Whale?

Triple Whale answers "where did my last sale come from?" — an attribution question. We answer "what's actually true about my business, and what should I do this week?" — a synthesis question. Triple Whale charges £450+/mo for the attribution debate. We charge £99 for everything else: site CVR diagnostics, inventory cover, creative vision scoring, true email attribution, cross-channel divergence detection, and a £-tagged action board. The two tools don't compete on the same axis.

What does "intellectually honest" actually mean in practice?

Six concrete things you can verify in the product: (1) trend engines refuse to fire below 14 days of data and the dashboard tells you why; (2) sale-day data points are excluded from baselines so BFCM doesn't contaminate January; (3) every tab shows when its data sources last refreshed; (4) the forecast tab is empty until you supply planning inputs — we won't extrapolate from a contaminated baseline; (5) P(hit) badges show "low confidence (n=1)" rather than a precise number until enough actions have closed; (6) when correlated channels decouple, the divergence engine flags it instead of letting you draw a smooth line over the gap.

What data sources do I need?

Shopify, Meta, Google Ads, GA4, and Klaviyo at minimum. Instagram organic and Microsoft Clarity add additional surface. We integrate via direct OAuth where possible, with Supermetrics as a fallback for accounts that prefer that pipeline. Setup takes about 10 minutes for the connections; the first weekly brief lands the Monday after.

What if my data is messy?

This is the point. On frkl's data, three of the four headline "crises" the team had been chasing turned out to be data artefacts — gifting orders contaminating the discount rate, wholesale revenue inflating the headline, draft orders polluting AOV. We segment first, surface findings second. Most platforms do the opposite.

How does the £/month figure get calculated?

Every finding has a "money recipe" attached — a deterministic calculation against your primary data with the basis shown on every row. Confidence labels (high / medium / low) tell you how much weight to put on the number. We are deliberately conservative — the total is illustrative, not additive (some findings overlap, e.g. site CVR work overlaps with specific JS fixes). The point isn't the precise total; it's that every recommendation has a £ tag rather than vague "this is important."

I already pay for Triple Whale. Why would I switch?

Most operators don't switch — they add. Triple Whale stays as the attribution debate venue for the media buyer. Operator Intelligence sits next to it as the operator's Monday-morning surface. The Scale tier (£599) costs less than Triple Whale Pro (£1,500) and covers eight functions instead of one. The Starter tier (£99) is small enough to add without procurement.

What's the catch?

Three things: (1) we're new — the platform was built and stress-tested on frkl as design partner over Q1-Q2 2026, with brand #2 onboarding next; (2) the Universal Pattern Aggregator (cross-brand intelligence) is live for new tenants but the network effect compounds with more brands on the platform; (3) we don't do attribution modelling — if that's the only thing you care about, stay with Triple Whale.

Is my data safe?

Multi-tenant Supabase with row-level security keyed to your brand_id. We never train models on your data. Data is read-only — we don't write back to Meta, Shopify, or Klaviyo from this platform. The dashboard is a synthesis layer on top of your existing sources, not a system of record.

Find £30k/mo in your business this Monday.

Seven-day free trial. No credit card to start. First brief lands the Monday after connection.