Lead Nurture · Analytics & Attribution

Nurture analytics and attribution: every touch credited, every deal explained.

A multi-touch attribution engine that credits every email, ad, WhatsApp, sales call, and webinar against the eventual closed deal — with plain-English explanations of what actually drove the close. Deploy it inside your stack, or co-deploy it with your agency.

Multi-touch attribution · live
shapley + markov · v3.4 · prod
Live
Acme Cloud38d
Enterprise SaaS · 240 staff
$48KClosed-won
Touch ledger · fractional credit AI · scoring
Email · welcome series
2.4%
Retarget · Meta · pricing creative
8.1%
Webinar · attended live
14.3%
WhatsApp · sales follow-up
22.6%
Demo booked + attended
31.0%
Proposal sent
21.6%
attributing credit…
What the system is doing
journey.readdeal=D-1124 · 38d window 6 touches · 4 channels

What the attribution engine does

Attribution as a system component — not a Monday-morning report.

One coherent revenue-attribution layer across every channel you actually run — fractional credit, plain-English explanations, and live anomaly detection wired into the systems your team works in.

Multi-touch fractional attribution to closed revenue

Every touch in the journey gets fractional credit against the eventual closed deal — not last-click, not heuristic splits. Markov-chain + Shapley-style modelling, calibrated to your actual conversion data.

Cross-channel coverage — every surface in the nurture

Email, SMS, WhatsApp, retargeting, push, in-app, sales activity, webinars, calendar holds. One coherent attribution layer across every channel you actually run, not the three your dashboard happens to track.

Plain-English deal explanations

Per closed-won deal, the system writes a 1-2 sentence explanation of what mattered — which touches anchored, which broke a stall, which were noise. Readable by your sales lead. Defensible to your CFO.

Anomaly detection on channels + creatives

Channel performance shifts and creative fatigue surfaced in real time, with confidence bands. 'Retarget creative B-3 · −24% credit · last 24h' lands as an alert before it eats a quarter of pipeline.

Leakage alerts on dropout cohorts

Cohorts that drop out of the journey unexpectedly get flagged with a plain-English root-cause hypothesis — missing follow-up, broken handoff, or stage friction — so the gap closes before the report ships.

Multi-tenant for agencies — branded reports

Workspace isolation, per-client attribution models, branded explainability outputs. Run attribution-as-a-service across a client roster from one deployment, without leaking signals between books.

Where attribution earns its keep

One engine, every revenue motion.

Same ingestion, same fractional-credit core, same explainability — tuned per motion. PLG SaaS, mid-market enterprise, e-commerce lifecycle, agency multi-tenant, marketplace, and long-cycle services all run on the same engine; only weights, channel mix, and integrations change.

01

B2B SaaS · trial-to-paid attribution

PLG signal weighting fused with marketing + sales attribution. Trial-activation events and paid-touch credit fold into a single revenue picture, not two disconnected charts.

02

B2B mid-market · multi-stakeholder

Account-based attribution rolled up across the buying committee. Champion engagement, CFO discovery, technical demo all credited fractionally — without a single AE having to log it manually.

03

E-commerce · acquisition + lifecycle

Attribution across acquisition, win-back, post-purchase, and reorder. SMS recovery codes get credit when they earn it; long-tail email lifecycle gets credit when it earns it. No more last-click overwriting it all.

04

Agencies · attribution-as-a-service

Workspace-isolated attribution per client book, branded reporting, monthly model retunes. Run it as a productised offering across a client roster without spinning up separate stacks per account.

05

Marketplace · two-sided attribution

Buyer-side and supplier-side attribution run as separate models with separate credit logic. Both sides surface in the right operator inbox; cross-side spillover gets flagged when it matters.

06

Long-cycle services · retainer + consulting

Six-month sales cycles get scored on momentum, not last-touch. Webinar series, white-paper downloads, podcast appearances all earn fractional credit when the deal eventually closes.

Live attribution

See exactly which touches are moving revenue — in real time.

Live touchpoints landing in the feed on the left, the closed deal explained in the middle, channel performance and anomalies on the right, AI attribution events streaming across the bottom. Every credit decision visible to operators as it happens.

attribution-engine.live
live
Revenue attributed · today$186.0K
Avg touches to close7.2
Top channelRetarget · 28%
Leakage detected3 cohorts
Live touchpointslast 60s
listening for events…
Closed deal · explained AI · attribution
Acme Cloud· 38 days
Enterprise SaaS · 240 staff · Series B · 4 channels touched
$48K ARR
Multi-touch breakdown
Email 2.4%Retargeting 8.1%Webinar 14.3%WhatsApp 22.6%Sales 52.6%
Journey timeline · fractional credit
Day 0
Welcome series · open + click
warmed account before pricing visit
2.4%
Day 4
Meta · pricing creative
anchored intent · 11 impressions
8.1%
Day 11
Live attendance · Q&A
1st value moment · stayed 38 mins
14.3%
Day 28
Sales follow-up · day 28
broke a 6-day stall · re-anchored champion
22.6%
Day 31
Demo booked + attended
primary anchor · CTO joined
31.0%
Day 35
Proposal sent · v1
closed within procurement window
21.6%
AI · why this deal closed
WhatsApp sales follow-up at day 28 broke the stall — without it, model predicts 18% lower close probability. Demo + proposal acted as a paired anchor.
Channel performancetoday
Retargeting
$52.4K
14%
Email
$41.2K
6%
WhatsApp
$36.8K
9%
Sales
$28.6K
3%
Webinar
$18.7K
4%
SMS
$8.4K
18%
Anomalies AI
Retarget creative B-3-24%
−24% credit · last 24h
Email Welcome v2+18%
+18% credit · last 24h
WhatsApp follow-up · APAC+11%
+11% credit · 7d
AI attribution logstreaming · last 5
listening for attribution events…

Models inside the engine

No single model gets attribution right. So we stack them.

Each model handles a distinct part of attribution — journey reasoning, fractional credit, anomaly detection, leakage classification — and the outputs combine into one auditable revenue picture.

LLM-BACKED JOURNEY ANALYSIS
Path Reasoner

Reads the full journey across every channel and identifies which touches drove the close — anchor touches, stall-breakers, passive keepalive — with plain-English reasoning per deal.

MARKOV + SHAPLEY-STYLE
Credit Allocator

Fractional credit assignment combining Markov-chain removal effects with Shapley-style coalition weighting. Calibrated to your conversion data, not a generic split.

CHANNEL + CREATIVE SHIFT
Anomaly Detector

Watches per-channel and per-creative attribution against rolling baselines. Significant shifts surface with a confidence band before they compound into a missed quarter.

DROPOUT COHORT DETECTION
Leakage Classifier

Identifies cohorts dropping out of the journey unexpectedly and labels the likely cause — missing follow-up, broken handoff, stage friction — for revops to action.

Channels and sources wired into every credit

Every input that earns credit — integrated.

CRM outcomes, marketing engagement, ad-platform conversions, behavioural events, sales activity, and commerce signals — pulled in parallel, normalised into a single journey schema, versioned alongside the attribution model.

Source
What it credits
Providers
CRM + pipeline outcomes
Stage transitions, won / lost outcomes, deal velocity, and meeting attendance — the closed-revenue ground truth that every attribution credit gets calibrated against.
HubSpotSalesforcePipedriveAttioClose
Marketing engagement
Email opens, link clicks, send-time response, unsubscribe signals — merged with multi-channel engagement (SMS, WhatsApp, push) for full nurture coverage.
HubSpotCustomer.ioMarketoIterableKlaviyo
Ad-platform conversions
Server-side conversion signals — view-throughs, retarget engagements, lead-form completions — credited fractionally against the eventual close, not double-counted across platforms.
Meta CAPIGoogle ECLinkedIn CAPITikTok Events
Behavioural events
Pageviews, button clicks, video plays, scroll depth, form interactions — captured client-side and server-side, normalised into a single event schema before crediting.
SegmentRudderStackSnowplowPosthogMixpanel
Sales activity
Sequenced outreach, dialer calls, meeting attendance, recorded-call talk-time, calendar holds. Sales-side touches earn credit alongside marketing — one revenue picture, not two.
OutreachSalesloftGongChorusCalendly
Commerce + revenue
Order events, subscription transitions, refund signals, expansion ARR — closed-revenue events that anchor the attribution model and feed back into weekly tuning.
ShopifyStripeChargebeeRecurlyNetSuite

Solution & technology — not a campaign service

Built to deploy. Built to co-deploy.

Revenue attribution shouldn't be locked behind a managed-service contract or a 12-month BI rollout. Our system is infrastructure you deploy inside your stack — directly, with your engineers, or with us riding alongside. Agencies productise it across a client roster; in-house growth teams embed it into their reporting layer; founders run it as the spine of how they answer the "what's actually working?" question.

  • In-house growth teams Run it as a permanent component of your stack — your engineers own the deployment, we tune the models against your closed-revenue data.
  • Agencies + consultancies Embed it in client engagements with workspace isolation and white-label reports. Attribution-as-a-service across the whole roster from one deployment.
  • Founder-led ops Spin it up on top of your CRM in weeks. No analytics team needed to operate it — the system writes the deal explanations and ships the alerts.
DEPLOYMENT MODELS
attribution.deploy · v3
Self-deployYour engineers · we hand over the system
Repo access · CRM + channel integration playbook · model weights tunable · monthly review
Co-deployWe embed alongside your team
Shared workspace · pair-build during rollout · we step out, you keep running
Agency white-labelRun it across your client roster
Multi-tenant · per-client attribution model · branded reports · agency-owned billing
ManagedWe run it end-to-end
Ops, model tuning, weekly reporting · your team owns strategy, we own the system

Why Axccelerate for attribution

Not a dashboard.
An attribution system.

Point tools deliver one channel's view. We deliver the whole substrate — fractional credit, plain-English deal explanations, anomaly detection, leakage classification, multi-tenant agency support — with the auditability your CFO and your clients can actually read.

Capability
Axccelerate
Point-tool vendor
In-house build
Multi-touch fractional credit (not last-click)
Varies
Cross-channel coverage (email + SMS + WA + ads + sales + webinar)
Varies
Plain-English deal explanation per closed-won
Real-time anomaly detection with confidence
Varies
Leakage alerts on dropout cohorts
Varies
Workspace-isolated multi-tenancy (agencies)
Markov-chain + Shapley-style attribution (not heuristic)
Varies
Varies
Plug-in to any CRM — not a CRM replacement
Varies
AI insights surfaced (not just dashboards)
No vendor lock-in · models exportable

Pricing

Priced to the deployment, not the touch volume.

We scope against your channel mix, your CRM, and your client/book footprint before quoting. Agency multi-tenant deployments price differently from a single in-house team — both are first-class citizens.

Launch
Enquirefor pricing
One book · agency-deployable

One brand or one client book in production. Multi-touch fractional attribution, cross-channel coverage, plain-English deal explanations, branded reporting. Scoping fits an agency embedding the system into a single client engagement.

1 brand or client book
Markov + Shapley-style attribution
Cross-channel coverage (6 channels)
Plain-English deal explanations
Branded weekly + monthly reports
Enquire for pricing
Most popular
Scale
Enquirefor pricing
In-house teams · agency multi-client

Up to 4 brands or client workspaces. Per-workspace attribution models, custom channel weights, weekly model tuning, anomaly + leakage alerts piped into Slack or your incident channel of choice.

Up to 4 brands or client workspaces
Per-workspace attribution models
Weekly model tuning
Anomaly + leakage alerts
Branded explainability for clients
Enquire for pricing
Fleet
Enquirefor pricing
Platform · agency at scale

Unlimited workspaces, dedicated attribution engineer, custom data sources, white-label dashboards, and on-call support. Designed for agencies running attribution-as-a-service across a roster.

Unlimited brands / client workspaces
Dedicated attribution engineer
Custom data sources + integrations
White-label dashboards + reports
On-call support · SLA-backed
Talk to us

FAQ

Common questions.

Don't see your question here?

Ask us directly
Deploy · co-deploy · white-label

Your attribution layer — engineered.

30-minute scoping with a senior engineer. You'll leave with a deployment plan, channel-by-channel integration sketch, and realistic timeline — whether you're shipping it for your own pipeline or for a client roster.