Attribution has always been an approximation. The version built on third-party cookies was deterministic in appearance but systematically incomplete — missing Safari users, Firefox users, anyone who declined consent, anyone using an ad blocker. What replaces it is more transparent about its uncertainty, but no less actionable. The shift is from false precision to calibrated triangulation, and the approach is available to most B2B marketing teams today without enterprise-scale tooling.
Why No Single Method Is Enough
Each attribution method available in a cookieless environment has a distinct angle on the truth and a distinct set of blind spots. Platform-reported conversions (Google Ads, LinkedIn, Meta) are biased toward crediting the platform's own channels and overlap with each other via modeled attribution. Web analytics (GA4, Plausible, or similar) show last-click or session-based attribution for consenting, cookied users — a subset of actual converters. Self-reported attribution captures intent signals that neither of the above can see. Media mix modelling (MMM) reveals macro spending-outcome relationships that cookie-level tracking never could.
Using any single method in isolation produces a distorted picture. Using all of them together — and looking for where they agree — produces the confidence needed to make real budget decisions. Our pillar on marketing attribution explained covers the foundational concepts behind each method if you need to orient a team around the vocabulary first.
Self-Reported Attribution: Ask How They Found You
Self-reported attribution — sometimes called "how did you hear about us?" or survey attribution — is the simplest and most underused method available to B2B marketers. At the point of conversion (demo request, form submission, purchase), you ask the prospect or customer a single question about which channel or source brought them to you.
The data is qualitative, subject to recall bias, and incomplete by nature — not everyone will answer. But it captures something no other method can: channels that are genuinely influential in driving decisions but do not produce a trackable click or conversion event. Podcast appearances, word-of-mouth, LinkedIn organic posts, newsletter reads on mobile — these regularly appear in self-reported data and are invisible to everything else in the stack.
Implementation is low-friction: a single dropdown or open text field on your conversion form, synced to your CRM. Aggregate the responses weekly. Look for patterns in which sources your best-fit customers cite, not just total volume. Even with partial response rates, you will quickly identify channels that are influential but appear invisible in platform dashboards.
MMM-Lite: Spend vs. Outcome Regression
Marketing mix modelling (MMM) at the enterprise scale involves econometric models built and maintained by dedicated data science teams. MMM-lite is the accessible version: a lightweight regression analysis that asks whether changes in your marketing spend, across channels, correlate with changes in your business outcomes over time.
The core idea is simple: if you have a weekly or monthly time series of spend by channel (paid search, paid social, email, events, display) and a corresponding series of outcomes (pipeline created, trials started, revenue), you can model the relationship between the two using regression techniques available in any spreadsheet or basic data tool. You do not need an advertising econometrician. You need clean data and the willingness to interpret coefficients with appropriate caution.
MMM-lite does not require cookies and is not affected by consent signals, browser restrictions, or cross-device gaps — it operates entirely on aggregated spend and outcome data. Its limitations are that it requires meaningful time series length to produce reliable coefficients, and it cannot resolve attribution at the individual journey level. But for strategic budget allocation decisions — where to shift spend across channels — it is often more reliable than multi-touch attribution built on incomplete cookie data.
Platform Data and Analytics Cross-Reference
Platform-reported data (Google Ads conversions, LinkedIn conversion tracking, Meta Pixel events) and web analytics (GA4 events, Plausible goals) both remain relevant in a cookieless setup — they just need to be read differently than they were when third-party cookies provided cross-site continuity.
Platform data is strongest as a within-platform efficiency signal: conversion rates, cost per result, impression share trends, and relative performance between campaigns and ad sets within the same platform. Do not use it to compare absolute conversion volumes across platforms — the overlap and attribution windows make cross-platform comparison misleading.
Web analytics tools show what consented, cookied users do on your site. GA4's default attribution is last non-direct click for the session, supplemented by data-driven attribution for accounts with sufficient conversion volume. Neither is complete for your whole audience, but trends over time — particularly in entry page performance, landing page conversion rates, and organic traffic growth — remain reliable signals even with partial data coverage.
The cross-reference discipline is to pull all three sources (self-reported, platform, analytics) into a single view — even a simple spreadsheet — on a monthly basis and look for consistency and divergence. Where they agree, increase confidence. Where they disagree sharply, you have a measurement gap worth investigating.
Conversion APIs and First-Party Event Matching
All major advertising platforms now offer conversion API integrations (Meta CAPI, Google's Enhanced Conversions, LinkedIn Insight Tag server-side events) that allow you to send conversion data directly from your server to the platform, matched against first-party signals like hashed email addresses. This server-to-server path bypasses the browser entirely — it is not affected by ad blockers, ITP, or consent mode restrictions on client-side pixels.
For B2B advertisers, where CRM data is typically the authoritative source of conversion events (lead created, opportunity opened, deal closed), conversion APIs allow you to pass those CRM events back to ad platforms with first-party matching. The match rate depends on the quality of your data — how often you have hashed email or phone number for matched conversions — but even partial coverage substantially improves signal completeness versus relying solely on client-side pixels.
This pairs naturally with the server-side tagging approach covered in our companion piece — together they form the technical foundation of a cookieless measurement setup that does not wait for platform-level deprecations to force a migration.
Building the Habit, Not Just the Stack
The technical components — self-reported attribution questions, a lightweight MMM spreadsheet, conversion API setup, server-side tagging — are only half the answer. The other half is the analytical habit of reviewing all three signals together on a cadence, drawing incremental budget decisions from the convergence, and being explicit about what you do and do not know.
Post-cookie attribution is not worse than what came before — it is more honest. The apparent precision of multi-touch last-click models built on third-party cookies concealed systematic gaps. A triangulated view that acknowledges uncertainty and looks for cross-method agreement is a more robust basis for marketing investment decisions, even if it is less satisfying to report in a single dashboard number.
Frequently Asked Questions
How many responses do I need for self-reported attribution to be useful?
There is no fixed threshold, but the method becomes meaningfully directional as you accumulate responses over weeks and months. Even partial response rates — say, half of converters answering the question — reveal patterns that are invisible elsewhere. Focus on trends and relative frequency across channels rather than treating individual data points as definitive.
Is MMM-lite appropriate for small B2B teams with limited budgets?
Yes, with appropriate expectations. MMM-lite is most reliable when you have at least several months of consistent spend data across at least two or three channels. With thin data or highly variable spend, the regression coefficients will be noisy. Use it to inform directional decisions rather than precise budget splits.
Do we still need GA4 if we use conversion APIs?
Yes — they serve different purposes. Conversion APIs send specific events to ad platforms for optimisation and reporting. GA4 (or another analytics tool) gives you site-level behaviour data, content performance, funnel drop-off, and audience insights that platform dashboards do not provide. They complement rather than replace each other.
Build a measurement plan that works without cookies
Hatch's free plan tool helps you design a post-cookie attribution framework tailored to your channels, team size, and technical setup — without needing enterprise resources to get started.
Free Plan Tool