Every marketing team trusts their attribution data. Most of it is wrong. Cookie deprecation, walled gardens, and multi-device journeys have broken the models companies rely on — and the replacements aren't much better.
The Problem
Last-click attribution was always wrong — now it's dangerously wrong
Last-click gives all credit to the final touchpoint before conversion. It worked poorly when journeys were simple. Now, with cross-device browsing, dark social sharing, and podcast-influenced purchases, last-click attribution credits the wrong channel for the majority of conversions. Companies relying on it systematically over-invest in bottom-funnel and under-invest in the demand creation that actually drives growth.
Multi-touch models create precision illusions
Time-decay, linear, and position-based models distribute credit across touchpoints using arbitrary weightings. They feel more sophisticated than last-click but they're equally arbitrary — there's no empirical basis for giving the first touch 40% credit versus 30%. The precision of the percentages creates a false sense of accuracy that leads to confidently wrong budget decisions.
Privacy changes made all digital attribution less reliable
iOS tracking restrictions, third-party cookie deprecation, and privacy regulations have created massive gaps in attribution data. The journeys you can't see — dark social shares, podcast mentions, word-of-mouth, multi-device browsing — increasingly drive purchasing behavior. Your attribution model only measures what it can track, and what it can track is a shrinking fraction of reality.
Attribution arguments consume more energy than actual growth work
Marketing teams spend weeks debating attribution methodology, building custom models, and arguing about which channel deserves credit. Meanwhile, the actual work of creating demand, building content, and optimizing conversion gets deprioritized. Attribution has become the goal instead of the tool — and the companies growing fastest are the ones who've stopped trying to solve attribution perfectly.
How We Help
We help companies replace attribution-dependent decision-making with a measurement framework that actually works in the current privacy environment. The approach combines three complementary methods: incrementality testing, media mix modeling, and qualitative attribution — none of which rely on individual-level tracking.
Incrementality testing answers the most important question attribution can't: what would happen if we turned this channel off? By running controlled holdout experiments on geographic, audience, or temporal dimensions, you measure the causal impact of marketing activities rather than the correlational data attribution provides. This requires discipline and patience, but it produces the only truly reliable channel performance data.
Media mix modeling uses aggregate data — spend levels, conversion volumes, external factors — to estimate channel contributions without individual-level tracking. Modern MMM approaches update weekly and account for diminishing returns, competitive activity, and seasonality. They're not perfect, but they're directionally accurate in ways that broken attribution models aren't.
Qualitative attribution asks customers how they found you — through post-purchase surveys, sales conversation analysis, and customer interviews. Self-reported attribution has biases, but it captures the dark social, word-of-mouth, and podcast-influenced journeys that digital attribution can't see. Combining qualitative data with incrementality tests and MMM creates a measurement mosaic that's more accurate than any single model.
What we deliver
- Attribution audit — evaluating current model accuracy and identifying where it misleads
- Incrementality testing program — controlled experiments measuring causal channel impact
- Media mix modeling — aggregate-level channel contribution analysis updated regularly
- Measurement framework — combining quantitative and qualitative methods for better decisions
The companies making the best marketing decisions aren't the ones with the most sophisticated attribution models. They're the ones who've accepted that perfect attribution is impossible and built measurement systems that are directionally useful instead of precisely wrong.
Our Methodology
Our measurement methodology starts with an honest assessment of what your current attribution can and can't tell you. Phase one audits your existing attribution infrastructure — identifying where data gaps, tracking failures, and model assumptions create misleading signals. We compare attribution-reported channel performance against actual business outcomes to quantify the accuracy gap.
Phase two designs the replacement measurement framework. We identify which channels are best measured through incrementality tests, which through MMM, and which through qualitative methods. The framework is designed for your specific channel mix, data infrastructure, and decision-making cadence.
Phase three implements the framework and begins producing actionable insights. Incrementality tests run on priority channels. MMM models calibrate against historical data. Post-purchase survey infrastructure captures qualitative attribution. Monthly measurement reviews synthesize all three data sources into channel performance assessments that drive budget allocation.
How We Work
Measurement framework engagements typically run 3-6 months for initial design and implementation, with ongoing retainers for test management and analysis. The first 30 days audit existing attribution and design the replacement framework.
Months 2-4 implement the measurement infrastructure — incrementality test designs, MMM model development, and qualitative survey deployment. First test results typically appear within 60 days, with MMM model calibration requiring 90 days of data.
Months 5-6 begin producing integrated measurement reports that synthesize all data sources. Budget reallocation recommendations are based on the combined evidence from incrementality tests, MMM, and qualitative data.
Ongoing retainers manage continuous testing, model updating, and monthly measurement reports. Quarterly deep-dives reassess channel performance and recommend strategic budget shifts.
- Decision accuracy — budget allocation based on causal impact, not correlational attribution
- Channel clarity — understanding which marketing activities actually drive business outcomes
- Privacy resilience — measurement infrastructure that works without individual-level tracking
If your general company needs thought leadership leadership, we should talk.
Expand your marketing team output with our experts
Let us take a custom approach to your growth goals by assembling and leading the best-in-class marketing team to support your next stage.
Frequently asked questions
How much does it cost to fix our marketing attribution?
Measurement framework projects range from $30K-$75K for design and initial implementation. Ongoing measurement retainers run $8K-$20K monthly. ROI is measured through better budget allocation — reallocating even 15% of spend from over-attributed to under-attributed channels typically improves marketing efficiency significantly more than the measurement investment.
How does can't we just use a better attribution platform?
Attribution platforms can only measure what they can track — and the tracking gaps are growing, not shrinking. Better technology doesn't solve the fundamental problem: individual-level tracking can't capture the full customer journey in a privacy-first world. The measurement framework we build complements your existing attribution with methods that don't depend on tracking pixels.
How long before we can make better budget decisions?
Initial incrementality test results appear within 60 days. MMM models require 90 days of calibration. Qualitative survey data produces useful signals within 30 days. The first integrated measurement report — combining all three methods — is typically available within 90 days. Budget reallocation decisions can begin as soon as the first incrementality tests complete.
What makes Winston Francois different from measurement and analytics firms?
Analytics firms build data infrastructure. We build decision-making frameworks. The difference is practical — our measurement outputs are designed to answer specific budget allocation questions, not produce dashboards. We also bring growth strategy context to measurement, meaning we can interpret data in the context of your overall growth plan, not just channel performance in isolation.
Does this mean we should stop using Google Analytics and attribution tools?
No. Existing attribution tools still provide useful directional data and operational insights — campaign-level performance, landing page optimization, and channel-level trends. What changes is how much you trust the data for strategic budget decisions. Attribution tools become one input among several, not the sole basis for million-dollar allocation choices.
What type of company benefits from this measurement approach?
Companies spending $500K+ annually on marketing across multiple channels. Ideal clients are making budget allocation decisions based on attribution data they suspect is unreliable, or are seeing discrepancies between attributed and actual business performance. If you're spending under $200K annually on two channels, simpler measurement approaches are sufficient.
