lucianfialho/synthetic-attribution-benchmark

Synthetic Multi-Touch Attribution Benchmark

A synthetic marketing dataset of multi-touch user journeys across interpretable channels, with a known data-generating process (DGP) — so the true per-channel credit (exact Shapley value) for every converting journey is included.

Real attribution datasets (e.g. Criteo) are hashed and have no ground truth, so you cannot tell whether last-click / linear / time-decay / Markov / Shapley attribution is correct. Here you can: compare any attribution heuristic against gt_shapley_*.

Fully reproducible from the seed. DGP parameters in ground_truth.json.

Files

• touchpoints.parquet — one row per touchpoint: journey_id, step, channel, t_hours, cost.
• journeys.parquet — one row per journey: journey_id, n_touch, n_channels, converted, conv_prob, and the ground-truth credit gt_shapley_<channel> (0 if the channel is absent or the journey did not convert).
• ground_truth.json — the DGP parameters (channel weights, base logit, half-life) and the aggregate true channel contribution.

Channels & true weights (DGP)

channel	weight	mix_weight	cost
paid_search	1.1	0.9	1.2
social	0.7	0.85	0.8
display	0.35	0.8	0.4
email	0.9	0.5	0.05
organic_search	0.8	0.6	0.0
direct	1.0	0.4	0.0
referral	0.5	0.3	0.1

mix_weight is a categorical mixture weight, not an independent arrival rate. Each touchpoint in a journey is sampled from a single categorical draw over all channels, with probabilities proportional to these weights — not from independent Poisson processes per channel. Journeys that happen to contain a channel multiple times do so by repeated draws, not by a separate channel-level process.

Base logit = -3.2; exposure half-life = 168h; journey window = 720h. Conversion ~ Bernoulli(sigmoid(base + Σ_channel weight·saturate(decayed_exposure))). Ground-truth credit = exact Shapley value of each channel under that value function.

Suggested uses

• Benchmark attribution models: score last-click/first-click/linear/time-decay/Markov against gt_shapley_* (MAE on channel credit shares).
• tabular-classification: predict converted from journey features.
• Teaching material for CRO / marketing attribution.

License

Synthetic data, no PII. CC BY 4.0.