1. Descriptive Analytics

Experiments and Causal Boundaries

Estimate what changed because of an action rather than after it

Experiments and Causal Boundaries

The missing outcome

For customer ii, let Yi(1)Y_i(1) be the outcome under a new free-delivery message and Yi(0)Y_i(0) under the current message. The individual effect is

τi=Yi(1)Yi(0),\tau_i=Y_i(1)-Y_i(0),

but only one potential outcome is observed. Random assignment makes groups comparable in expectation, allowing the difference in group means to estimate an average effect.

HarborMart experiment

GroupEligible customersPaid ordersConversion
control10,0002,40024.0%
treatment10,0002,52025.2%

Estimated lift is 1.2 percentage points, or 5% relative to the control rate. The standard error for two independent proportions is approximately

SE=0.24(0.76)10,000+0.252(0.748)10,0000.0061.SE=\sqrt{\frac{0.24(0.76)}{10{,}000}+\frac{0.252(0.748)}{10{,}000}} \approx0.0061.

A rough 95% interval is 0.012±1.96(0.0061)0.012\pm1.96(0.0061), approximately 0.0 to 2.4 percentage points. “Positive point estimate” and “precisely established business value” are different claims.

Design before significance

Design itemHarborMart choiceWhy it matters
unit of randomisationcustomer accountavoid a customer seeing both messages
eligibilitycustomers with a serviceable addresspreserve target population
primary outcomecontribution per eligible customerincludes discount cost and conversion
guardrailslateness, refunds, support contactscatch displaced harm
horizon28 days plus delayed refund windowinclude foreseeable effects
analysisintention to treatpreserve randomisation
stopping rulefixed or pre-specified sequential planavoid repeated peeking bias

First run an A/A or instrumentation check. A sample-ratio mismatch, duplicate assignment or missing outcome can invalidate a very small p-value.

Interference and marketplace effects

If treatment increases demand, control customers may face fewer slots. One customer’s treatment then changes another customer’s outcome. Randomising individual customers does not isolate the marketplace effect.

Possible responses include cluster or switchback randomisation, capacity-aware designs and explicit equilibrium analysis. The design must match how units interact.

Airbnb’s 2025 experimentation review illustrates current work on long-run ranking dynamics and batch-adaptive experiments. It is an organisational research case, not proof that one design fits every marketplace.

When randomisation is unavailable

Before/after and treated/untreated comparisons can be confounded by season, targeting, selection and concurrent changes. Use domain-supported designs—difference-in-differences, discontinuities, instruments, matching or synthetic controls—only when their identifying assumptions are defensible.

Write the boundary:

“The adjusted association is useful for forecasting and prioritising a future experiment; it does not identify the effect of sending the offer.”

Quick check

Treatment raises conversion but lowers contribution per eligible customer. Is the experiment a success?

Answer
Not under a contribution objective. Conversion is a driver; discount cost, order mix and guardrails determine value. Use the pre-specified outcome and investigate heterogeneity without redefining success after seeing results.

Foundation

Next: Predictive Analytics

Copyright © 2026