Proxies and Lift Studies

Measure impact when direct attribution is not available.

Some sales cannot be connected directly to advertising activity. We create useful proxy metrics for ongoing decisions and structure lift studies around actual sales when you need to measure incremental impact.

  • Comparable proxies

    Bring different platform metrics into one perspective that reflects the outcome you care about.

  • Actual sales lift

    Evaluate how channels, line items, creatives, and budget changes affect real commercial outcomes.

  • Maintained for you

    Keep definitions, platform metrics, and recurring processes current without rebuilding the plumbing.

When sales cannot be attributed directly

You still need a dependable way to understand performance, compare activity, and improve how budget is allocated.

Proxies support recurring optimization when sales are not observable at the required level. Lift studies answer a different question by measuring whether a controlled activity changed actual sales.

Unobservable sales

Important outcomes often sit outside the advertising platform

Direct attribution can be unavailable because the sale happens offline, across many locations, or inside a digital environment that does not share sales data.

  • Offline sales: purchases happen across physical locations without a direct path back to individual advertising activity.
  • Streaming destinations: purchases or rentals can happen on Prime Video or similar services without detailed sales exports.
  • Digital walled gardens: the destination controls the outcome data and does not expose enough detail for direct budget allocation.
You cannot see directly what performs and what does not, but you can still create a useful measurement structure.

Proxy design

Use the best measurable approximation of the outcome

A proxy is the measurable signal that comes closest to the sale or business outcome you cannot observe directly.

  • Outcome relevance: select a signal that reflects meaningful progress toward the real commercial result.
  • Consistent availability: use metrics that can be collected reliably enough to support recurring decisions.
  • Clear interpretation: document what the proxy represents and where its limits begin.
A proxy does not pretend to be a sale. It creates the closest dependable perspective available for optimization.

Quality views

Different platform metrics become one usable perspective

Advertising platforms often describe a similar behavior with different metrics. We understand those definitions and bring them into a shared proxy model.

  • Meta: fifteen second video views can represent meaningful attention.
  • TikTok: six second video views provide the closest available platform measure.
  • Google Ads: TrueView metrics include viewing and relevant engagement behavior.
  • Shared model: the source metrics remain transparent while the data model creates one quality view perspective.
The goal is not to claim that different metrics are identical. It is to make them comparable enough for the strategy and decisions they support.

Lift studies

Change an activity and observe what happens to actual sales

A simple lift study compares sales during a controlled change with the sales level that would otherwise be expected.

  • Choose the activity: test a channel, line item, creative, market, or budget level that can be turned on, turned off, or changed for a defined period.
  • Set the comparison: use a comparable time period, location, audience, or control group to establish the expected sales baseline.
  • Measure actual sales: evaluate whether sales moved beyond the baseline while the selected activity changed.
  • Repeat carefully: account for seasonality, promotions, and other influences before treating the observed difference as meaningful lift.
Lift studies do not use the proxy as the outcome. They measure the impact of controlled advertising changes on actual sales.

Infrastructure at scale

Focus on improving the process instead of maintaining the plumbing

Whatever proxy strategy or lift study structure fits your setup, we build the data foundation and maintain it as platforms change.

  • Platform knowledge: keep metric definitions and collection logic aligned with current platform reporting.
  • Unified data model: combine source metrics while preserving the context needed to interpret them correctly.
  • Repeatable tools: automate recurring proxy reporting, study comparisons, and decision outputs.
  • Ongoing maintenance: adapt the process when platforms, exports, or measurement requirements change.
Your team can spend its time making measurement and media decisions better instead of tracking platform updates and reconciling metrics manually.

Let's explore the right measurement setup together.

Tell us where sales happen, what the platforms return, and which activities you can control. We will map the right proxy or lift study structure for your setup.