Skip to main content

Reporting Reach in the AdCloudIQ Platform

How AdCloudIQ calculates Reach with speed and precision

Updated today

The AdCloudIQ Reporting Platform is designed to give you clear, actionable insights, without sacrificing speed or flexibility. Whether you're exploring performance trends, adjusting filtering, or analyzing custom date ranges, reporting should feel intuitive and immediate.

To deliver that experience, we focus on three core principles:

Insightful

Providing meaningful, decision-ready data.

Shapable

Allowing you to dynamically filter, segment and explore performance across any dimension(s) and any timeframe.

Performant

Ensuring fast, reliable results, even across large datasets.

Balancing these three priorities isn't simple. Deep insight often requires complex data processing, and dynamic filtering can slow performance at scale. To solve this, AdCloudIQ uses advanced technical solutions, including methods like HyperLogLog for Reach calculation, so you can interact with data seamlessly.

Objective

Our reporting experience is built to:

  • Provide robust, high-level performance visibility through the dashboard

  • Enable detailed, actionable insights within advertiser-specific reporting views

  • Maintain usability and speed, even when processing large volumes of data

Because your time should be spent analyzing results, not configuring reports or waiting for them to process.

HyperLogLog

To support scalable, fast, and flexible reporting, AdCloudIQ calculates Reach using a technique called HyperLogLog.

HyperLogLog is a widely adopted statistical method for estimating the number of distinct values in very large datasets. In our case, the number of unique households reached across a set of impressions.

Why not use exact distinct counts?

Reach is fundamentally a distinct household count, meaning we need to answer questions like: “How many unique households saw all of these impressions?

In the legacy dashboard with fixed reporting, this can sometimes be computed using exact COUNT(DISTINCT ...) queries. However, distinct counting becomes extremely taxing at scale, especially when users apply:

  • Dynamic filters (campaign, line, geo, device, creative, etc.)

  • Custom date ranges

  • Real-time dashboard exploration

  • Multi-dimensional cuts (e.g., Weekday by Daypart, or Creative by Zip Code)

Exact distinct queries require large amounts of memory and compute, and performance degrades quickly as data volume grows. This makes them difficult to support in a self-serve environment where responsiveness and usability are critical.

What HyperLogLog does instead

HyperLogLog provides a highly accurate approximation of distinct reach, while using only a fraction of the computational resources required by exact counting.

It works by summarizing household identifiers into a compact probabilistic structure, allowing AdCloudIQ to estimate unique reach efficiently even across billions of impressions.

Why this is better for customers

Using HyperLogLog allows AdCloudIQ to deliver reporting that is:

  • Fast — Reach metrics return instantly, even with complex filters

  • Scalable — Performance remains consistent as campaign data grows

  • Flexible — Users can explore data dynamically without long query delays

  • Reliable — Estimates are extremely close to true counts, with minimal variance

  • Industry-standard — This approach is commonly used by leading analytics platforms for large-scale measurement

Ultimately, HyperLogLog ensures that AdCloudIQ can provide Reach reporting that balances accuracy with the real-world needs of modern self-serve dashboards: speed, interactivity, and scale.

Accuracy and Confidence

While HyperLogLog is technically an approximation, it is not a simplistic or outdated estimate. It is a highly sophisticated algorithm originally developed by computer scientists at Google to solve the challenge of measuring distinct users at massive scale. This isn't a "best guess" as HLL is a statistically rigorous technique designed for internet-scale measurement, and it's widely used by companies like Google, Amazon and other to measure distinct users efficiently. It's also why many reporting systems will round reach to the nearest thousand.

In practice, HyperLogLog delivers Reach calculations that are extremely close to true distinct counts, with a very small and well-understood margin of error: <1.04% on average. This approach is trusted across the analytics industry because it provides the best balance of:

  • Precision

  • Speed

  • Scalability

For customers, this means AdCloudIQ's Reach metric remains both statistically reliable and operationally performant, even when reporting becomes more dynamic and data volumes grow.

Did this answer your question?