Research areas

Six areas, one aim: explain what the data shows

Each area has its own questions and methods. Many studies combine several of them.

01 · NETWORK TRAFFIC ANALYSIS

How traffic is composed, and how it changes

We analyse traffic volumes, flows, protocols, sources and destinations to describe how networks are used. Time-based views show daily and weekly cycles, growth and shifts in composition, so capacity and policy decisions rest on evidence.

Typical questions

  • What makes up our traffic, and how is the mix changing?
  • When and where do peaks occur?
  • Which flows account for most of the load?
02 · ONLINE DATA ANALYTICS

How digital services are actually used

We study usage, engagement and journey data to understand behaviour on websites, applications and online platforms, connecting what people do with how the service performs.

Typical questions

  • Where do users drop out of a journey?
  • How does usage differ by time, region or device?
  • Which changes affected engagement, and by how much?
03 · PERFORMANCE & RELIABILITY RESEARCH

Speed and stability, measured as distributions

We examine latency, throughput, error rates and availability across their full distributions, because averages hide the slow and failed requests users notice most. Results show where problems concentrate and what they relate to.

Typical questions

  • What do the median and the slowest requests look like?
  • Which conditions are linked to errors or slowdowns?
  • How does reliability vary over time?
04 · ANOMALY & PATTERN ANALYSIS

Telling real change from normal variation

We characterise normal behaviour first, including its cycles and spread, then identify deviations that are unlikely to be noise. Recurring patterns are described so they can be anticipated rather than rediscovered.

Typical questions

  • Is this spike unusual, or part of a known cycle?
  • Which patterns repeat, and on what schedule?
  • What preceded an unexpected change?
05 · ANALYTICAL METHODS & MODELING

The right method for each question

We select statistical and computational methods to suit the question and the data, test their assumptions, and document the choices made. Models are used where they add explanatory or predictive value, and their limits are stated.

Typical questions

  • Which approach fits this data and this decision?
  • How sensitive are the results to our assumptions?
  • How much can a model explain, and where does it fall short?
06 · RESEARCH REPORTING & VISUALIZATION

Findings people can follow and use

We present results in reports and visuals designed for the people who will act on them: what was found, how certain it is, and what it suggests. Supporting detail is available for those who want to check the work.

Typical deliverables

  • Written report with findings and recommendations
  • Charts and visual summaries for decision makers
  • Method notes describing data, assumptions and limits

Not sure which area fits?

Describe the question. We will suggest how to approach it.

Start a Conversation