M365Con 2027 – free online conference · 11–22 Jan 2027 · 120 sessions m365con.net

Data gravity

Concept

Data gravity describes how large accumulated data sets attract compute, tools, copies and teams, making the data increasingly hard to move.

✦ CONCEPTData gravityM365.FM
Title graphic: Data gravity (Concept)

Also discussed

Episodes about Data gravity also cover Cost to serve, Process criticality, Azure Autoscale, Scope creep, Non-human identity and Managed Environments.

Description

Data gravity is a concept from IT architecture describing how a large body of data attracts applications, services and other data, much as mass attracts objects in physics. The more data accumulates in one place, the more compute, tools, copies and teams gather around it, and the harder it becomes to relocate. Data lakes and data warehouses are commonly given as prime examples of systems that exhibit this effect.

The term was coined in 2010 by IT expert Dave McCrory. He later distinguished two forms: natural data gravity, which arises as data moves to the cloud, and artificial gravity, which is caused by outside forces such as legislation, throttling or manipulative pricing. In 2020 he released the Data Gravity Index, which measures the effect for large enterprises by metro area and industry.

The core mechanism is scale. Gravity grows with data volume, and moving large data sets is expensive and slow. As a result, applications and services tend to be placed near the data they use rather than the other way around. This has several practical consequences. Applications that sit far from their data face higher latency. Cloud egress fees can make a migration uneconomical. A migration also creates extra work for every application that depends on the data being moved. New gravity keeps appearing as AI, machine learning and edge sources generate additional data.

Typical decisions around data gravity concern where to place workloads and how much data to keep. Suggested responses include co-locating applications and data, applying strong governance and standards, integrating sources into a unified store, planning migrations for the eventual data size rather than the current one, and filtering or analyzing data at the edge before it travels upstream.

The M365.FM episode "Edge vs. Cloud in Manufacturing- Why the Wrong Architecture Costs You More" uses data gravity as one of several placement criteria. It asks how much data is created and what format it actually needs when moving upstream. The episode says that a design that collects everything forever ends up paying for every unnecessary byte more than once. In that discussion, data gravity is weighed alongside outage behavior, where a function should degrade gracefully and keep production running within defined local limits, and alongside scope, where a single sensor or asset favors local placement and broader fleet operations push work upstream.

The episode "Fix Common Azure AI Foundry Project Failures" treats data gravity as an argument for centralizing on one platform. It says that hosting data and AI in Azure reduces latency and complexity, and presents this as a reason teams choose Azure for larger AI projects. The same episode notes that Azure also supports on-premises, hybrid and edge inference, so AI can run where data is generated rather than moving the data.

Taken together, the two episodes show the concept from both directions: moving compute to where data already sits, and reducing or processing data near its source so that less of it needs to move at all.

Overview compiled with AI from M365.FM episodes and the facts on this page. You can improve it with the edit button below.

TypeConcept

Episodes about Data gravity

Episodes in which Data gravity is discussed:

Timeline of episodes about Data gravity

Frequently asked questions

Questions marked with AI were written with AI from the facts on this page.

What is data gravity? (AI)

Data gravity is the tendency of large data sets to attract applications, services and other data, which makes the data harder and costlier to move as it grows. The term was coined by Dave McCrory in 2010.

Why does data gravity matter for cloud migration? (AI)

Moving large data sets is slow and costly, cloud egress fees can make a migration uneconomical, and every dependent application needs extra work. Planning for the eventual data size and co-locating applications with their data are suggested responses.

How do the M365.FM episodes relate data gravity to architecture choices? (AI)

One episode uses it to judge what data should move upstream from manufacturing sites, while another cites it as a reason to host data and AI together in Azure to reduce latency and complexity.

Glossary

Community wiki

Short info above, community knowledge below: add practical tips, examples or corrections. New contributions are reviewed first; trusted members publish directly.

Experts and companies for this topic

Nobody has listed this topic yet. Add it to your profile.

Sources

Maintained by Mirko Peters (Microsoft MVP) · Last updated: