Fully Managed Service

A fully managed service is a delivery model in which the provider owns the entire data collection process end to end: source selection, extractor build, scheduling, monitoring, validation, ongoing maintenance and delivery into the customer's systems. The customer defines what data they need and where it should land, and the provider absorbs the continuing work of keeping it flowing as target sites change. The model is also described as data as a service. At Import.io it is one of three ways to work with web data, alongside a self-service extraction platform and Aperture, and choosing between them usually comes down to who carries maintenance rather than which one collects better data.

Why it matters

• Takes maintenance off the customer's engineering roadmap, which is where most in-house collection projects quietly stall
• Shifts the cost question from subscription price to total cost of ownership, counting engineering time, proxy infrastructure and rework
• Fits teams whose data requirement is permanent while their capacity to maintain collection is not

How it is used

• Multi-source programmes spanning many retailers, regions and refresh schedules
• Projects on defended sources where blocking and frequent layout change make in-house upkeep expensive
• Scheduled delivery into an existing warehouse, BI tool or dashboard

A fully managed service and a self-service platform with support attached are different arrangements. Under self-service the customer builds and owns the extractors, and picks the work back up when a site changes. Under a fully managed service the provider owns both the build and the upkeep. Managed hosting is a separate idea again and refers to infrastructure rather than to the data.