Search still on Microsoft's research radar
Developed by Microsoft Research, Viveri is designed to be a sandbox where Microsoft can try out new search ideas. — (Credit: Ina Fried/CNET Networks) — REDMOND, Wash.—Microsoft has a lot of ideas on how Web searching could be better.
Context & Ripple Effects
Microsoft's search ambition has been a running thread for years: a stealth Silicon Valley search team surfaced in 2007, followed by leaked internal plans the same fall. By early 2009, with critics arguing Redmond should stop chasing Google on its own turf, Microsoft Research's Viveri sandbox is the company's answer that the problem remains open.
First-order effects
- Microsoft gains an internal test bed (Viveri) where new ranking and query ideas can be trialed without touching its production search service — a direct response to arguments that web search is still an unsolved problem.
- Microsoft Research staff get a dedicated vehicle to move lab prototypes toward product, reinforcing the innovation pipeline showcased in the group's 100-innovation showcase two years earlier.
Second-order effects
- A working experimentation platform lowers the cost of iterating on search quality, which could let Microsoft close relevance gaps faster and force Google and Yahoo! to keep raising their own pace of feature releases.
Third-order effects
- If sandbox-to-product pipelines like Viveri become standard, search competition shifts from marketing scale to research velocity — a structural dynamic that later shows up in how Microsoft keeps repositioning search surfaces, as in its 2017 pivot to Windows taskbar search.
- Sustained R&D in search lays groundwork for the next paradigm shift; the same 'search isn't solved' thesis eventually underwrites AI-assisted retrieval, where publisher-facing controls become a battleground.
The trend: Two decades of persistent investment show Microsoft treating web search as a permanently unfinished research problem rather than a settled market it lost.