DataSift has no direct competitors in the market and without competition, there are no external forces that are really testing the internal assertion that the company is building a world class software platform. A few years ago, the things DataSift were working on were perhaps fairly novel. However, the processing pipelines DataSift are developing have largely been commoditized by big players through the likes of AWS and Google's big data and stream processing offerings.
DataSift's core technology (SaaS data stream processor deployed to physical servers) now looks a little antiquated in the context of what can be accomplished through more modern managed offerings from the big cloud platform players. To compound this problem, DataSift's biggest asset is has become its biggest risk I think: The commercial relationships with their data partners (eg Facebook) mean that usage of modern, managed offerings for stream processing are simply off the table. This enforces the continued use of more traditional, labour intensive, approaches and means that a lot of time is spent building and maintaining infrastructure components rather than focusing on true value adding software. This is a shame and the effect here is that the company's ability to deliver features that will really drive revenue is being limited.