Big data, analytics and predictive insights are very much the flavor du jour in technology marketing land. Vendors and their PR flacks are quick to jump on these buzz words that have seen much interest from customers. DataSift, a vendor that delivers a machine learning platform, is trying to deliver these buzzwords as a platform that is applicable to all and usable by any developer. Or is it?
There is an interesting trend occurring in the general machine learning space. On the one hand, the real value of a machine learning tool is that it is broadly applicable to whatever datasets and use cases individual customers have. On the other hand, the difficulty with a very broad platform approach is that it doesn't really provide bounds within which customers can operate. Organizations get the concept of what machine learning is all about, but without specifically tailored offerings, they often find it hard to know how to apply the technology to their situation.
It is for this reason that we have seen a huge number of startups offer very specific solutions that are tailored to sales and marketing use cases. This is in part because of the obvious value that sales and marketing sees in analytics, but also because sales and marketing needs (i.e. lead scoring and customer intent) are relatively simple ones to create and tend to be fairly generic across organizations.
Trying to move away from this vertical trend, and apparently going for a broad horizontal platform offering is DataSift, which is today announcing VEDO intent, a platform that aims to simplify machine learning but without constraining it to particular use cases. The idea of the product is to help developers to place machine learning into the hands of business users across different departments within an organization. While the offering is focused on social data as a source for insights, DataSift claims that it doesn't pre-suppose where those insights will be applied.
In justifying this approach, DataSift opines that while social media is vital for marketing and customer service, extracting industry or customer insights has been limited to analyzing brand health and measuring sentiment (positive, negative, or neutral). "With this broader approach to a platform, customers can take the analysis further, gaining more nuanced insights into market interest and surfacing customer intent to churn, purchase and more," says the company. "VEDO Intent builds on VEDO, the programmable intelligence engine built into DataSift's data processing platform. It allows people across the business, including marketers, customer relations managers and more, to easily build advanced machine learning-based classifiers for their organization without the need of a data analyst."
Using an approach known as Assisted Learning, VEDO Intent learns as posts are manually classified into categories, such as rant, rave, purchase intent, or churn. VEDO Intent then dynamically builds a machine learning-based model to first suggest, and then fully automate, the real-time classification of millions of posts to surface insights that previously would have been hidden. By having a system carry out the data analysis, VEDO Intent also protects individuals' privacy. Customers can generate answers to questions such as:
- Which products are people considering buying?
- How satisfied are my customers? Are they thinking of leaving me?
- How important is this campaign message?
- Are my target markets interested in my marketing program?
- Which parts of my service prompt the best reviews?
It's certainly a broader offering than the very focused predictive analytics solutions out there, but I'm not sure if it really goes far enough. While the company talks about this offering "democratizing the building of advanced applications for human-created data," it's still looking at the problem through a sales and marketing lens. There are so many use cases that this sort of number crunching can apply to - scoring credit, for example, or advanced insights into economic factors that may affect the business over time.
Of course, it is easy to understand why DataSift is somewhat constraining the areas this platform is moving into. And it is for the same reasons that other vendors are similarly constrained. Customers, while making platitudes about very broad analytical tool sets, actually need constraints to help inform them about the applicability of analytics for their particular organization. Democratizing machine learning is useful, but if that ultimate democratization means that customers don't know what to do with a platform, it is somewhat irrelevant.
I suspect that, over time, we will see a maturing in terms of customers' understanding of the opportunities and applicability of these sort of tools. Once we do, we will see vendors further broaden their propositions to truly be "blank canvases" within which customers can create their unique solutions. Until customers get to that position, however, DataSift and its competitors will need to introduce constraints into the system.
This article is published as part of the IDG Contributor Network. Want to Join?