In early June, Amalgam Insights attended Alteryx Inspire ‘18, where Alteryx Chairman and CEO Dean Stoecker led an energetic keynote to inspire their users to “Alter(yx) Everything.” Based on conversations I had with Alteryx executives, partners, and end-users, I came away with the strong impression that Alteryx wants to make advanced analytics and data science tasks as easy and quick as possible for a broad audience that may not know code – and they want to expand that community and its capabilities as quickly as possible. Data scientists and analytics-knowledgeable employees are in high demand, and the shortage is projected to worsen as the demand for these capabilities grows; data is growing faster than the existing data analyst and data scientist community can keep up with it.
Alteryx has built an enthusiastic community around the Alteryx Analytics platform, and it can credit its ease of use as key to that. Besides appreciating the product itself for the attention it pays to the end user experience, it was clear just how much Alteryx users actually love Alteryx. What other enterprise software company can fill an arena with a couple of thousand people, eagerly watching three of their most competent end-users using that software to solve three business problems inside a 45 minute sprint? Alteryx does it with their annual “Grand Prix” competition at Inspire. The people sitting behind me at the Grand Prix argued cheerfully the entire time over which analytics expert they thought would win, based on their performance in past Alteryx challenges, and whether they thought each was constructing their workflows with the right tools. (Congratulations to Nicole Johnson of T-Mobile for a photo finish of a win!) Alteryx made watching a competitive code sprint a compelling – and understandable! – event, even for first-time attendees discussing it in the hallway afterwards.
At Inspire, Alteryx announced the release of Alteryx Analytics 2018.2, highlighting features that emphasize making advanced analytics and data science capabilities accessible to all and straightforward to use:
- Analytic templates that cover functional analytic tasks as well as departmental- and industry-specific analytic tasks; these function as a shortcut to learn from or to construct your own analytic workflows.
- Global community search across Connect, Designer, and Promote will help users find the right data and answers more quickly, and reduce duplication of existing analytic assets and information.
- Support for Databricks via in-database connections as well as through the Apache Spark Code tool allows users to leverage the power of Spark in a Databricks cluster.
- Being able to leverage third-party Python libraries and Python code for model development – and even being able to drag and drop it into your workflow via a Python SDK.
- A more streamlined and personalized onboarding process (“analytic shepherding”) to enable end-users to more quickly learn how to conduct self-service data preparation and advanced analytics.
Alteryx already has an interface its users find easy to use – the addition of these templates and the enhancements to metadata collection and the search process will make building on existing analytics workflows even more simple. Being able to start from a template in a number of cases will help organizations standardize their analytics workflow creation process, make it easier to learn how to create workflows quickly for new users, and speed up the process for those already familiar with workflow construction.
What Alteryx Should Do Next
Now that Alteryx has expanded from a point solution to a more-complete suite, it’s all the “Alteryx Analytics” platform. It’s a suite of products, and making each product’s purpose clear is part of getting users to understand the platform as a whole, and the features they may not yet understand how to leverage. Their descriptions for Connect (“Discover and Collaborate”) and Promote (“Deploy and Manage”), the newer products, are fairly clear; their descriptions for the older Designer (“Prep + Analyze/Model”) and Server (“Share + Scale/Govern”) are still a bit vague for new users to grasp easily without seeing a task being performed, and slash together tasks that don’t overlap so well. From a marketing perspective, Alteryx should clarify and distill these descriptions.
The addition of Connect as a resource enhances end-users’ ability to quickly build and perform analytics tasks on their data, but it marks a shift in the natural starting point for longer-term Alteryx users. If a user starts to construct a workflow in Designer that looks similar to an existing template, whether Alteryx-provided or a custom template for their organization, a gentle nudge could help users take better advantage of the pre-built resources they already have on hand.
Finally, Alteryx identified the majority of its users as “citizen data scientists,” as distinguished from “data scientists.” The term is fairly common as a way to differentiate “coders” from “clickers” in a less-disparaging way, but “citizen” as a qualifier for “data scientist” doesn’t make it clear what either group does or can do. If Alteryx defines “data scientists” without the “citizen” qualifier as the deep specialists in coding complex machine learning models, then let’s “alter” the “citizen data scientist” term to one that better reflects the work they do. Earlier this year, they suggested that Alteryx users would “become the orchestrator directing disparate data, making it flow together and make sense.” I appreciate the level of complex coordination that “just works” implied in the “orchestrator” title reflected in the data tasks users are trying to accomplish when using Alteryx.
But overall, Alteryx holds two big advantages: their easy-to-use platform, and their community of enthusiastic and knowledgeable users that help Alteryx use spread quickly in the organization. Minor quibbles over clarifying terminology aside, the combination is a potent one to help Alteryx go “viral,” and their swiftly-expanding customer base demonstrates they understand the big picture – getting advanced analytic and data science work done quickly and easily.