- Mode 1 focuses on predictability and has a goal of stability. It is best used where requirements are well-understood in advance and can be identified by a process of analysis....
- Mode 2 is exploratory, involving experimentation to solve new problems, and optimized for areas of uncertainty. In this case, requirements are not well understood in advance. Mode 2 is best-suited for areas in which an organization cannot make an accurate and detailed predefined plan because not enough is known....
I continue to think the demarcation lines are somewhat blurry, and that this blurriness isn't a bad thing. Gartner even says, "The ideal for customers is to have both Mode 1 and Mode 2 capabilities in a single platform, with interoperability and promotion between the two modes."
This implies, to me, that Mode 1 and Mode 2 shouldn't really be associated with "traditional enterprise reporting platforms" and "modern analytics and BI platforms," respectively. And, in fact, we see Gartner call out Mode 1 and Mode 2 in several of the company write-ups.
The emphasis on the two modes also suggests to me that we will continue to see convergence in these platforms -- a trend that stands out when we look at the 2016, 2017, and 2018 Magic Quadrants together.
Here are three examples of where this seems to play out in the report.
- One of the six Strategic Planning Assumptions is, "By 2020, organizations that offer users access to a curated catalog of internal and external data will derive twice as much business value from analytics investments as those that do not."
- The Market Definition/Description section notes that "Modern analytics and BI platforms may optionally source from traditional IT-modeled data structures to promote governance and reusability" (emphasis in original). Later, in the Market Overview, they say, "The rise in data lakes, as part of the overall information architecture, forces analytics and BI teams to decide how best to model the data and where," with further discussion of data replication and model complexity.
- In one of the vendor write-ups, Gartner commented that complex data modeling needed to be created outside of the tool in question, perhaps in a data warehouse.
In all three cases, there's a need to create a repository of data in which people can place their trust. As we've seen in this industry for decades, that's not typically something that lies in the skill sets of business users: They need help acquiring the data, understanding how to create more complex models, defining the rules by which that data is prepared and cleansed, and (once the data's in place) doing any joins and blending that aren't pretty straightforward.
Moreover, there's a need to promote insights and content that have been generated by businesspeople to the rest of their organization -- or even to customers and business partners. This "hardening" process requires IT to take something that has been done on a single-user basis and bring it to hundreds of people or more. (We can see the need for this in the many references Gartner makes to scalability and large user deployments.)
Finally, from my perspective, it's worth noting that the creation of InfoApps -- highly interactive analytical applications, often with millions or billions of permutations of the data on a single screen -- serves the same purpose as creating the trusted data stores that Gartner talks about in these pieces of the report.
None of this is to say that self-service tools for data-savvy business users ("mode 2" tools) aren't important. They are, and Gartner does a good job of covering that terrain in this report. But I think the "traditional" to "modern" shift was less of a permanent change and more of a pendulum swing, similar to what we saw with reporting giving way to OLAP giving way to business intelligence back in the 1990s. As long as we're constantly moving forward as an industry, that pendulum swing will no doubt be a feature of most long-term conversations.
There are other interesting things to note. Here are three.
Gartner has changed its terminology from "smart" to "augmented," reflecting the fact that analytics and BI tools should be smart specifically in the way they help people uncover, identify, analyze, and promote insights. "Smart" started to seem a bit like getting a magic bullet; "augmented" grounds us in the reality that machine learning, search, and other forms of machine-driven intelligence will help us become more analytical rather than creating the infrastructure for Skynet.
Cloud is "past the tipping point," with most net new deployments originating in the cloud.
They've started to include social responsibility and workforce diversity as part of vendors' "Completeness of Vision" score.
There are twenty vendors in this quadrant. We're not sponsoring the report this year, but if you'd like to see a bundle of additional analyst opinions from other firms, please feel free to check out our analyst bundle -- and tell us your take in the comments section.