Data Analysis Teams with Data Synthesis to Navigate a Shaken Economy

Ted Cuzzillo,
Principal Analyst – Data Analytics

Summary Bullets:

• Analysis isn’t enough anymore. To be truly data driven, organizations also need synthesis.

• The analysis-synthesis duo can thrive if supported by a handful of conditions and practices within the organization.

The recent QlikWorld Online conference came and went with no new public roadmap, but it did offer something better: an intriguing vision of an analytics trend that’s built on a handful of new requirements — which altogether stand as a row of streetlights for the analytics industry to portend a new road.

In “2020 Trends: Analytics Alone is No Longer Enough,” Qlik senior director of global market intelligence Dan Sommer argued that today analytics needs its other half, synthesis. This seems especially relevant and timely for the new pandemic-ravaged world.

Analysis helps understand one piece of an operation at a time. It’s useful for navigating a known environment, such as one department or one function, or even whole parts of the pre-COVID world. But analysis tends to falter in an uncharted or fast changing reality. As Sommer puts it, analysis is “micro smart but macro dumb.”

Synthesis, however, is macro smart. It’s best in an unknown, even chaotic world. It reveals relationships and commonality among disparate parts to show how they relate and form a new whole. That can launch techniques like scenario planning, in which multiple futures are imagined and planned for.

Clearly, synthesis has a few requirements that analysis doesn’t. While analysis can look closely at one operation, synthesis can examines examine several at once, including:

• Synthesis benefits from real-time data to tie pieces together. Analysis can help improve a department’s efficiency, for example, while synthesis can help that department work among other departments and the organization’s environment.

• Synthesis also benefits from widespread use of data much more than analysis does. Several trends will improve accessibility.

• Metadata catalogs, made possible by machine learning, helphelps make all disparate data sets accessible across an enterprise and the organization’s environment.

• Dataops DataOps with self-service analytics speeds the data pipeline as it manages the data. It automates data testing and deployment in real time, and it at least partly eliminates self-service data preparation.

• Using AI and machine learning to make data instantly identified, evaluated for quality, made available, viewed, analyzed, and discussed. Sommer’s name for this is “Shazam your data” after the smartphone app that identifies a song based on a small sample.

• Data intelligence has to infuse the organization. That requires enterprise-wide data literacy. Literacy is the ability to read, work with, analyze, and argue with data. The now legendary standard of about 35% literacy just isn’t adequate. However, organizations have found themselves unable to fix the problem, so they need professional help.

Then there’s ensuring an ample supply of data. Trust is a big factor.

• Ethics and responsible computing. If data analysis depends on a plentiful and constant supply of data from a wide variety of sources, no data user can risk disrupting that flow by violating the confidence of suppliers. Ethics and social responsibility is required to ensure their enduring trust.

Of course, all these conditions map to Qlik’s capabilities, either current, emerging, or expected. Other vendors are on to these, too. It’s up to forward looking leadership to make sure the organization pays attention and actually builds in each of these conditions for the analysis-synthesis duo to do their work.