Looker. The next generation of analytics. Are you ready for it?

Written By:
Content Copyright © 2018 Bloor. All Rights Reserved.

Each winter you get the reviews of the market by all of the big US-based analyst houses, which is of interest because so many technologies start off in the US and take quite some time to get to EMEA. Over the last few years, I have seen a steady increase in interest and respect for Looker. But Looker has had a very low profile in the UK and EMEA, so I have read from afar with growing interest. They set up a European base a couple of years ago, so I thought it was time to learn more.

From the Looker perspective, they see that we have had three waves of BI tools, something that I would agree with. This is important because the first wave was around for a long time, and is the most established in the market place. In the last five years, it has been to a large part supplanted in the minds of those at the more leading edge of the market, by second wave tools. Those second wave tools are probably still not making much of a profit after their various rounds of gaining venture capital, and they now are being challenged. It’s a tough market.

In the first wave of tools, we have the big beasts from the likes of SAP with Business Objects, Oracle with OBIEE and Hyperion, IBM with Cognos and TM1. Big monolithic stacks designed to work with traditional data warehouses, typically batch ETL tools loading a third normal form layer, which then is transformed into a presentation layer, and often moved to a data mart, with the BI tool sat on top of that stack. The whole design philosophy being dictated by the nature of the database technology, which dominated, being largely RDBMS, which was designed to support transactional data loads, to do the insert of records accurately and to do keyed reads. A few brave individualists went for columnar databases, but they were not the norm. Worse still those machines were often virtualised in data centres, and BI applications had to wait their turn for I/O execution. These applications are functionally rich, but in all surveys fair badly in customer satisfaction, and have very low rates of adoption within the business units, things tend to be done by IT for the business.

Then we had the second wave with the emphasis on visualisation, and self-service. This era has been dominated in the mind share by products like Tableau and Qlik and has resulted in the big vendors following with Microsoft BI etc. These were initially departmental tools, and they were supplemented by many other tools to try and match the rich functionality of the first wave of tools, which was a bit messy as integration even within a single vendors range has never been a strong point of the IT industry to date. So adoption rates were better as more users wanted to use them, but there was still not universal user satisfaction because of the issues of lack of some functionality.

Then, of course, the world started to change rapidly, and we had the explosion in the scale and speed that the Cloud brought; now everyone could access an MPP box, not just those with the deep pockets to afford a Teradata box. People now wanted self-service, they wanted to be able to embed BI into business workflows, they wanted things to scale to cope with the Internet of Things and the data in the Cloud, and they wanted speed, and all at a price they could afford. Enter Looker, a single integrated analytics platform, with no issues around backward compatibility with legacy solutions, designed to be part of the technology revolution heralded by the Cloud.

There are many BI users who are now on the cusp of their existing technology it is unable to match the requirements of the business. Those requirements are themselves changing with people starting to become aware that to succeed you now have to have analytics as a partner not as something tucked away in a back office. The most successful companies have developed a data culture, and are using facts to aid decisions, and not just at the Boardroom level but throughout the organisation. They have a choice you can stay with your incumbent and wait for them to bring to market tools that meet your need, which may be tricky because most of those companies are based on big margins, and they have a large installed base that does not want to be neglected, or you can make a change.

If you decide to make that change, then Looker has an awful lot to commend it.

Looker is designed to work with SQL compliant databases, it is an ideal match to the cloud databases like Big Query, and RedShift, but it can operate with existing SQL databases like Oracle. Looker is based on SQL, but it has extended the capabilities of SQL, with LookML, which allows reusable business logic to be created. It connects directly to the database it is not creating things like OLAP cubes. It has connectors and APIs to enable the data to be used to create scheduled output, to be fed to data-driven applications. It has the governance mechanisms and version control that was so difficult with the second wave of tool assemblies that people had to create to deliver insight in a controlled fashion.

A feature I want to highlight is Looker Blocks. Blocks are pre-built libraries that can be used to accelerate building a new application. They can be reused, and they can be customised. They address things like analytics patterns, source data extract, external data augmentation to your data, source data access, visualisations, embedding, and data tools. It is designed to enable self-service, more than that it encourages self-service, by creating robust (so no fear of breaking things), visually meaningful (so you have a clear picture of what it means), and tailored directly relevant output. People can answer their questions, with output that can be shared.

So there is a reason why Looker has appeared in so many of the BI reviews as a leading product, it is amazingly capable, it is an enterprise solution, it is designed to cope with the demands of the world of the Cloud, and it addresses the needs for self-service, scale, embedding in business workflows and speed. It is a thoroughly modern response to what is for many a pressing need; it’s hard to see what’s not to like! But for me, the clincher is how can you ignore a company that calls its Customer Support team the Department of Customer Love, has to be worth a serious look!