Looker held a webinar today. I recently blogged about their presentation to the BBBT community, but it’s an interesting company so was worth another visit. The company is a business intelligence (BI) firm. With the presenters being Colin Zima and Zach Taylor, the presentation stayed at a much higher level than the previous presentation and was aimed at a business audience rather than analysts. It is always good to see a different view of things.
The focus of their presentation is why it’s good to embed BI in other applications in opposition to pure BI tools. It’s a good message but needs to be strengthened. Colin and Zach quickly mentioned embedding as if everyone understood it, then dove into the issues in evaluation the build v buy decision. They should have spent a couple of minutes explaining what they mean by embedding and their focus on what they focus on as places to be embedded into.
Their build v buy decision discussion was standard and hit all the right points about letting companies focus on their competencies and leverage the BI industry’s competencies for analysis. Where embedding and build v buy really blend, and they could have hit harder, is the difference in ROI between embedding and having a separate BI visualization tool.
They did have a couple of case studies that were interesting. Ibotta is a company providing analytics to their consumer packages goods clients. That’s a great application and a powerful use of BI in a business network, but I didn’t see much on what it was embedded into or how. That meant it didn’t fit into the overall scheme of the presentation.
The other key one was HubSpot using Looker to provide analytics to sales on sales performance. That’s done by embedding the analytics directly into the normal Saleforce.com windows the sales team see every day. That’s a powerful message and one that I felt deserved a bit more time.
The only questionable message I heard was during Q&A, when somebody asked about their performance issues. As in the previous presentation, they talked about using the source data and not replicating for BI. They therefore said they didn’t have performance issues when scaling users but it was one for the databases. Well, that’s not quite true.
It’s not likely that all a company’s various data sources have been built to scale to lots of users. Companies will still use ODS’s, data warehouses and other methods to parallel data and have multiple versions of the truth which require strong compliance to control. Companies will still have to spend time to analyze and prepare appropriate data sources that can handle large numbers of concurrent users. The advantage of Looker is not that it means that you don’t have to add to the confusion to get performance, whatever is provided to get good performance for Looker isn’t unique and limited to it but can serve other applications as well.
Looker is that rare young company that seems to not only have a good early generation product, but understands how to market their product to multiple audiences. As someone focused on software marketing, I think that’s great.