Looker is a well-known analytics solution that lets you model your data and create analytical interfaces for non-technical people. It uses an in-house language called LookML so you're expected to learn this language to be able to use it. Here is how it works:
- You connect to your database from Looker.
- Learn and adopt LookML or get their help which costs around $10K - $15K for a typical setup. You usually need at least one engineer since SQL is often not enough to learn the language if you don't have a technical background.
- Create LookML specifications and make Looker learn about your data layout. Then, you can create reports (Explores or Dashboards in their term) for different departments in your organization.
- The non-technical teams in your organization can use their "Explore"s, select dimension, pivot, measure, filter their data, see it in a chart without writing any SQL code so that they can ask questions in an ad-hoc manner.
- You also need to maintain the Looker specification which requires you to dedicate at least one technical person in your organization.
Looker is a big step in terms of empowering non-technical teams to ask more questions. However; as your data structure and the reporting needs become more complex, it fails to provide a seamless experience. We have common features and use-cases but there are a few differences that make us unique:
Looker has only one reporting feature which is called Explore. In Rakam, we have Segmentation which has the same features to Looker's Explore but in addition to that, we have Funnel, Retention and SQL reporting capabilities. Rakam mainly targets product data, you can model your event data in Rakam and seamlessly analyze the customer behavior. While Looker has a SQL Runner feature, our SQL reporting feature lets you reference variables in your SQL queries and create reports similar to Segmentation reports so that you can add them to your dashboards.
You can model your data programatically as SQL at rakam. If you're using dbt (Open-source data transformation tool), you can integrate dbt project with rakam to automatically create datasets from your dbt resources. If you're using one of the supported analytics solutions like Firebase, Google Analytics, Segment, etc. to collect the data into your data warehouse, you can install the recipes, pre-defined datasets, for the solutions and start analyzing your data without any extra modeling.
Looker uses their in-house IDE for LookML while we make use of VSCode and develop an extension for our data modeling language. You can develop your data modeling in your local environment, have access to all the advanced IDE features of VSCode which is being used by millions of developers globally.
Your materialization engine is based on the open-source ELT tool called dbt while Looker developed its own materialization engine which is called derived tables. dbt has many advantages over Looker's derived tables: a. It supports incremental materialized views so that you can update the materialized tables periodically for the time-series data. It's a deal-breaker in event modeling. b. You wouldn't want to rely on your BI tool for data transformation since you may be using the summary tables for other use-cases as well. You can start using dbt with Rakam and export your dbt project, maintain them yourself or use dbt Cloud that lets you create dependency graphs from your ELT workflow. c. dbt support Jinja thus provides a way for you to write much concise and readable data models compared to Looker's Liquid templating.
We have the Recipe feature that lets us parametrize the data models. It lets you use Rakam if you're using one of the solutions that we support without any learning curve. Looker has an alternative which is called Blocks but the integration is not seamless, you still need to deal with LookML which is not something small companies can afford.
Having said that, being a small company requires us to stay focused. It wouldn't be fair to tell you that Rakam is better than Looker in general. We focus on a few use-cases where Looker can't provide a seamless experience by design.
Rakam is relatively new compared to LookML so the community is not big enough. We have a Slack channel and the founding team is mostly online though.
We don't have an alternative for Looker Actions yet. We plan to implement it in Q3 2020. (will be open-sourced)
We do not provide a solution for embedded analytics yet.
New: We have developed a migration tool for you to convert your LookML files into Rakam Models so that you can try Rakam without too much afford. Feel free to contact us if you would like to try it out.