L'approche intuitive de l'exploration des données de Looker rend la diffusion d'une culture axée sur la curiosité à la fois efficace et efficiente. Une interface web intuitive permet aux utilisateurs d'exploiter le génie de leurs équipes d'analyse de données. Ils ont la possibilité de créer et de partager des rapports à la volée, afin que d'autres groupes fonctionnels puissent bénéficier des questions qu'ils posent et des connaissances qu'ils ont créées. Et puisque toute l'expérience de Looker donne des résultats aussi satisfaisants, ils commencent à se tourner vers les données pour guider leurs décisions et stimuler leurs activités.
Au cœur de la recherche de données (data discovery) de Looker, il existe une nouvelle façon de créer ces expériences web. En utilisant un langage flexible et facile à apprendre appelé LookML, vos analystes de données ont la possibilité de créer des mini-applications qui ajoutent de la puissance et de l'efficacité à l'exploration de données. La syntaxe LookML simplifie le développement de modèles solides pour toute entreprise, tout en améliorant les capacités du SQL. Il n'est pas nécessaire de consacrer du temps et des ressources au codage des requêtes de façon ad hoc ou à la mise en place de plateformes logicielles de Business Intelligence (BI) complexes. Cette approche basée sur le langage permet une exécution plus rapide des requêtes, optimisant ainsi les performances.
Like I said, it's a great tool for analysts that don't need a lot of flexibility or maybe just aren't used to working with a lot of data. To get the full value out of this software, you'll need a core group of experienced Data Analysts that can learn fast (to learn Lookers Table Calculation code which is similar to SQL) who can build a bunch of dashboards and views. You really don't want people who aren't savvy with data doing a lot of the creating because measures are always aggregated based on the dimensions you pull into each table...if a user forgets to add a specific dimension then measures could show the wrong value. This could be offset by having really good developers who architect the data in a way that doesn't allow this to happen.
It's a great data exploration tool for users who normally would get anxious when given a large dataset in something like Excel. It's generally simple to use and there can be unlimited Viewers (users at the lowest permissions level) through the enterprise product. Data Visualization is essentially done for you and Exploers (users at the second level of permissions) can build new visualizations & turn them into interactive dashboards.
On the Looker platform, Explorers can locate a view (called an Explore) and then are given a list of dimensions and measures they can pull into a table. Users can create new dimensions using calculated fields and then save their views so any users with Viewer permissions (and up) can see what was put together.
With simplicity comes lack of flexibility. More skilled data analysts will want more flexibility than Looker can provide but at the same time you need to have highly skilled Data Analysts to be the Developers (high level of permissions) to set up the data so that less skilled analysts (the users who benefit from Looker most) who are Explorers don't build views that grossly misrepresent the data.
My biggest pet peeve with this software as an Explorer is that you cannot make calculations on the base data...only the data that you pull into your table (called table calculations). That is a major feature that Tableau has that Looker does not have at the Explorer permission level. Of course developers can use LookML (Lookers data language) to create a new column...but a Developer license is more expensive and creating the fields take time and it's something you constantly will need.
When I first saw Looker (during a long BI evaluation process where we demo'd perhaps 5-10 tools) I knew it was what I had been hoping for in a BI vendor. The ease of dragging dimensions and metrics onto a canvas, the fluid way that you could generate reports, and the simplicity of drilling down into the source data all suggested this was a tool that had been built well from the ground up.
Other vendors seemed to trap you into a narrow view of your metrics, but with Looker the only restrictions were the ones built into your data model. It could easily flex to accommodate whatever data you wanted to look at.
The LookML language and the way Looker is designed to interact with your data is pretty brilliant. Even a non-data scientist and non-SQL user like me could get the hang of the basics ;)
The range of options, the configurability, the flexibility...all aces. Really you only get stuck by your imagination, the tool is powerful enough that you can solve most problems if you have the know-how.
The team is also great. Awesome culture, friendly support.
They have a continual flow of new features and improvements.
You do need a good BI person or ETL developer to really get the most out of this tool.
Setting up your data model really well and going to the outer limits to solve some particularly difficult queries are all things I would have had a hard time doing without technical resources on the team.
Of course it's possible Looker helps with a lot of this during implementation; we pretty much controlled our own because we had the folks in house to do it.
We get to force-multiply our impact by making it easy for users in all departments to make data-driven decisions and help allow that to drive their day-to-day.
I love the customization of Looker, and that you can put large amounts of data into the hands of anyone -- in a way that makes sense [to them]!
While their product is great, it speaks volumes to the team. It's an amazing group of people that continue to receive feedback from their customers and iterate time and time again on their product. They truly care about data and their customers, and they want to empower them the best they can.
When modeling views, I wish you could define better filters for measures but it's limited. There are workarounds, meaning modeling out several additional dimensions to arrive at the measure you need to pull regularly, but it'd be much more efficient to allow for measures to be more nuanced.
Large dashboards (with 25+ tiles) can take awhile to load, which is understandable. Once their "merge results" feature is fully rolled out and we're able to save them as new tiles or looks, it should help resolve the need for so many tiles.
Neither of these two things are a big deal, and I wouldn't let this deter you from considering this software.
It's a great product for giving people in all departments of your company access to data and allow them to explore what your data can provide. Rather than having to always go to the Business Intelligence or Analytics team, people of any level can query data without any SQL knowledge or understanding and yield results that are either directly actionable or ask more questions that can be answered using Looker. We ran an initial 30 days trial with a view that only our analysts and BI team members would use it. By the end of the 30 days, everyone from the COO down to the account teams were using it. It's amazing.
Furthermore, the vendor is smart, responsive and constantly improving the product, delivering a new version approximately every month. Their customer service is second-to-none and they are always happy to hear about new features that you need in the product. Very pleased overall with their service, strategy and plan to expand our use of them to our clients over the next few months.
Looker are still very heavily US based. They have just started an EU office, but support is still largely limited to US West Coast hours (although they seem to wake up at stupid o'clock to deal with some of our issues and plan to have support in the EU soon).
Our business users love the self-serv options that this provides. It drastically cut down the effort our analysts have to spend answering routine questions and allow them to focus more on the value added activities.
Things I like about this product:
1. Ease of use by our users - very intutive
2. Integration with our big data platform and databases. We use ours with Redshift and Aurora and they take most advantage of both systems.
3. Their super awesome customer support - always available chat who are really good at what they are doing.
4. Custom install options along with their hosted solutions.
5. Their whole team is involved with the product and customers top down. Really appreciate their responsiveness and customer-centric approach.
We really don't care about visualizations, but that can use some improvement to those who care.
The LDAP integration setup is a pain. It works well once it's all setup.
There is no roll-back once a version is installed. Work around is to restore from a backup. Would love to be able to do that as sometimes things don't always work the way we are used to and might require a rollback.
Voici quelques-unes des questions fréquentes sur Looker.
Types de licences disponibles pour Looker :
Type de licence: Abonnement
version d'essai gratuite: Disponible
Nous n'avons pas d'informations sur les fonctionnalités de Looker.
Utilisateurs habituels du logiciel Looker :
Grandes entreprises, Entreprises de taille moyenne, Organismes à but non lucratif, PME
Langues dans lesquelles Looker est disponible :
Types de licences disponibles pour Looker:
Nous n'avons pas d'informations sur les appareils pris en charge par Looker.
Applications s'intégrant à Looker :
Alooma, Dropbox, HipChat, Improvado, Kustomer, Marketo, Panoply, Twitter, Workable, Xplenty
Ressources d'aide disponibles pour Looker :
FAQ, Forum, Base de connaissances, Support en ligne, Tutoriels vidéo