Today's Deep-Dive: Metabase

All right. Kicking off another deep dive and today we're taking a look at Metabase.

You know, you sent over their GitHub repo and the website,

definitely two sides of the same coin, right? So let's, uh, you know,

see what kind of picture they paint for themselves.

Is it truly a data tool for as they say, everyone,

or is that just some marketing hype?

That's the question, right? You got the kind of like under the hood view on GitHub.

That's where the devs live. And then you have the polished, you know,

marketing side on the website. So let's see how those line up. Okay.

So the GitHub repo first thing hits us with this tagline,

the simplest fastest way to get business intelligence and analytics to everyone

in your company. And I got to say, I love the little emoji, you know,

they're trying to be our data BFF, it seems like, but simplest and fastest.

That's a, that's kind of fighting words when it comes to the world of data,

for sure. You know, usually you get one or the other, right? Not both. Yeah.

And then there's that, you know, the everyone claim, like that's the holy grail

of business intelligence,

getting insights out of just the analyst's hands and into, you know,

the folks making those daily decisions. That's a big claim.

So the website better back that up. Let's see. Oh, here we go.

Set up in five minutes. Kim on really,

I timed myself making coffee this morning and it took longer than that.

I think they're serious about that. The installation is like super streamlined.

There's no, no messing around with servers or like crazy configurations.

Especially if you go to the cloud route, they want to get you up and running,

like ASA. Okay. Intrigued, but color me skeptical as always,

you know, got to see the proof, but this next part definitely caught my eye.

Ask questions without knowing SQL. Finally, someone gets me. SQL's powerful. Sure.

But sometimes I just want to point at my data and just yell,

tell me what's going on. Well, it looks like they actually let you do that.

The visual query builder. It's like a whole different ball game.

Imagine you just drag and drop the fields you want, pick your chart or analysis.

And boom, you've got insights without writing a single line of code. Okay.

That sounds dangerous. Tempting how suddenly I'm picturing our marketing team

actually like understanding our customer data without like a million slack

messages to me. All right. All right. Color me cautiously optimistic on this one,

but you know, what about when you do need to get down and dirty with the SQL?

Can I handle the heavy lifting too? Oh yeah, definitely.

They've got a full on SQL editor for those who love to write those custom

queries and really get into the nitty gritty.

So it caters to both ends of the spectrum,

the SQL newbies and the seasoned data rangers.

That's good to know. It means you're not sacrificing power for that ease of use.

But speaking of power, what about the types of data it can handle?

Is it just for those trendy new databases or can it connect to the, you know,

legacy systems that I feel like so many companies are still stuck with?

That's where it gets really interesting.

They've got a seriously long list of supported databases.

You've got your usual suspects, of course, like MySQL and Postgres,

but they also play nice with those big data warehouses like Redshift and

Snowflake. Oh, and a bunch of those specialized databases,

you might not even expect.

So whether you're like a startup on cutting edge tech or a company with data

just spread across a patchwork of systems, Metabase is saying, Hey,

we can connect to it.

I'm starting to see why they're making these bold claims about reaching everyone

in the company, but talk is cheap, right? Sure.

Do they have any like real world examples to show this thing actually works?

Oh, they do. They've got testimonials on their website from, you know,

companies of all sizes, different industries. Take Holland and Barrett,

for example, that they're a big global health and wellness retailer.

And they were, you know,

struggling to get data into the hands of their business users who weren't

analysts. With Metabase,

they finally empowered those non-technical teams to build their own dashboards

and answer their own questions.

So they actually walk the walk on that everyone promise. I'm impressed,

but let's be real, you know,

a shiny website and some nice words can only get you so far needed like dig

deeper, see what's really going on, like under the hood with those features.

Agreed. Let's start with that thing that caught your eye,

that visual query builder. It's not just, you know, drag and drop.

It's about fundamentally changing how people interact with data.

Like imagine you're a marketing manager and you want to see which campaigns are

driving the most sales.

You can just drag the campaign name field in the sales field right into the

builder, pick a bar chart and boom,

in seconds you've got a clear picture of what's working and what's not.

So no more waiting for the data team to get back to you with that report. Okay.

That's, that's a game changer.

But what about the folks who are comfortable with SQL?

I don't want them to feel like they're being like pushed into a simplified tool.

Don't worry. They haven't been forgotten. Like I said,

they've got that full SQL editor. So if you need to write complex queries,

join multiple tables or use those advanced functions,

you've got all that SQL power. And here's where it gets really interesting.

They let you switch back and forth between the visual builder and that SQL

editor seamlessly.

So you could start with a simple drag and drop query.

And then if you need to refine it with them custom SQL,

you can do that right there in the same interface.

That's pretty slick. Okay.

I'm definitely liking what I'm seeing so far,

but I think we're just scratching the surface here.

Tell me more about these models that keep popping up. They sound intriguing,

but also kind of mysterious.

Models they're essentially a way to create a simplified like curated view of

your data. They're really helpful for those non-technical users who, you know,

might not know all the ins and outs of how your database is structured.

Imagine you've got data spread across, you know,

tons of tables with like these complicated relationships and cryptic names.

A model can take all that complexity and present it in this neat and

understandable format.

So it's like creating a user friendly map of your data with clear labels and

directions.

I can see how that would be helpful for someone who just wants to get to those

insights without getting lost in the, you know, the data jungle.

But what about security? If we're making this data accessible to everyone,

how do we make sure that people are only seeing what they're supposed to see?

That's where sandboxing comes in.

It gives you that granular control over data access.

You can create different sandboxes for different teams or departments.

That way each group only sees data that's relevant to them.

So for example,

we could have a sandbox for the marketing team that only lets them see marketing

data while the finance team has their own sandbox with the financial data.

Exactly. You can even get more granular than that.

You can control access at the row level.

So you could restrict access to certain types of customers or even transactions

within a data set.

Okay. That's reassuring. They're not just, you know,

throwing the data doors wide open and hoping for the best.

There's some real thought put into that security and governance.

Absolutely. And that level of control is crucial when you're talking about

democratizing data. You know,

you want to give people access to information,

but you've got to do it responsibly and securely.

I'm sensing a theme here.

Metabase is all about that balance between power and simplicity, accessibility,

security, you know,

they want to make data exploration easy and intuitive for everyone,

but they're not skimping on the features and controls that those data pros

need. Yeah. You hit the nail on the head.

And that's what makes Metabase so interesting.

It's not just for one type of user.

It's designed to be that data hub for the entire company.

And we're just getting started. There is so much more to unpack about Metabase,

but let's take a quick pause here. And when we come back,

we'll dive even deeper into some of those more advanced features. Welcome back.

We're a, we're knee deep in Metabase now,

trying to see if it lives up to those ambitious claims.

And we, we left off talking about those models. Remember those things that

can supposedly make a messy data landscape. Well, less messy.

Yeah. You made them sound almost too good to be true.

Like a magic data fairy that just swoops in and cleans up all your messy tables

and cryptic column names and stuff. But like, how does that actually work?

Like in the real world? All right. Let's say your company,

they've got data about customer orders spread across a bunch of different tables.

Right. You've got one for customer info, one for order details,

another for shipping, so on and so forth.

A model can basically stitch all those tables together and present it as one

unified view.

So instead of having to like navigate this tangled web of tables and try to

figure out how they're all related, you get a nice clean interface that makes

sense, even if you're not, you know, a database guru or whatever.

Exactly. And you can define custom fields within the model, you know,

combining data from multiple sources. For example,

you could create a field called like total order value,

and that combines say product price, quantity, and any discounts.

That way anyone looking at the model, they can instantly see the total value of

each order without having to, you know, break out the calculator.

That's, that's really smart. Those pie in the sky promises about, you know,

empowering non-technical users. It's starting to feel a lot more real now,

but models, they sound like they could be helpful even for like the season data

folks, right? Absolutely. Even if you're a SQL whiz,

models can save you a ton of time and effort.

Imagine you're building a dashboard to track sales performance, right?

Instead of writing all those crazy queries to join tables and calculate those

metrics, you can just use those predefined fields in the model and just focus on

the visual presentation analysis much easier.

So it's like having a whole team of data engineers working behind the scenes to

like organize and optimize your data. Even if you're a one man show,

I like the sound of that, but let's switch gears for a second.

Let's talk about something that's often,

I feel like an afterthought in data tools documentation.

How does Metabase handle that?

Surprisingly well, actually every model, each metric you define,

it all has that built in space for documentation. You're not just throwing,

you know, raw data at people. You can actually explain what it means,

where it came from, any caveats they should be aware of.

So no more like cryptic column names that leave everyone scratching their

heads. That alone could save countless hours of frustration and,

uh, and misinterpretations. Okay.

I'm really starting to see how all these, you know, seemingly small features,

can add up to a huge difference in how a company uses data,

but we haven't even touched on one of,

I think the most exciting aspects of Metabase embedded analytics.

Tell me more about that. It sounds like something straight out of a,

like a sci-fi movie.

It's not quite telepathy, but it's pretty close. Imagine this,

you're a sales manager, right?

And you're using your company's CRM to track, you know, your team's progress.

Normally you'd have to switch to a separate like BI tool to get insights into

sales performance, customer trends and all that. But with embedded analytics,

you can have all that data displayed right there in the CRM.

No switching back and forth.

So you're not just looking at a list of contacts and deals.

You're actually getting real time insights into what's working, what's not,

where you should be focusing your efforts.

Exactly. And it's not just dashboards, you know,

you can embed interactive charts, reports,

even like full blown data exploration tools right into your apps.

And that's what I call like actionable intelligence.

No more switching between tools,

no more exporting data and trying to make sense of it in spreadsheets.

You get the insights right where you need them.

But I can also see how this could be like a slippery slope, right?

Too much data. It can be overwhelming,

especially for folks who aren't used to working with it all the time.

So how do you like, how do you strike that balance?

It's all about being like strategic about what data you embed and how you

present it. The goal is to make the user experience better,

not just bombard people with information.

Think about those key questions that your users are trying to answer in a

specific app.

And then you design the embedded analytics to give them those answers clearly

and concisely.

So it's like a choose your own adventure for data where you can tailor the

experience for different users and their needs. I like that.

But I'm also curious about the technical side.

How easy is it to actually embed Metabase into, you know, my own applications?

That's another area where they've done a great job.

They give you a full API and have a bunch of pre-built components that make

embedding super easy. You don't have to be, you know,

a coding genius to get it working.

So even if you're not a developer, you can still take advantage of this,

this powerful feature, huh? That's impressive. Okay. We've, uh,

we've covered a lot here, but there's,

there's one big elephant in the room we haven't really talked about.

Metabase is open source.

What does that actually mean for someone who's thinking about using it?

That's a great question. And it's one that trips a lot of people up. Let's, uh,

let's break down open source a bit.

Basically it means that the code that makes Metabase tick,

it's publicly available. Anyone can download it, look at it, even modify it.

So you're not like locked into a black box where you have no idea what's

happening behind the scenes.

Exactly that transparency. That's one of the big perks of open source.

You can see exactly how your data is being handled, which is, you know,

super important for companies that have, you know,

strict security or compliance rules. And because the code is open,

there's this huge community of developers constantly working on it, you know,

improving it, adding new features. It's constantly evolving.

So it's like having this global team of data experts,

like working on your BI platform for free. That sounds, that sounds amazing.

But what about, what about support? You know, if something breaks,

who do you call? Ghostbusters.

That's, that's a common worry with open source. I get it.

But Metabase, they've got this really active community forum where you can get

help from other users. And they also have, you know,

those paid support plans for companies who need like a more formal level of

help. So it's like, uh, it's like choosing your own adventure. Again,

you can go the DIY route, you know,

rely on the community or you can pay for that peace of mind that comes with that,

that dedicated support. Precisely.

It's all about finding that level of support that works for you,

your needs and your budget. Okay. I'm, I'm getting the open source picture now,

but there's still one piece missing licensing.

You mentioned different licenses earlier. What's the deal with that?

So the core of Metabase, it's released under the AGPL license.

It's a type of open source license that, you know,

it encourages sharing and collaboration. Basically,

if you modify the Metabase code and use it in your products,

you're required to make those changes public as well.

Makes sense. You know, keeping with that open source spirit,

sharing is caring and all that.

But what about companies that want to keep those customizations private?

Is that even possible?

That's where those commercial licenses come in.

They give you more flexibility, more control. So for example,

you might want to, you know,

integrate Metabase into your own software without having to make those

modifications public. Makes sense.

So the open source license, it's great for that transparency,

community involvement, but the commercial licenses,

they provide that flexibility that some companies need.

It sounds like they're trying to, you know, appeal to a wide range of use cases

and philosophies, but enough about, you know, legal stuff.

Let's get back to the features.

You said before that Metabase supports a ton of different databases.

Can you give me a sense of just how broad that support is?

Yeah, they're not kidding about that wide range. You've got your big names,

of course, like, you know, Postgresql, MySQL, SQL Server.

Then you've got the cloud-based data warehouses like Snowflake and BigQuery,

but it's not just SQL databases either. They also support NoSQL databases.

Like MongoDB and Cassandra. Oh, and those specialized databases,

like Elasticsearch and Redis. It's a lot.

Wow. That's, that's a mouthful.

So whether you're a tiny startup with a simple database or like a Fortune 500

company with data just sprawled out everywhere, Metabase has got you covered.

That's the idea. And it's not just about, you know, connecting to those databases.

It's about making it easy to work with data from all those different sources

within like the same platform.

Okay. I'm starting to get a headache just thinking about all the possibilities,

but that's a good problem to have, right?

It means you're not limited by the tools. You're only limited by, well,

your imagination, but let's not get too carried away. There's,

there's still more to explore,

especially when it comes to the like the nitty gritty details of like deploying

and using Metabase. Let's take a quick break. And when we come back, we'll,

we'll dive into all those practical details.

Welcome back to the deep dive. We've,

we've really gone in depth with Metabase from its user friendly query builder to

its ability to, you know, talk to a whole bunch of different databases.

And we touched on that open source aspect of the platform, right?

Which can be a real plus,

but also brings up those questions about like support and licensing.

Exactly. So let's tackle the, the elephant in the room.

The question is probably on everyone's mind.

Like how do you actually get started with Metabase? Where do you even begin?

Well, that's where it gets pretty interesting. They've got options.

And I mean options for everyone from free and DIY all the way up to, you know,

fully supported and managed. Let's start with the,

the option that everyone loves to hear free.

Ah, yes. The magic word free,

but free often, you know, comes with strings attached.

Is this a limited version or like a trial that expires or you know, what's,

what's the catch? Nope, no catch. The open source version of Metabase.

It's like totally functional.

It has all those core features we've been talking about. You can download it,

install it on your own servers and start exploring your data. No limits.

So I could like literally go to their website right now,

grab that open source version and like be building dashboards by lunchtime.

Yeah, pretty much. It's that simple. Of course you'll need some, you know,

technical skills to set it up and manage it yourself,

but their documentation is like really good. It'll walk you through it.

Okay. That's impressive. But let's say I'm not a, you know, a tech wizard,

or I just, I don't want the headache of managing servers and updates and all

that. What are my, what are my options then?

Well then you've got Meta-based cloud. That's their, you know,

fully managed cloud based option. Basically you sign up,

pick a plan and they handle the rest hosting security maintenance,

backups, you name it.

Sounds like, uh, the, the set it and forget it solution,

but I'm guessing that that comes with a price tag.

It does. They've got different tiers based on, you know, what you need,

how many users you have, how much data you're working with,

the level of support you want.

So you're basically paying for that convenience and peace of mind. Exactly.

And for a lot of companies, that's, that's worth it.

It lets them focus on the important stuff, you know,

getting insights from their data and making better decisions.

That makes sense.

But let's not forget about those paid support plans you mentioned before for,

for that self hosted version, right? What kind of like handholding did they offer?

Their support plans,

they range from like basic email support to dedicated account managers and

super fast response times.

They also do like training and consulting to really help you get the most out of

Metabase.

So they've really thought of everything, haven't they? Whether you're, you know,

a solo data nerd running Metabase on your laptop or you're like a huge company

with a whole data team. They've got something that fits.

And that brings us right back to their, like their core mission,

making data accessible to everyone, you know,

breaking down those barriers that have always prevented people from really

exploring and understanding their data. Yeah.

It's not just about building a fancy tool. It's about, you know,

democratizing data and giving people the power to ask questions, get answers,

and make those data driven decisions. Exactly.

And that's what's so exciting about Metabase. It's not just another BI tool.

It's a platform that could really change how companies use data. Okay.

So we've covered a lot today. Let's, uh,

let's wrap things up with those key takeaways. What have we learned?

So first off Metabase is like super user friendly,

even if you're not a, you know, a data expert, that visual query builder,

it's, it's a game changer. And those models, don't forget those.

They're like having your own personal data organizer. Right. Second,

Metabase is, it's more powerful than you might think.

It can handle complex queries, connect to like tons of different databases,

and even be embedded like we talked about right into your own applications.

Third, and this is a big one, it's open source.

So that means transparency, flexibility, and you know,

a huge and active community of developers constantly improving it.

And lastly, they offer like, you know,

different deployment and support options to fit any need, any budget.

So if you're looking for a way to really unlock the power of data in your

company, I think, you know, Metabase is definitely worth a look.

Head over to their website, try out the open source version, or, you know,

sign up for a free trial of their cloud offering. It's worth it.

You might just be surprised at what you find.

Happy analyzing.

Happy analyzing.