Segment Analytics: Pros and Cons of Using It With Mixpanel & Other Tools

There’s a special analytics tool out there that is generating a lot of buzz and questions. You probably have seen it or heard of it. I’m of course referring to Segment.com (or Segment Analytics).

Almost every company is trying to understand how this tool fits into their analytics strategy and how to use it with other tools like Mixpanel and Google Analytics.

I’m hearing the same questions over and over again including things like:

• How does Segment work with Mixpanel (or any other tool)?
• Should I send the data to Segment and then to Intercom?
• What happens if I want to load the Facebook Marketing pixel into my app? Can Segment help me with that?
• The support team at tool X said that I shouldn’t use Segment.com. Are they correct?

You can also see similar questions over at Quora, just in slightly different words. Companies are trying to understand why they need Segment and how to implement it properly.

By the way, Segment.com isn’t the only player in this space. You also have mParticle and Keen.io. You could even build an internal tool that does the same thing as these third-party tools.

Let me demystify the role that these tools play by helping you understand the short and long term challenges that these tools are helping you solve.

*Spoiler alert*: By far and large, using tools like Segment alongside your other analytics tool is a good idea.

We first need to start by understanding how a “hub of analytics” can help your company and what you need to set this up.

Understanding Why You Need a “Hub of Analytics” Inside Your Company

Throughout the remainder of this article, I will use Segment in most of my examples simply because they are the most well-known tool in their space and because I think they have a solid product.

Segment.com describes themselves as:

“Segment is a single platform that collects, stores, and routes your customer data to hundreds of tools with the flick of a switch. We take care of the messy analytics installation process for you, so you can spend more time using your data and less time tracking it down.” – Source

You might notice that they used the word “customer data” instead of analytics data (or maybe you didn’t). Either way, this is a pretty accurate description of what companies these days are experiencing.

Analytics data isn’t just limited to what dedicated “analytics tools” like Google Analytics, Mixpanel and Kissmetrics offer you.

You can now find useful data in almost every single marketing tool. Your Mailchimp (email marketing) has metrics on your email campaigns, your Salesforce (CRM) has metrics on sales performance and your Zendesk (customer support) has metrics on customer happiness.

Your average customer will interact with dozens of tools like this, each collecting something valuable. The problem is that all of your data is now stuck in silos and getting out can be a pain.

Segment.com Becomes Your Hub

This is where Segment.com comes in (and tools like them).

They become the “hub” through which all data flows through. They help you tackle two things:

1. They make it easy to push data to tools like Mixpanel and Kissmetrics.
2. They make it easy to pull data from sources like Mailchimp or Salesforce

The above two things might seem simple but they usually require hours of programming simply to get data out of one service like MailChimp. Instead, Segment.com would let you pull this data in a few seconds.

The need for this data isn’t sometimes obvious. After all, companies just want answers to simple questions. How hard could that be?

Answering Simple Questions like “What is Our CAC (Cost of Acquisition)?”

A “simple question” like “What is our CAC (Cost of Acquisition)?” turns out to be more complex than you think. After all, CAC is basically:

Total Marketing Spend / Total New Customers = CAC

To answer this question, you may have to pull data from tools like:

• Facebook Ads (part of Total Marketing Spend)
• Google Adwords (part of Total Marketing Spend)
• Mixpanel (for Total New Customers)
• Stripe (for a more accurate version of Total New Customers)

All of sudden, you need 4-5 sources of data to get one “simple number. Are you able to quickly pull this data into one single dashboard or report?

This is where a “hub” comes in. A hub helps you collect all of this data and store it in one place for easy access.

Segment.com doesn’t just make it easy to answer important questions like the one above. It makes your data portable which is a crucial aspect of a long term analytics strategy.

Making Your Data Portable and Moving it to Wherever You Want

One of the things that companies fail to consider is the portability of their data. After all, why should you worry about this?

Data portability becomes important the moment you decide to switch tools. You realize that all of your historical data is stuck in tool X and moving to tool Y means losing all of that data.

Let’s imagine a common scenario that you’re almost guaranteed to run into at some point:

You decide to set up Amplitude and track key activities within your app. You use Amplitude for 6 months before realizing that Mixpanel might be better suited for your app.

You then realize that moving to Mixpanel means losing the past 6 months of data and you would have to start from scratch. All of your reports like retention, engagement, etc would take weeks to collect enough data for them to be useful.

You could export the data from the Amplitude API and import it to Mixpanel. However, this means writing custom scripts to work with both APIs and these can get messy.

You will also have to rewrite your Amplitude event tracking code into the correct Mixpanel structure. Worse of all, this could happen to you again if you decide to change tools in the future.

This is where data portability comes into play.

Why You Should Keep a Full History of Your Data

Tools like Segment.com (which act as your hub) make it easy to store all of your data in a database that you control.

Segment.com calls this Warehouses which supports Amazon Redshift, and Postgres. As data flows through Segment, it gets sent to tools like Mixpanel/Amplitude AND it also gets sent to your custom database.

Moving from Amplitude to Mixpanel becomes easier because you now can easily import historical data from your custom database (which has your entire data history) into Mixpanel (Segment will recognize timestamps for past events.).

Related: We have created a free video course that will teach you everything you need to know about Amplitude. You can view the Definitive Guide on Amplitude Analytics. here.

This ability to import historical data into new tools makes the cost of switching tools relatively low but this can only be done if your data is easily accessible and in a structure that isn’t specific to any tool.

Finally, let’s look at some of the Pros and Cons of using Segment.com with popular analytics tools like Mixpanel or Google Analytics.

How Segment Plays With Tools Like Mixpanel, Google Analytics, Intercom and Hundreds of Others

It is common to wonder if you should use Segment with *INSERT TOOL HERE* especially since all analytics tools have their own dedicated APIs.

Let’s look at a classic Pros and Cons analysis to understand why you might or might not want to use Segment.

Pros of Using Segment:

The first benefit is that your implementations become easier. Most tools will receive roughly the same data so it doesn’t make sense to write the same event code multiple times.

For example, you might want to send an event to Mixpanel & Intercom. The code these events would look like this:

Mixpanel track function:

mixpanel.track(‘Signed Up’, {

plan: ‘Enterprise’

});

Intercom track function:

var metadata = {

plan: ‘Enterprise’

};

Intercom(‘trackEvent’, ‘Signed Up’, metadata);

They are almost identical but you still have to write two functions. Instead, you could send the data to Segment.com with one function and Segment.com will translate it to Mixpanel & Intercom in the backend.

Segment track function:

analytics.track(‘Signed Up’, {

plan: ‘Enterprise’

});

This also means that if eventually want to switch from Mixpanel to Amplitude, you don’t have to rewrite your event code.

Related: Before you implement Segment.com, you should create a tracking plan that organizes all of your events and properties in one location. You can download the same tracking plan that I use with my own clients.

Segment functions as a tag manager letting your marketing people load small things like marketing pixels. This can be done in a few clicks without technical assistance.

As your hub, Segment also lets you build a historical copy of your data in a database that you control. This is huge especially when you want to switch tools without losing your historical data

Cons of Using Segment:

Like everything in life, there are some cons to consider when using Segment:

1. You will need a developer to help you implement Segment. There’s no way around this and I’m sure you could tell by the overall technical language in this article.

2. Debugging issues becomes slightly harder since you now have two systems in play. This is commonly brought up by the technical support at popular tools like Mixpanel, Kissmetrics, etc.

3. Some tool-specific features will not work through Segment. This is due to a lack of integration or simply because they are incompatible. For example, Mixpanel has a feature called Autotrack which lets you track events in Mixpanel without having to use their APIs. These events would be tracked outside of Segment Analytics.

4. The final con is around pricing. Using Segment means you have to pay their pricing on top of any analytics tool that you want to use. Costs can quickly increase as you get more users and data.

Conclusion

It is easy to get confused trying to figure out the best analytics tools for your company. Instead, you should focus on handling the fundamentals first and the details will fall in place.

Understanding the portability of your data and how easy it will be to collect and store it, is a fundamental issue that needs to be addressed before considering any tools.

Your universe of tools will change quickly so make sure you get this right before you are trying to figure out how to combine data from 10 different tools or how to switch to a different tool.

I would also love to hear any thoughts or questions that might have. You can leave a comment below.