I was recently working with a startup doing $50k+ of MRR. The company was quickly growing but their analytics couldn’t be more screwed…
During our first meeting, I asked: “What works best regarding your Marketing?”. I was shocked when they replied: “We don’t really know. Content is our main traffic driver”.
I realized that they weren’t measuring anything at all. They had several thousands of visitors per day, but they had no way to measure where their customers were coming from.
Screwing up your analytics will hold you back on making data-driven decisions and effectively growing your startup.
Gut will guide your decisions. Half of which will be wrong… Don’t just hope that your competitors will do the same…
This article is about to give you an overview of the Analytics Lifecycle in Startups, what’s wrong about it and how to fix it.
Read on if you want to avoid waking up one night unable to answer the question:
- Where are they coming from?
- What do they do?
- What needs to be fixed?
- How can I grow more?
Your Skewed Analytics
Most people will install Analytics tools and forget about:
- Configuring these tools
- Verifying data consistency
- Making sure they track the right things
Dropping a line of code into your software isn’t implementing Analytics.
Failing to implement tools correctly leads to your Analytics being useless because:
- You track way too many metrics
- You don’t track the right metrics
The result? You take decisions based on what you think is right.
- “We think Content Marketing is working”
- “We think they’re using our Trial for 15 days”
While the industry is trying to reduce the guesswork out of Entrepreneurship (e.g. Lean Startup, Growth Hacking), you’ll be doing the complete opposite…
I’ve seen tons of companies failing at this. What would generally happen is:
- Content Marketing is driving Traffic
- Let’s set up Analytics to see what we can do better
- Well. We just realized our Marketing wasn’t as effective as we thought
You need to make sure that you can measure your actions from the early beginning. Just investing a few hours will save you a ton of sweat.
Creating a data-driven culture among your organization will ensure that you can get better over time. You’ll learn more thanks to quantitative data and you’ll improve your likelihood of success.
The Analytics Lifecycle Stages
Most startups will follow a 4 steps lifecycle with their Analytics:
- They’ll implement Google Analytics right from the beginning
- After growing a little (+/- $10k MRR), they’ll implement an Event-Based Analytics like KISSmetrics or Mixpanel
- After growing to $50k+ MRR, they’ll implement a Business Intelligence software
- After scaling their business, they’ll start investing in a Custom Analytics Stack
Besides is an explanation of the different stages and what often goes wrong.
1. Google Analytics (GA)
One of your developers simply plugged in Google Analytics on your website. Everyone forgot about it. You just check out the number of visitors sometimes.
You drill down into your Traffic Sources and look at your bounce rate. Most people will stop here.
You will not be able to measure anything meaningful. You’ll see where your visitors are coming from, but not where your customers are coming from.
The issue: Without any further configuration, Google Analytics won’t tell you the traffic source behind conversions.
The fix: Implement Google Analytics Goals
2. Event-Based Analytics
At some point, you’ll start to wonder what’s happening within your app. Since GA doesn’t tell you much (unless well configured), you’ll install an Event-Based Analytics like KISSmetrics, Mixpanel or Amplitude.
You “set it up”, track as many events as your developer is willing to integrate, forget about it in a while.
Tracking every metrics doesn’t make sense. Most of the thing you’ll track will not matter. You won’t be able to move the needle.
The issue: You have no real metrics. You know how many people “Added X” during the week but that doesn’t show you anything.
- Don’t try to measure everything because you’ll end up in analysis paralysis.
- Don’t rely solely on event-based Analytics for your Traffic Sources & Conversion since most of them have limited possibilities when it comes to attribution.
3. Business Intelligence
You’re now rocking. You’re at $50-100k+ of MRR. Everything is amazing.
You current stack is only limited to events happening within your app. But that’s only a small portion of what you can actually analyze.
Your MySQL database, MailChimp, your CRM and all the software you use also hold an incredible amount of data.
You want to gather all this data into one unified datasource. You call in the big guns and implement a RJMetrics, Looker, Chartio…
4. Custom Analytics Stack
You became so big that your current Analytics stack can’t answer your questions anymore. You have too much data. You just can’t stay like this.
You therefore decide to build your own Analytics stack with Amazon Redshift and dozen of other software.
(Please note that Amazon Redshift isn’t the only solution available)
I’m not going to get into much details here, because at this point, you’re not a startup anymore.
If you want an example of how to set that up, the folks at 500px built their own Analytics.
The main issues with data can be summarized by:
- Our Team doesn’t understand why (or how) they can use data
- We didn’t track the right metrics / Optimized the wrong things
- The tools we implemented don’t fit our business
The 3-Step framework below will ensure that you can create something great with your analytics.
Your Analytics success (or failure) mainly rely on the people using this data.
At the moment, if no one is willing to look at data within your team, changing your software won’t change anything.
Making sure that employee’s have access to data and are willing to use it is therefore the Step 1.
Avoid data silos within your organization by creating a data-driven culture.
2. The Measurement Plan
Now that your team is aware of how data is important to your success, it’s time to define your metrics.
Here are the steps involved in designing your measurement plan:
- What are your business objectives? (e.g. Grow our Customer Base)
- What are the Strategies & Tactics? (e.g. Twitter)
- What are your KPIs? (e.g. CR from Twitter)
- What are your targets for each KPIs? (e.g. 15%)
- What are the segments? (e.g. Traffic Sources)
By doing so, you’ll ensure that your business objectives, strategies & KPIs are all tied together.
To help you create your own plan, Avinash Kaushik wrote an article ab Digital Marketing & Measurement Model.
This plan isn’t set in stone. It’ll evolve with your business’ goals and marketing strategies.
3. The Stack
Your analytics stack then flows easily. You want a stack where you can implement all the metrics included in your Measurement Plan.
You also want something that will be accessible across your whole organization. Everyone shouldn’t have to write SQL to query the datasource.
I strongly advise you to implement Segment so that you can manage your stack without too much development.
Where most people go wrong is that they think about software first. After implementing it, they realize a while later that they can’t measure what they need.
The Day 1 Analytics Stack
You had 500 sign ups during your first week, but no one ended up paying? What went wrong? You need to know. You need to be able to answer questions like that from Day 1.
For the early-startups out there, I wanted to simplify the process:
- Think about the Metrics that you want to track (aka Measurement Plan)
- Implement Google Analytics & Configure Goals
- Implement an Event-Based Analytics
If you’re bootstrapped, you can use Amplitude or Mixpanel, which are free Event-Based Analytics.
Investing in Analytics during your early days will pay off big time later on.
You don’t understand anything about Analytics? Get a friend or a consultant to help you with your Measurement Plan + Implementation.
Few hours of work on your Analytics can have a huge impact on your bottom line.
Don’t ship until you have a way to measure success (or failure), otherwise you won’t learn anything.
Sure, you might have a couple customers, but what happened to all the people who didn’t convert?
Hope you now have a clear picture into what your Analytics should look like depending on your stage.
What success did you have implementing an Analytics culture? Did you struggle with your tools?