Ruben Ugarte

New video: How to Analyze Customer Onboarding Data in Amplitude.

In this 15-minute walkthrough, I cover 5 steps for discovering were you're losing users and how to fix them.

2 months ago | [YT] | 1

Ruben Ugarte

Should every idea be an A/B test?

Not quite.

It’s easier than ever to spin up a test. But running too many wastes time and can create new problems, like more support tickets.

A better approach is to prioritize using a framework like ICE (Impact, Confidence, Ease). Consult other stakeholders, talk to customers and then test the best ideas.

This slows things down a little, but it ensures the tests you run are worth it.

Quality beats quantity when it comes to experiments.

3 months ago | [YT] | 1

Ruben Ugarte

Are you confused by duplicate dashboards?

Here’s how to fix it.

Open your reporting tool and you might see three acquisition dashboards, ten funnels, and five retention reports. Which one is the “real” version?

The fix is simple: adopt a naming convention. Mark one version as the “production,” “approved,” or “master.” Borrow the language of engineering.

That way, when someone asks for acquisition metrics, they know exactly which report to trust. Fewer duplicates, more confidence.

3 months ago | [YT] | 0

Ruben Ugarte

I recently signed up for ‪@OnePageCRM‬

And I loved their onboarding flow

Their homepage says setup takes four minutes. It took me 10 but that’s still impressive.

The flow was simple: create an account, add company info, add a contact, set the next action. That’s it. No endless steps, no over-engineered journey.

They’ve nailed the idea that a CRM should make the next action the star. It’s a great reminder: onboarding should deliver core value quickly, not make users jump through hoops.

3 months ago | [YT] | 0

Ruben Ugarte

Why did I change my mind on data autocapture?

Here’s why.

Ten years ago, when I was setting up tools like Mixpanel and Amplitude, I thought you had to instrument everything: every event, every property, every user trait for maximum value.

The downside is that it can take weeks or months to get data into production. Autocapture can get you data in hours or days.

You can start fast with page views, clicks, scrolls, form submissions, and mobile equivalents. It’s not the most complete data, but it’s enough to deliver early wins.

Those wins help you earn resources for deeper tracking later.

Props to ‪@MixpanelAnalytics‬ , ‪@Amplitude_HQ‬ for adding autocapture and to companies like ‪@Heapio‬ for pioneering the concept.

3 months ago | [YT] | 0

Ruben Ugarte

Your team doesn’t need more data.

You need a mining process.

Raw data is like raw ore — valuable, but useless until refined. The gold comes from consistency: clear questions, repeatable analysis, and documented insights.

Do this weekly and you’ll surface patterns no one else sees. Ignore it, and you’ll drown in numbers with nothing to show. Data only becomes valuable when it’s mined. Otherwise, it’s just rock.

3 months ago | [YT] | 0

Ruben Ugarte

Forget about data-driven decisions.

Yes, really.

Data should support decisions, not drive them. If you only move when the numbers tell you to, you’ll always be late. The best teams use data as a compass that guides intuition, strategy, and experiments, but they never let it sit in the driver’s seat.

Judgment, experience, and context still matter most. Don’t let data paralyze you. Use it to sharpen your instincts instead.

3 months ago | [YT] | 1

Ruben Ugarte

Is your data architecture sound?

Here are the signs it’s not.

A missing property here, an inconsistent definition there — and months later the foundation collapses. These cracks seem small at first, but they destroy trust over time.

Bad data architecture doesn’t break overnight. It leaks.

Invest in data QA early. It’s far cheaper to patch a small leak today than to rebuild the whole system tomorrow.

3 months ago | [YT] | 0

Ruben Ugarte

Are you falling into these analytics traps?

Many teams are.

Three stand out: vendor lock-in that traps your data in one tool, overpaying for enterprise plans you don’t fully use, and stacking too many tools that create noise instead of signal.

Each of these slows progress, wastes budget, and erodes trust. Analytics isn’t about adding more complexity. It’s about creating foundations you can actually grow on.

Avoid the traps, and you’ll move faster than before.

3 months ago | [YT] | 0

Ruben Ugarte

How do you know if you have too many tools?

Here’s how.

If you have 10+ analytics tools but still can’t answer a basic question like why users churned last week, you have too many tools. The solution isn’t to add another product. The solution is focus.

Start by mastering one or two. Extract every ounce of insight. Then decide if you really need the next one. More tools don’t equal more clarity. They usually create more confusion.

3 months ago | [YT] | 0