Event Tracking in MVP: Key Metrics to Measure Early

Today, tracking key metrics is what defines a startup that bakes analytics into its DNA. The evidence is solid when it comes to outcomes of such data-driven organisations compared to those who are simply “data-informed”, or those who rely on guesswork or intuition. For instance, the AWS research shows that 65% data-driven companies outperform their ‘data-informed’ competitors. This financial success is driven by stronger performance across key metrics, such as: 

• customer satisfaction – 69%
• revenue growth – 65%
• operational efficiency – 65%
• marketing performance – 61%
• cost savings – 55%
• risk reduction – 50%

Overall, data-driven companies are consistently about 2 to 2.5 times more likely to report positive business outcomes. So, what is a data-driven startup? According to the Segment’s co-founder, it means not simply tracking data but building the entire startup’s operations around it.

In practice, it often looks like:

1. Setting up events tracking and meaningful startup metrics that matter for your product;
2. Setting up dashboards to ensure visibility for the entire team;
3. Making the data impossible to ignore, up to the point of buying a TV to display dashboards and positioning it at the center of the startup’s workspace. Anything to ensure that everyone, from a founder to a new hire, actively thinks about how to move the key metrics. 
4. Every week or month, there are initiatives that target the metrics, and their impact is monitored carefully.
5. Based on the results, the team designs new initiatives/projects to target the metrics.  

In this post, we’ll discuss why no analytics is not an option anymore, key tools and how they’ve evolved, and setting up event tracking with them. 

Data-Driven Approach & Data Analytics Market

The widely-known quote by American statistician W. Edwards Deming says:

“In God we trust. All others must bring data.”

Data-informed decision-making is a much stronger pillar for driving a company’s success than opinions, guesswork, or intuition. In business, it helps to overcome the HiPPO—the Highest Paid Person’s Opinion, which often leads to higher failure rates, costly mistakes, and stifled creativity.

The debate whether a startup needs to set up analytics from Day 1 has decidedly ended by 2026. Launching without it is seen as a risk. More precisely, simply poor analytics data leads to a loss of 12% annual revenues. When one thinks about the difference between extending a startup’s runway and running out of cash within year 1, this 12% can easily mean sink or swim. And this 12% is about poor analytics; the results for no analytics at all are not even measured. 

The market for analytics tools has long passed the experimentation stage and is moving into strong adoption and growth. Its size is expanding, and the growth is accelerating. According to Grand View Research, the data analytics market is in for rapid growth, and it is not oversaturated yet.

This means that new tools will be appearing, and, on average, even now, best-in-class analytics tools tend to change every 3-5 years. Therefore, for a startup founder, it may not make much sense to spend time on the tool selection. Choosing what’s popular now and offers a free or very affordable startup plan is good enough. 

Tooling for Key Metrics and Event Tracking

To illustrate the market shift from the introductory stage to the growth stage, let’s look at the analytics tools then and now. Here is a screenshot of major analytics tools 6 years ago. In here, you have 4 major categories:

1. tools for managing campaigns such as Mailchimp, Marketo, customer.io, and Intercom.
2. pageview-era tools like Google Analytics, Omniture, KISSmetrics, Quantcast.
3. Emerging behavior analytics tools that constitute the core of Event Tracking – Mixpanel, Flurry, Crittercism, Amplitude, and FullStory.
4. paid acquisition and CRM tools – AdRoll, Salesforce, and Facebook Conversions API.

This represented the earlier fragmented analytics approach, which also focused on casting wide nets to capture massive amounts of data, yet only a small percentage of it was actionable.

Market Trends Now

Since 6 years ago, the tooling trend shifted:

• Focus on platforms and tooling consolidation. Adobe acquired Marketo and Omniture to add to its comprehensive Adobe Analytics suite. This serves well enterprise-grade decision-making, but so it comes with a hefty price tag.
• Newer analytics tools focus on privacy, and data collection moves from tool-owned to company-owned. Tools don’t use third-party cookies and assign in-product identities. And most tools allow you to self-host the data on your own servers, or pipe it into data warehouse tools like BigQuery, Amazon Redshift, or Snowflake. 
• Shift from pageviews, clicks, and sessions to event tracking, user journeys, and product usage. Most tools offer event tracking as well as an array of other options. For instance, in line with Amplitude and Mixpanel, there is PostHog with a developer-friendly SDK and APIs. After all, event tracking is a foundation that enables product experimentation: if we add X, do people do Y more often? If we relocate Z, do people perform A and B now more often? Etc. Event tracking makes sense with other tooling, such as feature flags, session replays, surveys, and error tracking. And, now it can all be found in one tool, while others manage data storage, security, and querying.

The key here is that event tracking of the user actions is the foundation for all the tooling. You set up feature flags to measure the impact different features have on user behavior – their actions expressed through the events. Querying databases focuses on crunching different events to find correlations and determine what causes certain user behaviors to occur. Error tracking is also somewhat related to events, as it can answer the question of what users are trying to do that makes the app fail. 

Event Tracking for Startups: Funnel vs. Viral Loop

The way startups bring their product to market is also undergoing a shift: linear marketing funnel vs viral growth loop.

The image below shows a traditional funnel approach.

This way, event tracking revolves around defining key events for these stages:

1. Acquisition – the top of the funnel that tells you how many people visit your app and how they get to it.
2. Activation – this is the first moment when the user experiences value. However, sometimes, startups measure user milestones here, like completing the onboarding or filling out the profile. This is useful but secondary data.
3. Retention – this event measures when a user does the value action again, which might be the same as in the activation stage. It can also be a different event if your MVP has a range of functionalities, or core functionalities differ based on the pricing plan.
4. Referral – this is to measure participation in referral programs, such as “Invite a friend, get 20% off”.
5. Revenue – here you set up events to track monetization from upgrades to trials to cancellations.

Viral Loop 

Other startups build viral growth loops into their products during MVP Development. 

So, in a viral loop approach, the stages are:

1. Awareness – this is mostly focused on tracking a non-user referral visit, such as what a non-user does following the link from an existing user.
2. Download – this measures if a referred visitor decided to create an account or download a mobile app. Usually, the link gives access to shared content, and there is an invitation to sign up. So, this tracks whether users experienced enough value to sign up or download an app if it is mobile.
3. Activation – after a referred user signed up, did they do the core value action? For the loop to perpetuate, new users should repeat the same core action. 
4. Share – generating & sending a link for a non-user to access the artifact (video, piece of content, payment, calendar event, etc.) or an external invite link.

What is an Event?

Before we dive into event tracking with examples, it is essential to understand that modern events, like in PostHog, are not a single piece of data like page_views. Events correspond to a user action and often have properties. For instance, default properties in PostHog coming from its documentation are as follows:

• distinct_id, 
• session_id, 
• $device_type, 
• $current_url, 
• $browser, 
• $os, etc.

However, you can set up measuring more properties, such as those shown in the screenshot below. You can measure screen dimensions, feature flags, whether the event is a click-activated or keyboard-activated, IP addresses, etc. 

Most importantly, though, tools like PostHog allow startups to track events from registered users and anonymous ones. The difference will lie in the fact that when a registered user creates events you track, they will be linked to the user’s profile. When the user is anonymous, they will receive a unique ID, and you can still view all events in the session, but there is no personified data. 

Finally, many tools, including PostHog, offer default events like $autocapture for forms, buttons, and links inside your app, $pageview / $pageleave, etc. These allow you to link up a new tool and immediately start getting some data. However, it is best to create custom events and disable default ones. After all, for PostHog, the free limit is 1M events. So if you track every form field and every button click, it can be up to 100 events per visitor. This really limits the number of users you can observe per month interacting with your app. Though if this is a closed beta, where you limit a number of users you allow in, it can be a non-issue. Though, of course, you will get a barrage of data for those when combining custom and default. 

Event Tracking For Funnel-Based Product Growth

Here are some concrete examples of events and their properties that you might use for your startup.

Stage	Event Tracking	Event Properties
Acquisition	“Landing Page Viewed”,  “Signup Page Viewed”	“utm_sourced_detected” – where the visitor came from, “utm_campaign” – specific marketing campaign, such as a spring sale, or 50% promo “utm_medium” – is it organic, PPC, social media, email, etc.
Activation	Primary Event: “Core Action Completed” – based on key problem your app solves (e.g., a stay booked, a purchase made, etc.) Secondary Events: “Onboarding completed”, “Profile Filled Out”, “Account Created”	“Device”, “time_to_complete” – these can help detect early friction and optimize for time-to-value; You can assign properties like “value”/”outcome” – to track the early usage/value signals. For instance, it can be the direct value of purchase, or the size of the project, the stay booked for a day or 15 days, which also vary in their significance for your business
Retention	Primary Event: “Core Action Completed” – based on the key problem your app solves (e.g., a stay booked, a purchase made, etc.) Secondary Events: “Onboarding completed”, “Profile Filled Out”, “Account Created”	“frequency” or “time_since_last_action” is a key property, which is helpful for early cohort formation: light users, habitual users, power users, etc. For secondary events, you’d want to track the ‘source’ – what triggered return behavior, was it an in-app notification, or direct
Referral	“Referral Signup Completed ”Secondary: “Referral Invite Sent”, “Referral Link Clicked”, “Referral Reward Granted”	“referrer_user_id”“channel” (email, link, social) “invite_count” “incentive_type” (discount, credit, etc.)“Campaign_id”  – these help to determine the most efficient referral programs and channels
Revenue	“Subscription Started” or “Payment Completed” “Trial_started”, “Plan_downgraded”, “Subscription_cancelled”	“revenue_amount”, “currency”, “plan_type”, “billing_cycle”, “discount_applied”, “is_trial_conversion” – all these help analyze the quality of monetization For downgrades and cancellations, you can also have properties like “tenure”, “last_active”, “cancellation_reason”

Event Tracking for Loop-Based Product Growth

With the viral growth loop, the focus shifts to which content generates the most new users to convert, over which channels, and what steps non-users take to experience the value, become registered users, and trigger a new loop.

Stage	Event Tracking	Event Properties
Awareness	Here, you should start with a trinity of events: “Referral Link Clicked” “Referral Content Consumed” “Referral Landing Page Viewed” or “Sign Up Started” These should help answer the question if the loop artifact is valuable enough for the user to consume it and check out the value proposition	“referrer_user_id” “invite_id”  “content_id” “content_type”, if you have several options
Download	“Signup Completed” “App Installed”	Generally, you just carry forward these properties: “referrer_user_id” “invite_id”
Activation	“First Value Action”	You would still keep carrying forward the referrer_user_id, but might also track “time_to_activate”
Share	“Invite Link Sent”	The mandatory ones are “content_id” and “user_id”, but if you have share options, you might also track “channel”, such as ‘copy link’, WhatsApp, Facebook, Instagram, etc.

FAQ: Event Tracking in MVP: Key Metrics to Measure Early