How To Use Analytics To Improve Engagement

In today’s digital economy, capturing a user’s attention is only the first step, the real challenge lies in keeping it. Whether you are running a streaming platform, a mobile app, or an eCommerce site, sustained engagement is the key to unlocking long-term value. Users who return frequently and interact meaningfully are more likely to convert, subscribe, recommend, and remain loyal over time.

But boosting engagement is not just about adding new features or sending more notifications. It starts with understanding your users at a behavioural level: what they do, what they avoid, and what triggers their continued use. With the right data and analytical frameworks, you can uncover what truly matters, and design strategies that shift occasional users into highly engaged advocates. This article explores how to move beyond vanity metrics and build an engagement model grounded in behavioural insights.

Behavioural analysis is the practice of studying what users do, and just as importantly, what they do not do in your product. It involves examining patterns in user actions, such as how often they return, which features they use, what sequence of steps they follow, and where they drop off. The goal is twofold:

1. Understand how your active users currently behave — What are they doing when they engage with your product? Are they browsing, creating, sharing, or purchasing? Which actions are performed most frequently and which are rarely touched?
2. Identify what influences or hinders key behaviours — What prompts a user to return, complete a purchase, or recommend your app to others? Equally, what might be discouraging them — confusing interfaces, lack of perceived value, or intrusive prompts?

By analysing user behaviour over time, you can begin to map out what meaningful engagement actually looks like for your business. This helps distinguish between fleeting attention and genuine value creation.

From Insight to Impact

When done well, behavioural analysis gives product teams the insights they need to craft more effective user experiences. It enables you to move beyond surface-level metrics (such as daily active users) and uncover the real drivers of loyalty, frequency, and intensity of use.

It is the essential first step in any data-driven engagement strategy; one that informs everything from onboarding design and feature prioritisation, to personalisation, messaging, and lifecycle marketing.

Defining Active Users and Engagement

Before you can effectively analyse user behaviour, your organisation must reach a shared understanding of what an active user is, and how engagement is measured. This may sound simple, but it is one of the most important and commonly overlooked steps in building a robust analytics framework.

What is an Active User?

The definition of an active user will vary depending on your product and business goals. For example:

• A news platform may define an active user as someone who opens the app and reads an article.
• An eCommerce site might count those who add a product to basket or make a purchase.
• A B2B SaaS product could define activity as completing a key workflow or using a premium feature.

Whatever your chosen definition, consistency is key. If different teams within your organisation interpret "active user" differently; for example, marketing tracking logins, product measuring content views, and engineering focused on API hits — you will quickly run into misalignment, conflicting reports, and faulty conclusions.

Where Does User Data Come From?

User behaviour data typically originates from a number of sources across your technology stack:

• Backend databases: These contain the foundational records of user actions: such as logins, transactions, clicks, and feature use.
• Customer Relationship Management (CRM) systems: These hold contact details, sales notes, support interactions, and marketing preferences.
• Web and mobile analytics tools: Tools such as Google Analytics, Mixpanel or Amplitude capture frontend interactions like page views, button clicks and time on site.
• Third-party platforms: Data might also live in support platforms (e.g. Zendesk), marketing automation systems (e.g. HubSpot), or subscription managers (e.g. Stripe).

Each of these systems may track users differently, and unless there is a common thread between them, such as a universal user identifier, joining this data together becomes more difficult. A universal ID allows you to stitch user journeys across platforms, devices, and tools, giving you a complete view of behaviour over time. Without one, your understanding of engagement will be fragmented and vulnerable to duplication or misclassification.

Aligning Teams with a Data Dictionary

To avoid ambiguity and ensure reliable reporting, it is crucial to maintain a Data Dictionary. This is a centralised reference document that clearly defines key metrics, dimensions, and data entities, not only in technical terms (e.g. table names and schema), but also in plain business language that all teams can understand.

Your Data Dictionary should include:

• Definitions for “active user”, “conversion”, “retention”, etc.
• The logic or events that trigger each metric.
• Ownership: who is responsible for maintaining each definition?
• Version history and changes over time.

Be Wary of Vanity Metrics

In the early stages of growth, it is tempting to celebrate metrics that look impressive on the surface, especially those that show rapid adoption or large audiences. Figures such as Daily Active Users (DAU), Monthly Active Users (MAU), and App Downloads are commonly reported to stakeholders as signs of product success. These are known as vanity metrics — numbers that look good on a dashboard, but may not accurately reflect the true health or value of your business. While these metrics do serve a purpose, particularly for tracking visibility, reach, or technical performance, they should not be used as primary indicators of engagement or value creation. When used in isolation, vanity metrics can lead to overly optimistic conclusions and misinformed product decisions.

Why Vanity Metrics Can Be Misleading

Vanity metrics often fail to account for quality, intent, or retention. They may spike due to external campaigns or one-off events, but offer no indication as to whether those users found genuine value in your product, or whether they are likely to return, purchase, or convert. Below are three common scenarios where vanity metrics may paint a misleading picture:

1. Spikes from Campaigns Without Long-Term Retention: A company runs a promotional email campaign targeting dormant users, offering a limited-time discount. The next day, DAU doubles — a great result, seemingly. However, a week later, engagement drops back to previous levels and subscription revenue remains flat. The issue: The DAU spike was driven by temporary attention, not sustained interest or deeper product adoption.

2. App Downloads That Do Not Lead to Usage: A mobile app gets featured on the App Store and sees a surge in downloads. But after 30 days, less than a quarter of users have opened the app a second time. Without measuring retention, the team assumes their product is successful based solely on installation figures. The issue: Downloads are a poor proxy for product-market fit or user satisfaction unless accompanied by data on activation, return usage, and value delivered.

3. Active User Counts That Mask Inactivity: A product team defines an active user as someone who “opens the app.” Weekly active user numbers look steady, but deeper analysis shows many users are logging in and exiting within seconds, without taking meaningful actions. The issue: The definition of “active” is too loose — it counts users who technically meet a threshold but are not truly engaged.

Focus on Metrics That Reflect Value

To move beyond vanity metrics, organisations should focus on value-based metrics that are more closely tied to business outcomes and user satisfaction. These might include:

• Retention rate (e.g., 7-day or 30-day retention)
• Conversion rate (e.g., from trial to paid)
• Time to value (e.g., time taken to reach Aha! Moment)
• Net Promoter Score (NPS) or Customer Satisfaction (CSAT)
• Feature-level engagement (e.g., number of playlists created, forms submitted, or sessions completed)

Stickiness

Stickiness refers to how frequently users return to your product over a given period. A sticky product is one that becomes part of a user's regular routine, something they come back to repeatedly, often without needing prompting. This kind of habitual use is a strong indicator of perceived value, and is directly linked to customer retention, monetisation, and long-term profitability.

One of the simplest ways to measure stickiness is through the DAU/MAU ratio — that is, dividing your Daily Active Users (DAU) by your Monthly Active Users (MAU). This metric provides a baseline estimate of how often the average user interacts with your product in a month. For example:

• A 50% DAU/MAU ratio suggests the average user is active about 15 days per month.
• A 20% ratio would indicate more sporadic usage — around six days per month on average.

The higher the percentage, the more consistently users are returning, and the more "sticky" your product is considered to be.

Why Stickiness Drives Profitability

A sticky product creates value in several important ways:

1. Improved retention – Users who return regularly are less likely to churn and more likely to build habits around your product.
2. Higher monetisation potential – Frequent users are more likely to engage with premium features, renew subscriptions, or complete purchases.
3. Greater customer lifetime value (CLTV) – The more often users engage, the longer they tend to stay, increasing the total value generated per user.
4. Stronger viral growth – Sticky users are more inclined to share the product, write reviews, or invite others, which can reduce customer acquisition costs.

In short, stickiness indicates that users are not just trying your product, they are depending on it.

Metrics That Reflect Stickiness

While DAU/MAU is a good starting point, there are other metrics that can provide a more detailed view of how embedded your product is in users’ lives:

• Session frequency – Average number of sessions per user per day or week.
• Recency – Time since a user’s last session or key activity.
• Feature-specific engagement – How often users interact with your product’s core features (e.g., number of playlists created, workouts logged, projects updated).
• Time in app/site – Average session duration, particularly for content-heavy platforms.
• Retention curves – Cohort-based retention over time gives a more granular view of habitual usage.

It is also useful to monitor segment-level stickiness, for instance, how stickiness differs between free and paid users, new vs. long-term users, or by geographic location. Averages can hide wide variation that may require different product or marketing strategies.

Power Usage Curve

A user curve visualises usage frequency in the form of a histogram. It shows the percentage of users by the total number of active days in a month. You can choose other metrics that better reflect the health of your business, e.g. total number of swipes or views in a month. You can use a different time period if your product has a different usage cycle such as weekly. You can also use the power user curve to evaluate a particular feature, e.g. the number of times users shared playlist in chats.

Power users refer to those who are the most active, e.g. more than 25 active days in a month. They sit in the last bars on the right of the histogram and are the most desired users. Ideally you would like to move everyone further to the right. Below is a power user curve with a very nice right-leaning smile aka a group of very sticky users who come back almost everyday.

From the power user curve you can segment your users into different categories and start comparing them for deeper insights. For example, which feature your power users return to the most? Are one-day users aware of or fully understand the benefits of this feature?

Discovering Aha! Moments Through the Power User Curve

The Power User Curve is a powerful visual tool that helps product teams understand how frequently users engage with their product over a given period, used correctly, it becomes a strategic lens for uncovering what truly matters in the user experience — including your product’s Aha! Moment.

What Is an Aha! Moment?

The Aha! Moment is the point in a user’s journey when they suddenly perceive the core value of your product, a turning point where casual interest turns into consistent usage. It is often described as the moment the product “clicks” for the user. For some apps, this could be when a user completes their first successful workflow (e.g., sending a message, creating a playlist, saving a project). For others, it might be when a user hits a certain threshold of activity that signals genuine engagement. The earlier this moment occurs in the user journey, the better, it significantly increases the likelihood of long-term retention and deeper usage.

Aligning Behavioural Data to Discover Your Aha! Moment

To identify what triggers this lightbulb moment, product teams can use behavioural correlation analysis. One effective method is to model the pairwise relationship between key behavioural metrics (e.g. “number of searches”, “videos watched”, “comments posted”) and a power user metric (such as “active days in 30-day period”).

By looking for positive correlations, you can discover the actions most closely linked to high engagement. These are the behaviours you want to nudge users towards early, through onboarding, product design, and UX interventions.

Conversely, this analysis may also reveal negative correlations, actions that disrupt flow or indicate disengagement. For example:

• Referral or payment prompts shown too early can frustrate users.
• Overly aggressive upsells might harm retention if the perceived value is not yet clear.
• Complex onboarding flows could reduce the likelihood of reaching that Aha! moment altogether.

The key is to optimise both the timing and frequency of these behaviours. Prompting the right action at the wrong time can be just as damaging as never prompting it at all.

Example: Facebook

Facebook famously discovered that users who added 7 friends within their first 10 days were significantly more likely to stick around long-term. This insight, derived from behavioural analysis, led the growth team to reframe early user journeys around friend discovery, rather than just account creation. The result: a sharper onboarding strategy, faster time-to-value, and stronger long-term retention — all by designing the product experience around the behaviours that actually mattered.

Behaviour Is Only One Side of the Value Equation

While user behaviour helps us understand how people interact with your product, it does not always reveal which users are most valuable to your business. Two users may use the product in similar ways; browsing, clicking, completing workflows, yet contribute very different amounts to revenue. One may be a loyal customer who converts and refers others; the other, a habitual free user who never upgrades. Understanding behavioural patterns is essential, but so too is knowing who is driving your bottom line.

This is where combining behavioural data with commercial data becomes crucial. The challenge? These different data points often reside in separate systems: behaviour might live in your product analytics tools, while purchase history, subscription value, and campaign responses live in CRM platforms, finance systems, or customer support databases. Without a central data warehouse to integrate this information, you are left with fragmented views that hinder both insight and action.

A modern data stack allows teams to create a unified customer view. This not only helps you identify your most valuable users, but also allows you to model and forecast the behaviours that lead to revenue, not just engagement. And it is important to remember: what users say they want is not always what drives value. There are many famous cases where listening to feature requests did not increase profitability. For example:

• Twitter (pre-X) rolled out features like Moments and Fleets based on user feedback and trends, but they failed to drive meaningful increases in ad revenue or user retention and were eventually retired.
• Netflix once introduced an advanced star-rating system based on user demand, only to discover that viewer habits didn’t match their stated preferences; they later replaced it with a simpler thumbs-up/down model, driven by actual viewing behaviour.
• Snapchat made interface changes in response to usability complaints in 2018, only to see a loss of 3 million users the following quarter and stagnant ad growth, as the core engaged users resisted the new design.

These examples highlight the importance of grounding product decisions not just in feedback or usage frequency, but in holistic user value, informed by robust, centralised data.

Final Thoughts

Understanding and improving user engagement is one of the greatest challenges facing digital product teams. However, with the right data, metrics, and behavioural insights, you can identify what truly drives value for your users, and grow accordingly. If you are interested in how this can apply to your organisation, the friendly team here at 173tech would be delighted to have a chat with you.