Carnivore diet app: Securing a 54% DAU:MAU ratio with engagement-driving user flows

Conversion-driving onboarding quiz

To personalize meal plans and suggestions, the app needs information about the user. The quiz helps collect it fast, without friction. It asks users about their age, diet goals, level of knowledge about the carnivore diet, weight, physical activity, nutrition habits, and food preferences.

Thanks to social proof and relevant evidence added at critical points in the quiz, the drop-off rates reduced by 13% in our A/B testing. The quiz itself proved an effective conversion tool, with conversion rates reaching 12% within the first six months.

Cost-efficient AI coach

The hallmark feature of the app, the AI coach is there 24/7 to answer users’ questions about the diet, provide advice aligned with their goals, and help them stay on the right track.

After some deliberation, we opted for OpenAI’s GPT-4o to both power the coach chatbot itself and enable it to identify meals in photos (more on that below). At the time, its multimodal capabilities were ahead of the competitors, while API costs were in the sweet spot for our client.

We minimized input tokens and optimized the model output for shorter answers, reducing API costs by 21% compared to unoptimized requests. Our testing ensured that the model’s performance remained reliable in production, with an error rate below 0.02% across key user interactions.

‍

Personalized diet plans

Every day, users can check in with the app to review a set of meal options tailored to their profile: dietary type, allergens, and calorie target. Rather than following a fixed plan, users get a curated selection of meals they can actually prepare based on their individual input.

We pinpointed the sweet spot between detail and conciseness to prevent cognitive overload. Every element is linked to a relevant section so that users can easily get more information. Users can also quickly log their activity or meals without leaving the screen. All of that helped reduce friction and improve engagement, as proven by the DAU:MAU ratio of 54%.

Motivation-boosting weight tracking

As weight loss is a common reason why people may turn to the carnivore diet, we added a separate dashboard to help track it. Based on the user’s input, it displays a chart of how their weight has changed over time.

Users can pick one of three views (weekly, biweekly, and monthly) to gain more insight into their weight loss. They can also review all the logged data points listed below the chart.

To keep users motivated, we added a progress bar that displays how close they are to reaching their desired weight. This decision increased the feature’s adoption rate by 8%, according to our A/B testing.

Searchable recipe library

The app comes equipped with 1,000+ carnivore recipes. Thanks to this library, users don’t have to look recipes up themselves. They can simply tap the suggested meal on the home screen or search for a specific recipe based on the ingredients they have.

In the meal plan, each recipe is accompanied by crucial information about the meal’s caloric count, preparation time, and meal type, easy to grasp at a glance. Once the user taps the recipe, they can see a more detailed protein, fat, and carb count, as well as step-by-step preparation instructions.

‍

Barcode & meal scanning

Grocery shopping and eating out can be significantly more stressful when you’re trying to follow a diet and log your meals. The app’s scanning feature makes both easier.

Thanks to OpenAI’s GPT-4o, users can simply scan their meal to get a breakdown of key ingredients and its nutritional value (protein, fat, carbs, calorie count). Or, if they’re grocery shopping, the barcode-scanning feature will help them track which ingredients they have on hand.

Of course, we put these features through extensive testing and iterative fine-tuning to ensure reliable results. The feature achieves 99% accuracy across common meal and barcode scanning scenarios.

‍

Payment processing

No one wants to jump through hoops to confirm or change their subscription. That’s why subscription management and payment processing had to be embedded directly in the app.

We integrated the app with Stripe and Mollie, which handle online transactions around the world. Thanks to these two integrations, users can pay with their preferred payment method without leaving the app.

To reduce friction even further, we added Google Pay and Apple Pay integrations to enable users to pay even faster with their digital wallets.

‍

Admin dashboard

To facilitate ongoing app monitoring and management, we built an admin dashboard that tracks all key app metrics, from traffic volumes to uptime. Using the dashboard, the client’s team can also tweak the content within the app (e.g., to add new recipes).

Although not part of the initial requirements, we suggested implementing alerts. With our client’s approval, we set up the alerts via email for critical events like unexpected downtime. Thanks to them, the team will instantly know when metrics pass established thresholds, enabling them to intervene in time.

‍