Accelerating Growth Cycles with Analytics, Part 5: Closing The Loop.

July 8, 2020 | By Candice Ren.

In previous posts, we went through the four stages of user journey and how analytics can be applied to accelerate growth at each step: acquisition => activation => engagement => retention.


Now it is time to mix and match and create the 1+1 > 2 effect.


Churn Prevention


Churn Propensity + User Segmentation => Customised CRM

When it comes to user churn, like health matters, prevention and early intervention are way more effective than trying to bring someone back to life. Your churn propensity model will alert you when users are showing ‘early symptoms’, detected by a change in behaviour patterns and user state. Your segmentation model then provides a targeted response based on their unique characteristics.


For example, you are an e-commerce platform selling clothing online and have segmented your user base by three unique selling points (USPs): variety, quality and price competitiveness. Your prevention email for high churn propensity users who sought variety can read: “We have added new brands, come and check them out!” vs. price conscious consumers “Your special code inside!”. For those who did not react to the first message, add a follow up CRM mechanism with increased level of intervention.


Don’t forget to apply A/B testing methodologies when designing and continuously improving your CRM strategies.


accelerating-growth-part5

User Acquisition Revisited


LTV Prediction + Marketing Budget Optimisation

This is the single most ROI effective data science project every company should be doing. We have seen wild successes beyond clients believe with the following approach. You can read more about it in our other post here.


  • Create your own centralised single customer view (SCV). It gives you visibility on each user’s demographics, acquisition source, and all subsequent interactions with your product and CRM.

  • Build LTV prediction models on individual customer levels based on historical and live data collected in your SCV.

  • Use predicted LTV as an immediate signal to optimise your marketing spend on Facebook or Google.

Spice It Up With User Segmentation + Audience Selection

High LTV users come in different forms, so should your acquisition strategy.


Let’s reuse the USP segmentation example above from the e-commerce site. You may create three separate Facebook Lookalike Audience for each USP segment. In the ad for a design and quality-driven audience, use more inspirational materials such as how you selected new brands during the latest Milan Fashion Week. For price-driven audiences, opt for a more direct copy such as “Best price guaranteed!”.


If you have also done your homework on LTV per segment, you know how much different types of audiences are worth to you. You can assign bid strategies accordingly and maximise the collective LTV on your campaign.


Final Thoughts


Above are the most common examples. However, the potential is endless. For example, AB testing can be applied and combined with any aspect of your products and services. You can create specific landing pages and conversion funnels targeting certain user segments.


We hope you enjoyed our growth series. With a robust data structure and processes in place, your analytics journey can be streamlined and really enjoyable. It can be integrated into all parts of your products and services and guide your growth.


As always, we are happy to chat if you have any questions or exciting data ideas. Reach us on hello@173tech.com!


Subscribe to our monthly newsletter for more insights.

By subscribing, you agree to receive a monthly newsletter from 173Tech.

Get In Touch

Send us a quick message about your data challenges, we are always happy to chat!