Wilde

In this episode, we dive deep into the world of modern data architecture with Patrik Devlin, Co-Founder and CTO of Wilde. Discover: ??How Wilde AI uses DuckDB and MotherDuck to power flexible, scalable solutions ??The truth about data clean rooms - clearing up misconceptions and marketing hype ??Strategies for building secure and efficient data sharing ?? Listen now https://spoti.fi/3Z0Fp3j

Users expect great products now more than ever before. However, that quality does not have to come at the expense of time. At Wilde, velocity is baked into the architecture. Last week we received a lot of asks about building out a better cohort visualization. This week we’ve launched the new feature, and here are a few guiding principles: dbt is not just for analysts - Transformations are directly consumed by our dashboards.? - Data contracts and testing keeps things in check from unintended breaking changes. Move less data (shoutout DuckDB/MotherDuck) - No need for a backing analytics api to serve up data - Object types are closely shared between the sql that produces them and the consuming FE viz The dashboard below blends Wilde's predictive and historical ltv. If you’re interested in Wilde, shoot me a message! Would love to connect

The secret sauce for e-commerce darlings? We analyzed 247 brands to uncover the truth. Here's the answer: ACTIVE CUSTOMERS (Active customers => 80% probability of repurchasing) Growing brands have TEN TIMES as many Active customers. That's crazy. You need a very strong, loyal customer-base as a foundation to growth. Growing brands' Actives may not be worth as much as Shrinking brands, but I'd trade a few $ for 10x more loyal customers. Your costs dictate how valuable your Active customers need to be. NEW CUSTOMERS (One-time purchasers & younger than median days to 2nd order) Growing brands have twice as many New customers. And these brands can afford to be more aggressive with CAC because Growing brands' New customers are worth 50% more than Shrinking Brands. This is a dramatic difference, but you can see how retention spins the flywheel. ------------------------------------------------ TL;DR Growing brands have 10x as many Active customers, which is the foundation on which a brand can scale acquisition. Without loyal customers you won't grow. IN CASE YOU MISSED IT -- we released Wilde's E-com industry report yesterday. This is just one insight that changed my perspective. Comment below if you want a copy.

247 brands, $2.3B yearly GMV, 18M consumers, and 150 hours of my time. ???? We're releasing Wilde's E-commerce Industry Report ???? This has been the most difficult, rewarding, and insightful project I've ever undertaken. Questions we answer: - Is the industry growing? - Is inflation driving higher AOV? - What do Wilde's models predict will happen in the next 12 months? - What separates growing brands from floundering brands? (not what I was expecting) Overall, things are better than they appear. But we found a few key areas you'll want to focus on if you want to be one of the 53% of brands that are growing. If you care about the e-com industry and where we're headed then this is a must-read. Personally, I'm excited to hear your feedback and share follow-up analyses based on your questions. ------------------------------ Comment below and I'll dm you the report (we need to be connected first)

Where are you on the LTV Maturity Curve? I've seen this play out a 100 times. And you're probably immature. But getting to a good LTV calc is a journey. Here's how it generally plays out: 1?? COHORTS Your company knows LTV is important. Maybe the board asks for quarterly updates to prove the economics are sound. The finance team is tasked with owning the LTV calculation. They open an Excel workbook, record the number of new and returning customers every month since inception, throw in a pivot table, and voila! You're calculating the avg historical LTV per cohort per month after first purchase. Want predictions? Look at other older cohorts and use those numbers to guess. 2?? RFM Profitability is becoming more important across the org. The marketing team would like to use the LTV calculations for marketing actions. But the cohort-level calculations aren't actionable. The marketing team proposes using an RFM (Recency of last purchase, Frequency of repurchases, average/total Monetary value) model. They say it's more actionable. To make segmentation easier, the marketer decides to use only Recency and Frequency. Eight segments are created based on each customer's combination of RF. Names for each segment are created to make it easier to remember. It's been six months since implementation and no one can remember the difference between a Loyalist and a Champion. 3?? PREDICTIVE The data science team steps in. They predict LTV for each customer individually based on the customer's buying pattern, traits, and behaviors. Now that only one metric (predicted LTV) is being reported, it's easier for marketers to take action. And because there's only one metric to segment (LTV vs RF), they can create fewer segments. The DS team reports cohorted and aggregated LTV predictions to the finance team to be used for forecasting and reporting to the board. The DS team also measures model accuracy over time to increase trust. Now you have a true North Star Metric. Everyone in every org can look at the same metric and know exactly what they need to do. This stage is the most complex to calculate, but results in the most value created. ------------------------- TL;DR - Calculating LTV is a journey, keep iterating - Ownership of the calculation will probably change a few times - The most mature stage is being able to predict customer-level LTV (If you don't have a data science team, you'll need software or a consultant)