Lampoo: +80% YoY ROAS through Cassandra’s Marketing Mix Modeling Software for non-technical marketers

In recent years, marketing mix modeling (MMM) has evolved from its traditional form to a modern, automated tool that can be used to measure the holistic impact of media and inform decision making around marketing budgets and tactics. 

‍

Because of its increased utility in an evolving advertising landscape, we at Cassandra have invested in building our very own MMM SaaS tool which is helping our clients like Lampoo better invest their ad budget. 

‍

The software enabled the Marketing team to quickly build their Marketing Mix Model and gain insights that improved its overall marketing strategy ahead of a major sales event, even without any data scientist resources available in-house

‍

This case is based on collaborative work between Lampoo, Cassandra and Meta Marketing Science team, aiming at gaining a strong practical understanding of how MMM SaaS solutions can effectively support decision making for fast-moving digital-first advertisers 

‍

Lampoo: the Luxury Resale Scaleup with Sustainable Elegance

‍

Lampoo is a fast-growing startup in the luxury second-hand market. Their platform, a haven for authenticated pre-owned luxury clothing and accessories, not only offers a unique shopping experience but also champions sustainability. 

‍

By encouraging the reuse of high-quality luxury items, Lampoo is at the forefront of promoting a more responsible and environmentally conscious approach to fashion.‍

‍

Challenge: Navigating Marketing Measurement Complexities Amid Privacy Policy Changes

‍

Recent shifts in privacy policies have posed significant challenges for businesses trying to accurately evaluate the performance of various marketing channels. 

‍

Lampoo’s expansion relies heavily on strategic advertising to balance its platform's supply and demand dynamics, and these privacy changes have hindered Lampoo's ability to precisely assess advertising effectiveness using multi-touch attribution, creating uncertainties in their marketing approach.

‍

This led them to embrace a privacy-focused Marketing Mix Modeling (MMM) for more accurate advertising analysis and optimization.

‍

Marketing Mix Modeling is an econometric approach to measure Marketing Mix effectiveness without using user-level data and without being affected by privacy restrictions and cookies sunset. 

‍

This methodology uses only aggregate level data to measure:
‍

• Online and Offline Media ROI

• Efficiency saturation curves of each media

• Incremental contribution of Media and organic factors in the marketing mix

• Receive personalized media plans that maximize revenue

‍

The usual approach of hiring a consultancy for MMM presented drawbacks for Lampoo, notably high costs and a typical time-consuming process spanning 3 to 6 months. 

‍

The cost and the timing make this solution accessible only to big companies that can afford long waiting time and big measurement investments and was not a fit with Lampoo’s team.

‍

With an urgent need for swift, efficient solutions and Lampoo's dependence on a variety of marketing channels in a privacy-centric digital world, they needed an easy-to-use solution that could help them build an MMM fast and for a budget.

‍

Solution: Gaining Clear Incrementality Measurement Insights with Cassandra's Marketing Mix Modeling software FAST

‍

In this context, Lampoo, Cassandra and Meta Marketing Science team partnered together in a measurement project to evaluate the potential of MMM to drive effective decision making for a fast-moving digital first advertiser. 

‍

Lampoo’s objective was to quickly measure omnichannel campaign performance, detect where they were overspending and understand how they could improve their media mix.

They needed to be fast. 

‍

Cassandra was a perfect match, allowing Lampoo’s team to create their marketing mix model in-house, without the necessity for deep technical knowledge and seamless and rapid implementation.

‍

Implementation: Quick Data Integration and Model Development

‍

Integrating essential data into Cassandra’s system was expedited, completed within a week thanks to API connectors. 

‍

This was followed by an intensive three-week modeling phase, resulting in a comprehensive MMM that empowered Lampoo to analyze and refine their media strategy, improving decision-making processes and aligning all stakeholders in the company on each Marketing Channel’s ROI.

‍

Continual Improvement: Monthly Updates for Peak Precision

‍

After the initial build, the Model has been refreshed each month for three months. 

‍

The three monthly refresh cycles showed steady improvement in model performance.

The Marketing Mix Model has been evaluated based on 2 specific metrics:

R^2: A Metric that shows the percentage of variation in the output variable described by the model

NRMSE: Normalized Root Mean Squared Error - a Metric that shows the normalized average error between Sales Prediction and Actual Sales

‍

The first refresh had an R^2 of 79% and NRMSE of 12.3%,

the second improved to an R^2 of 80% and NRMSE of 10%,

and the third further improved to an R^2 of 84% and NRMSE of 7.4%,

‍

indicating a consistent enhancement in model fit and reduction in error.

‍

Optimization: Record ROAS for Black-Friday

During November Lampoo has ran the budget allocator through Cassandra receiving a new media plan that would maximize their revenue with -25% of the budget compared to the previous year.

They have implemented the strategy and after 4 weeks the results showed an increase of 80% YoY in ROAS.

Key takeaway

Lampoo’s experience with Cassandra’s MMM software illustrates the accessibility of complex marketing measurement analysis (previously accessible only by large corporations and substantial technical teams) to non-technical marketing teams to enable them to measure their media strategy in a privacy-centric era. 

‍

A word from the Lampoo's team member:

‍

‍

Cassandra’s MMM solution fits perfectly in this difficult scenario: for us, it is a great way of attributing the right credit (and hence the right budget) to each channel. We think the impact was already outstanding, but still, we are looking forward to seeing how much more it can impact our marketing activities as we foster further this collaboration. 

• Nico Giammarino - Digital Media Analytics Manager