Understanding promotional performance metrics for eCommerce

Promotions are a powerful way to drive sales and improve conversions on eCommerce platforms. However, successfully running a promotion requires more than offering discounts.

How do you know if it's working as well as it should? Understanding the right metrics can make or break your promotional strategy.

Let's break it down.

ROI vs. ROAS for smarter spending

When tracking promotion performance, ROI (Return on Investment) and ROAS (Return on Ad Spend) are critical metrics. But they aren't interchangeable. Knowing when and how to use each is essential for effective decision-making.

Basic definitions and formulas

ROI (Return on Investment): Measures the profitability of an investment relative to its cost.

ROAS (Return on Ad Spend): Focuses on revenue generated directly from ads relative to the amount spent on running them.

Key differences  

• ROI looks at profitability, considering all costs (e.g., production, distribution, and promotions).  
• ROAS focuses on ad performance, evaluating how much revenue advertising specifically generates.  

When to use ROI vs. ROAS

• Use ROI when assessing the overall value of a promotion or campaign.  
• Use ROAS to optimize spending on specific ad channels or campaigns.  

ROI and ROAS drawbacks

While ROI and ROAS are potent tools for evaluating success, they come with limitations.

• ROI is a broad metric that provides a high-level view but often fails to capture the nuances of individual campaigns, making it difficult to pinpoint specific areas for optimization.
• ROAS focuses solely on advertising spend and revenue, overlooking other operational costs such as production, shipping, or team salaries.

Relying too heavily on either can lead to a one-dimensional understanding of performance, which may result in missed opportunities for comprehensive growth. This is why a balanced approach, considering a variety of metrics, is essential for informed decision-making.

Real-world example  

A retailer runs a 20% discount promotion on a popular product. While ROAS shows a 400% return from ad spend, the overall ROI is just 20% once production and shipping costs are accounted for.

Misreading this data could lead to doubling down on an unsustainable discount strategy.

Attribution complexity and cross-channel challenges

eCommerce often involves multiple touchpoints—from search ads and emails to social media. Understanding where credit lies can be tricky but is essential for optimizing conversions.

Attribution model examples

• Last-click: Gives credit to the last interaction before purchase. Simple but often undervalues upper-funnel efforts like awareness ads.
• First-click: Credits the first interaction that drives users to your site. It highlights early discovery channels but ignores the impact of retargeting.  
• Linear: Distributes credit equally across all touchpoints. Fairer, but assumes all touchpoints have equal significance.
• Time-decay: This gives more weight to recent interactions. Focused on conversions but risks undervaluing mid-funnel marketing.
• Data-driven: This method uses machine learning to assign credit across touchpoints based on impact. It is the most accurate but complex to implement without proper resources.

Cross-channel challenges

Missing or siloed data is a common challenge. For example, a user may click on a Google ad, subscribe to your newsletter, and later purchase via an Instagram post.

Without proper attribution, you'd likely undervalue your email strategy.

Common mistakes to avoid

• Overreliance on one model: Using only one attribution model (e.g., last-click) can lead to misinformed spending decisions.  
• Neglecting channels: Ignoring assist channels like content marketing can skew ROI evaluations.

Incrementality testing for true impact

Your promotion does not drive every sale. Incrementality testing measures your promotion's true lift by separating what would have happened anyway.

What does incrementality measure?

Incrementality determines how many additional conversions or sales are directly attributable to your promotion, compared to a "business as usual" baseline.

Testing methodologies

• Holdout groups: Track the performance of a "test" group that receives the promotion and a "control" group that does not.  
• Geo-split testing: Running the promotion in some geographic regions and comparing the results with those of areas where it doesn't run.

Control group design

Ensure your control group properly represents your audience—this prevents bias in your results. For example, don't exclude top-performing segments from a control group, as this skews baseline data.

Statistical significance

To avoid jumping to conclusions, ensure your test results are statistically significant. A basic rule of thumb is to aim for a confidence level of 95%.

Implementation challenges

• Gathering clean, reliable data for the test group vs. control group.  
• Accounting for external factors (e.g., holidays or competitor promotions) that might influence results.
• Ensure consistent communication and messaging across test and control groups to avoid bias.  
• Allocating sufficient resources and time to conduct a thorough and accurate test.  
• Managing internal stakeholder expectations significantly when testing disrupts established workflows.  
• Monitoring and mitigating technical issues arising during the experiment, such as platform glitches or data tracking errors.

Bringing it all together for smarter decision-making

Promotions are powerful, but they require clarity in measurement. By understanding ROI, ROAS, attribution, and incrementality, eCommerce marketers can craft more effective strategies that optimize spending, boost conversions, and protect profit margins.

How metrics complement each other

• Use ROAS for tactical decisions (e.g., which channel to prioritize for an ad spend).  
• Use ROI for strategic overviews (e.g., which promotions drive long-term growth).  
• Test incrementality to validate the actual lift caused by promotions.  

Creating a balanced measurement approach

• Combine multiple attribution models for a more comprehensive view.
• Balance short-term revenue goals with long-term brand equity (e.g., not over-discounting or over-incentivizing).
• Incorporate customer feedback to refine campaigns and attribution models continuously.
• Leverage predictive analytics to anticipate future trends and adjust strategies proactively.
• Monitor cross-channel interactions to understand the holistic customer journey more clearly.
• Regularly benchmark performance against industry standards to identify areas for improvement.
• Invest in advanced tools that unify data from diverse sources for seamless analysis and actionable insights.  

Setting realistic expectations

No tool or metric tells the whole story. Set achievable performance benchmarks based on past data, industry averages, and customer behavior insights.

Making data-driven decisions

Set up systems to measure these metrics before launching your next promotion or discount campaign. Educate teams on the importance of accurate attribution and regularly audit your analytics setup to avoid costly errors.

Practical next steps

• Audit your existing promotions—do you know their true ROI or incrementality?  
• Experiment with different attribution models to uncover hidden insights.  
• Create your following campaign roadmap with clear KPIs for success.  

The customer-centric bottom line

A customer-centric approach involves evaluating your promotion's impact in the context of your customer's journey. This means considering multiple metrics, understanding cross-channel complexities, and leveraging incrementality testing to understand ROI and ROAS better.

By digging deeper into the data, you can make smarter decisions to improve both short-term performance and long-term growth. So, regarding metrics, remember: don't just measure what's easy; measure what matters most to your customers.