Start from the north star metric, which is usually the revenue. 1. Fix the single source of truth(SSOT) for north star metric. 2. Ensure the revenue in executive or financial report(SSOT) is adhered by the marketing channels. 3. There would be exceptions(billing delays, sessions post-processing), bake that into the system 4. Understand the deviation for channels to the SSOT, each channel has their own intricacy. 5. Understand which attribution system works best for business(or stakeholders) 6. Test and implement fast, to reduce mistrust.

Before diving into data, start with a clear KPI framework. Identify KPIs for long- and short-term goals. Understand how they correlate and influence each other to guide campaign evaluation. Next, set clear objectives, targets, and benchmarks for each channel and campaign. A branding campaign may excel in impressions, engagement, or earned media while driving minimal website traffic—does this mean it failed? Well, it depends what it was supposed to do. Manage expectations accordingly. Whilst different campaigns have different objectives, upper-funnel success should still connect to long-term outcomes, using tools like brand health trackers & MMM/CMM, you complete the picture and are able to act strategically.

1. Goals. Many will go straight to KPIs but KPIs miss the nuance. Know your goals and see if this data is even relevant to them before you invest the time digging further. If it’s not, keep this on a watch list to see if this is a pattern. If so, then start investigating. 2. Confirm the discrepancy. Sometimes data appears to conflict, but is pointing to the same issue. 3. Understand the platform and algorithm. Is there a difference because the measurement process is different or is there a mismatch in tactics? 4. Validate your root cause. This may be qualitative or quantitative. I like to do both to ensure I’m not just looking at metrics. The most effective marketing often cannot be measured.

Measure three times, cut once. OK? It is essential to approach data collection and analysis carefully. In the data analysis, we should not be biased toward a particular option. The analysis should be as professional and objective as possible.

Having clear objectives is what helps us to manage the data in the right way by taking the data that is useful and discarding the data that will not be relevant to achieve our goal. Markets are changeable, so focusing and segmenting can help to channel the research according to the different strategies you want to implement.

We need to have clear objectives and goals, and have solid communication strategy so that we can choose the best decision. And not to forget to keep the focus on the customer

First of all, we must know what we are aiming for. Clear KPIs help you get clearer results. 2- We must first evaluate the data source for quality and reliability. 3- We should check data consistency. Inconsistent data may indicate errors or biases. 4- We must analyze data correlation with other relevant metrics or industry benchmarks. This helps identify potential discrepancies or anomalies. 5- Visualize the data using charts, graphs, or heat maps. This can help identify patterns, trends, or outliers that may indicate data inconsistencies. And if all else fails... 6- Seek Expert Input

When faced with conflicting data from marketing channels, prioritize reliability by evaluating the source and methodology of each channel. Assess factors like sample size, data collection methods, and alignment with your target audience. Cross-check the data against historical performance and industry benchmarks to identify anomalies. Lean on first-party data, as it often provides the most accurate insights about customer behavior. Where discrepancies remain, A/B testing or controlled experiments can help validate findings. Ultimately, trust the channel that aligns with your goals, has a proven track record, and provides actionable, consistent insights.