Marketing Data Analysis: From a Business Question to a Tested Decision
Shusaku Yosa
Marketing data analysis turns a business question into evidence, plausible explanations and a decision. Start with the decision you need to change, not the software you want to use.
Instead of “leads are down,” ask: “Did qualified inquiries decline because traffic, traffic mix or completion changed, and what should we adjust next month?” Define the outcome, comparable periods, audience and observation limits.
Check data quality before performance
Review missing records, duplicates, hostnames, date ranges, time zones, outcome definitions, exclusions and whether the period is final. If a decline coincides with a form release, check successful submissions and tracking before assuming weaker demand.
Check | Example problem | Response |
|---|---|---|
Completeness | CRM records exist but analytics shows none | Reconcile tracking and transfer |
Duplication | Reloading creates repeated completion events | Provide a reproducible case to the implementer |
Period | One source contains a partial month | Align finalized periods |
Scope | Test or unrelated-host activity is included | State and enforce the reporting scope |
Decompose a fictional lead decline
Here conversions are successful submissions and the denominator is sessions. Every number is fictional.
Period and traffic group | Sessions | Conversions | Conversion rate |
|---|---|---|---|
Previous A | 3,000 | 150 | 5% |
Previous B | 2,000 | 20 | 1% |
Previous total | 5,000 | 170 | 3.4% |
Current A | 1,000 | 50 | 5% |
Current B | 4,000 | 40 | 1% |
Current total | 5,000 | 90 | 1.8% |
Total traffic is unchanged, but conversions fall from 170 to 90. Neither group's conversion rate changes; the overall decline is explained mathematically by the changed mix. Redesigning every form might miss the main issue in this example.
In real data, investigate why A declined: budget allocation, demand, search visibility, landing pages or a tracking failure. Explaining an aggregate calculation is different from identifying the business cause.
Test competing explanations
Hypothesis | Evidence to examine | Possible counterevidence |
|---|---|---|
High-intent traffic declined | Queries, source and landing pages | Large conversion changes within the same group |
A form failure occurred | Device completion, errors and submission records | Stable completion with lower traffic only |
Seasonality reduced demand | Prior-year period, business days and demand | A sudden change isolated to one release |
Customer mix changed | Industry, size, region and sales acceptance | Stable mix but slower follow-up |
Look for evidence against a favored explanation. Higher sales among people exposed to an ad do not prove the ad caused the difference; those people may already have had stronger intent.
Convert the analysis into a test
Before changing the experience, specify the audience, primary measure, guardrails, cost, period and comparison method. Where a controlled comparison is impractical, record seasonality and concurrent changes. A higher rate from a small number of conversions does not by itself establish effectiveness.
The decision memo should contain the question, observation and denominator, data-quality checks, leading explanation, alternatives, action, owner, deadline and reevaluation rule. A chart is an intermediate deliverable; the analysis is useful when the next decision and remaining uncertainty are clear. For presentation, see the dashboard design guide.
Related practical guides
- How to Build a Marketing Dashboard: Metrics, Layout and Data Quality
- Ad Performance Measurement: Metrics, Attribution and a Worked Example
- Marketing PDCA: Plan, Test, Review and Decide with a Practical Worksheet
- B2B Marketing KPIs: Funnel Definitions, Sales Handoff and Targets
Connect execution, costs and reviews
Xtrategy supports project management, monthly budgets and actuals, and customer and deal records. Use a 30-day free trial to evaluate the workflow against your team’s needs.