Skip to content

Initial CRO Data Baseline

Complete this once per client

Start after Client CRO Readiness. Build this baseline once, record the sign-off in the client's private hub or task system, and use later Campaign Refreshes to keep it current. This page does not store client completion in the browser.

Use at least 90 complete days at daily granularity. For a newer store, use all available complete days and state the shorter coverage. Extend toward 12 months when history exists and seasonality materially affects the business. Record timezone and currency, and exclude partial current days.

1. Define the data contract

  • Record the client, store, market, and region scope.
  • Record timezone, currency, baseline dates, and source granularity.
  • Assign readiness, data, analysis, and approval owners.
  • Define the private source location and naming convention.
  • Log promotions, stockouts, outages, migrations, theme launches, and major traffic changes inside the window.
  • Record privacy, consent, and PII handling constraints.

2. Gather quantitative sources

  • Shopify totals and funnel data: sessions, visitors where available, add-to-cart rate, checkout reach, conversion rate, orders, sales, average order value, and revenue per session.
  • Shopify acquisition or attribution data by channel and day.
  • Landing-page or page-level sessions and conversion-quality metrics.
  • Product sales, orders, units, and average order value where available.
  • Device split.
  • Shopify Web Performance or equivalent LCP, INP, and CLS.
  • GA4 acquisition, landing-page, device, and ecommerce reporting needed to cross-check or fill a Shopify gap.
  • Klaviyo campaign, flow, form or popup, and attributed-revenue context relevant to the storefront journey.
  • Google Search Console page and query data relevant to organic entry points.
  • Prior A/B test and direct-rollout history, including current live control states.

Manual exports remain the default for private client reporting hubs. Record missing inputs instead of silently omitting them.

3. Gather behavioral and customer evidence

  • Gather heatmaps and session recordings for the priority pages and funnel surfaces.
  • Confirm at least 1,000 recorded sessions for each surface interpreted independently and 250 recorded sessions for each device segment interpreted separately.
  • Gather reviews, support tickets, surveys, interviews, and relevant ad comments.
  • Gather client and team notes explaining promotions, merchandising, operational issues, and audience changes.
  • Add competitor evidence only when it informs a real customer question or observed leak.

4. Validate data quality

  • Confirm source coverage, timezone, currency, granularity, and complete periods.
  • Merge relevant periods and deduplicate overlapping days, with the newer valid source winning.
  • Reconcile totals across files that should agree.
  • Identify partial periods, missing dates, and mismatched comparison windows.
  • Flag orders without sessions, impossible conversion rates, unknown attribution, null product titles, and mapping gaps.
  • Identify bots, staff or QA traffic, broken recordings, consent gaps, tracking outages, and duplicate events where practical.
  • Confirm metric definitions before comparing platforms.
  • Preserve raw sources and document every material cleaning or exclusion decision.

Do not invent a value to repair missing evidence. Record the gap and its effect on confidence.

5. Analyze the baseline

  • Analyze funnel loss, daily and weekly movement, channel quality, traffic mix, device performance, landing pages, products, offers, and technical stability.
  • Review behavioral patterns and customer-language themes.
  • Log external events that could explain unusual movement.
  • Separate traffic-quality problems from on-site conversion problems.
  • Use at least two evidence types before presenting a cause as a supported hypothesis.
  • Keep observation, interpretation, and recommendation distinct.

6. Publish the baseline pack

Record the source register, KPI and funnel baseline, segment views, ranked leaks, behavioral and customer themes, data-quality log, missing inputs, event timeline, historical benchmarks, sign-off date, and owner.

Close gate: The baseline is current, reproducible, caveated, privately recorded, and ready to support the first Data & Strategy campaign.

Continue to Data & Strategy