Shopping feeds reporting dashboard metrics explained in one place

A shopping-feed dashboard is useful when it helps someone choose what to investigate, rather than simply displaying every available total. Read its measures as connected evidence: feed status, product exposure, advertising cost, recorded orders and retained value each describe a different part of the result, and none explains the whole story alone.

What decision should the dashboard help you make?

Begin with the decision the team needs to make, then select figures that inform it. Deciding whether a channel earns its place calls for order, cost and margin information as well as activity measures; feed-status numbers help identify a separate catalogue issue.

Agree what each label means before comparing products or channels. Keep the date range, product ID and channel scope consistent, and document the order event used for conversion. The nine key metrics give the detailed calculations, and demand signals show how to read product activity.

In brief, impressions count recorded displays, CTR is clicks divided by impressions, and conversion rate is completed orders divided by clicks. Average order value is revenue divided by completed orders. CPC is ad spend divided by clicks, and ROAS is revenue divided by ad spend. Margin per order deducts returns and the listed costs before dividing by completed orders.

Worked example: every input shown

For example, a small shop sells rucksacks through Google Shopping for one month. These figures are illustrative, not benchmarks. Substitute your own.

Retained revenue is £10,000 less £1,000 returns, which is £9,000. Deduct cost of goods sold, ad spend and shipping: £9,000 - £4,500 - £1,200 - £600 = £2,700. Margin per order is £2,700 / 100, or £27. A high ROAS of 8.3 therefore sits beside a £27 margin per order.

Keep revenue, cost of goods sold, advertising spend, returns and shipping together for each product and channel. State whether revenue is before or after returns, then use that basis for average order value and margin. If the revenue figure already excludes returns, do not subtract them a second time.

Where each figure comes from

Clicks, impressions, CTR, cost, CPC and conversion value come from Google Ads, and Merchant Center performance reports show clicks and impressions for free listings as well. The Products section of Merchant Center lists items by status and issue, which gives approval and disapproval counts. Orders and revenue come from your shop platform or GA4 ecommerce reporting.

Cost of goods sold, returns and shipping come from your own accounts or stock system. Looker Studio can join these sources in one view, and a spreadsheet works for a small catalogue. Platform ROAS counts conversion value before returns, so compare it with your net figure before judging a channel.

How should I read activity and sales together?

Read exposure, clicks and orders in sequence, but do not treat the sequence as proof of cause. Impressions show exposure, CTR shows the share that became clicks, and conversion rate shows the share of clicks that became recorded orders. Order value and margin add different views of the sales.

Compare the same products, channels and dates before investigating a change. Feed-health measures describe catalogue readiness, while sales measures describe recorded outcomes. The measurement and reporting guide sets out the wider context, and attribution methods explain why the way of assigning channel credit should be stated alongside channel totals.

What do approval and error counts tell me?

Approval and disapproval figures describe the status of submitted products, while feed errors count reported processing problems. Keep these measures distinct from clicks and orders: they describe feed condition, not sales performance. Use the same submitted-product scope and reporting period when calculating each rate.

For approval rate, divide approved products by submitted products; for disapproval rate, divide disapproved products by that same set. An error count is the number of reported feed errors, not a sales measure. See how UK merchants fix Merchant Centre disapprovals for the usual causes.

How do I know the totals can be checked?

Reliable totals depend on repeatable joins between feed, order and advertising records. Align product identifiers, channel names and dates, and record each source, field definition, refresh time and owner. Check those joins before presenting a dashboard total as complete, and keep the rules so another person can repeat the check.

For a smaller catalogue, export the Merchant Center product report and order list as CSV, then join them on product ID in Excel or Google Sheets. Larger catalogues can use a warehouse such as BigQuery. The feed-tool options include the Merchant API.

Use the UK Statistics Authority's quality assurance guidance to frame checks of the source and the match. Test the setup against a peak trading week. If totals fail to reconcile, inspect what the dashboard counts before changing its calculation.

How should I compare channels fairly?

State how orders are assigned to channels before interpreting channel totals as a comparison. A dashboard can show the figures side by side, but readers need the crediting method to understand what each total represents. Keep product, date and channel scope aligned when making that comparison.

A channel view should distinguish turnover from retained value and make its revenue basis visible. When totals do not reconcile, check the underlying counts and attribution approach before revising the calculation.

How current should each dashboard view be?

Choose a refresh rhythm to suit the decision and display a refresh time on each view. Daily figures can support stock and pricing checks, weekly updates suit margin review, and monthly views help with planning. The timestamp tells readers when the figures were current.

Different measures need not change at the same pace. Keep refresh timing visible, so readers do not mistake an older total for a current result.

What may I share with a partner?

Share only the fields an agency, marketplace or fulfilment partner needs for the agreed purpose. Name an owner, set an access expiry date and use aggregated information when it is sufficient. A narrower view is preferable to sending a complete export.

If a view may contain customer data, limit fields to those needed for the partner's task and review access when the catalogue or partner arrangement changes.

What should I do first?

Start with one decision, then check that its measures are defined and can be traced to reconciled records. Expand the dashboard only when the team needs additional detail to act.