The digital shelf is everything a shopper sees between typing a search term and adding a product to the basket. For a brand, it spans dozens of retailers and marketplaces, each with its own search ranking, content template and seller model. Measuring it well means choosing a small number of indicators, defining them precisely and collecting them consistently enough to show change over time.
This playbook is written for brand, e-commerce and sales teams building or rebuilding a digital shelf programme. It focuses on what to measure and how to collect it, rather than on any particular dashboard.
The five pillars
Most useful digital shelf programmes rest on five pillars. Each answers a distinct commercial question, and each needs different data.
| Pillar | Question it answers | Data captured | Typical cadence |
|---|---|---|---|
| Search visibility | Do shoppers find us for the terms that matter? | Rank position, sponsored flag, page, result count per keyword | Daily, by retailer and location |
| Content quality | Is our product page complete and correct? | Title, images, bullets, description, specifications, video presence | Weekly |
| Ratings and reviews | What do shoppers say, and is it changing? | Star rating, review count, review text, date, verified flag | Daily or weekly |
| Availability | Can shoppers buy the product today? | Stock status, delivery promise, location, variant availability | Daily, hourly for key lines |
| Offer ownership | Who wins the sale on shared listings? | Featured seller, price, fulfilment type, number of offers | Several times a day |
Search visibility
Share of search measures how much of the first results page your products occupy for a defined set of keywords. The metric is only meaningful if the keyword list is stable and representative. Build it from three sources: category terms shoppers use ("wireless earbuds"), benefit terms ("noise cancelling earbuds") and competitor brand terms. Keep branded searches for your own products separate, as they inflate the figure without saying much about competitive position.
- Record organic and sponsored positions separately. Many retailers mix them in the same grid.
- Capture results from a consistent location and device type, since rankings vary by both.
- Fix the number of positions you measure, such as the first 24 or 48 results, and keep it constant.
- Store the full result list, not just your own positions, so competitor movement is visible.
Content quality
A content audit compares what each retailer displays against the brand's approved master content. The aim is not a perfect score on every page, which is rarely achievable, but a prioritised list of fixes on the lines and retailers that matter most.
Content audit checklist
- Title contains brand, product type and the key differentiator in the approved order.
- Minimum image count met, with the primary image matching the current pack design.
- Bullet points or feature list present and aligned with approved claims.
- Specifications complete: size, weight, capacity, compatibility, warranty.
- No outdated claims, discontinued variants or incorrect pack sizes.
- Arabic or other local-language content present where the retailer supports it.
- Rich content or video present where the retailer supports it.
Score each item against this checklist and weight the score by commercial importance. A missing image on a top seller at a major retailer matters more than a missing specification on a slow line.
Ratings and reviews
Average rating and review count are lagging indicators. The text of recent reviews is where early warnings appear: a packaging fault, a delivery problem in one city, a confusing instruction. Collect review text with dates so you can track themes over time, and separate reviews by retailer, since each platform attracts a different shopper profile. Personal data should be kept to a minimum. Reviewer names are rarely needed for analysis and can usually be excluded at collection.
A digital shelf score is only useful if it can be traced back to a specific page, retailer and date that someone can fix.
Availability
Out-of-stock and delayed-delivery states are the most direct form of lost sales on the digital shelf. Availability should be captured by location where retailers localise stock, which is increasingly common in grocery and quick commerce. Distinguish between a product marked out of stock, a product that has disappeared from the listing entirely and a product available only with an extended delivery date. Each has a different cause and a different owner.
Offer ownership
On marketplaces where several sellers share a single product page, the featured offer determines who captures most of the sales. Monitoring which seller holds that position, at what price and with what fulfilment method shows whether authorised partners are winning, and whether unauthorised sellers or aggressive pricing are displacing them. This pillar connects directly to brand protection and pricing policy work.
Setting up the programme
Setup checklist
- Agree the retailer list per country, with a named owner for each relationship.
- Build and freeze the keyword list, with a quarterly review date.
- Load master content and approved imagery as the audit reference.
- Map each product to its listing on each retailer, and record the identifiers used.
- Define locations for capture where results or stock vary by city.
- Decide the scoring method for each pillar and document it.
- Set alert thresholds for availability and offer ownership.
- Agree the delivery route: warehouse table, API or scheduled report.
Reading the results
Look at trends before levels. A share of search figure in isolation invites debate about methodology. A decline over six weeks on one retailer invites action. Pair each pillar with a short list of owners and actions, and review it on a fixed rhythm. Most teams find a weekly operating review and a monthly management summary strike the right balance.
Finally, resist the temptation to add pillars before the first five are working. A small, trusted set of indicators changes behaviour. A large, disputed one does not.