Build a Traffic Rankings Shortlist Without Losing Context

You import a traffic dataset, filter it to your market, and find a few websites worth investigating. Before sending that shortlist to someone else, check what the screen and the download actually contain.

Shane’s Traffic Rankings gives you a filtered table, summary cards and a CSV export. In the current version, those outputs do not all use the same set of rows. Understanding the difference helps you prepare a clear research handoff instead of a spreadsheet that surprises its recipient.

Start with one comparable dataset

The tool works with supplied data. It does not fetch traffic figures from a provider or discover competing websites for you. The CSV template contains eight columns: rank, domain, visits, desktop, mobile, change, industry and market.

Before importing, choose records with a consistent source, reporting period, geography and measurement method. Keep a separate note of those details because the template has no dedicated source or reporting-period columns. The table labels visits as monthly, so make sure your input matches that period.

Use plain numeric values. Enter 180000 rather than 180,000 for visits and 45 rather than 45% for a share. Missing or nonnumeric values become zero in the current importer. Keep unknown figures in your original research file and resolve them before using the tool’s totals; an imported zero may otherwise look like a measured result.

See what the filters change

Industry and market selections filter the rows you imported. They do not request a new market dataset. The domain search further narrows the view using a case-insensitive text match.

The summary cards follow this filtered set:

• Websites shown counts matching rows.
• Combined visits adds the visits values on those rows.
• Top domain shows the first row after the selected sort is applied.

That last point matters. Top domain follows your current ordering, so its meaning changes when you switch from rank to visits or change.

A small example you can check

The following rows are fictional arithmetic examples, not observations about real websites:

rank,domain,visits,desktop,mobile,change,industry,market
2,alpha.example,180000,40,60,8,Retail,United States
1,beta.example,120000,55,45,-2,Retail,United States
3,gamma.example,90000,30,70,22,Software,Canada

With all three rows visible and Sort by rank selected, beta.example appears first because its supplied rank is 1. Websites shown is 3 and Combined visits is 390000.

Switch to Sort by visits and alpha.example becomes the Top domain with 180000 visits. Switch to Sort by change and gamma.example becomes the Top domain with a change value of 22. Neither action recalculates the imported rank or verifies what the change percentage measures.

Now select Retail and United States. The view contains alpha.example and beta.example, with two rows and 300000 combined visits. Under Sort by change, alpha.example appears first. Under Sort by rank, beta.example appears first again.

Describe the result precisely in your notes: “First by supplied rank among the imported US retail rows” communicates more than “the top website.” The wording preserves the dataset, filter and sorting context.

Filtering does not create new ranks

The rank column keeps the values supplied in the CSV. A filtered list may start at rank 14 or contain gaps such as 2, 7 and 19. Those values describe the imported ranking, not a newly calculated ranking within your filter.

If a rank is missing, zero or nonnumeric, the current importer substitutes a position based on how many rows are already stored. That fallback is an import-order convenience. Check the source rank before using it in a report or comparison.

Combined visits also deserves a careful label. It is the sum of the visible row values, not a calculation of distinct people across the websites. Use it to describe the selected dataset and keep your source’s metric definition with the handoff.

Check the CSV before sharing your shortlist

The current Export results button downloads all stored rows in their original import order. It does not restrict the file to the visible filters or apply the selected sort.

In the example, filtering to the two US retail rows still exports all three rows, including gamma.example. Sorting by change does not move gamma.example to the top of the downloaded file; the file retains alpha.example, beta.example and gamma.example in that order.

For a shortlist, open the export in a spreadsheet, apply the intended filters and order there, and save a separate clearly named copy. Check the final row count and domains before sharing. Keep the full export separately so you can trace the shortlist back to its source.

Also inspect the columns if your labels contain commas or line breaks. The current exporter joins values with commas without adding CSV quoting around those fields. Simple, consistent labels are easier to carry through this workflow reliably.

Avoid importing the same snapshot twice

Import adds rows to the data already stored in the browser. It does not replace the previous import or merge duplicate domains. Importing the three-row example twice creates six stored rows and, with no filters, 780000 combined visits.

That higher total reflects duplicate rows, not traffic growth. Check the stored row count before adding another file. Keep a clean source spreadsheet and treat each import as an addition, especially when working with refreshed monthly exports.

The same caution applies to mixing reporting periods. Without a period column in the tool’s standard record, two snapshots can look like duplicate entries rather than a time series. Compare periods in a clearly labelled external worksheet instead of adding both and interpreting the combined total as one month.

Make the handoff easy to audit

A useful shortlist can be brief. Include the selected domains, source and reporting period, the industry and market filters, the ordering used, and the question you want to investigate next. Add the cleaned shortlist file, with its row count checked against your intended selection.

Shane’s Traffic Rankings can help you explore a supplied dataset quickly. A few deliberate checks around filters, sort order and export scope make the resulting research easier for someone else to understand and reproduce.