Data literacy · 4 min read
Read a remote-hiring index without inventing a trend
How to interpret live samples, category counts and salary coverage without treating a rolling feed as a census.
Ask what the denominator represents
Before interpreting a chart, identify the records it summarises. The WRILS Remote Work Index uses the current Jobicy sample requested by the site. It is not a count of every remote vacancy, every employer or every opportunity in a country. The denominator is the returned sample, which can change as records enter or leave the recent feed.
This distinction matters even when the interface looks precise. A figure displayed as a percentage can still describe a small or selective group. The number becomes useful when its scope is visible, not when its label sounds like a complete market measure. Read the accompanying method before comparing it with another website’s catalogue total.
A refresh is not a historical observation series
An hourly refresh replaces or updates the current view. It does not automatically create a historical archive collected under stable rules. If the composition changes between two visits, the change could reflect timing, a large employer’s import or the publication window rather than a movement in demand for a profession.
A defensible time series needs saved snapshots, consistent filters, deduplication and a policy for closed or missing roles. Without that foundation, a line connecting two sample values would create an appearance of evidence the data does not support. WRILS therefore presents the current distribution without claiming a measured historical trend.
Read category totals as labels, not separate populations
A vacancy may have more than one category, and categories can describe different levels of a professional hierarchy. That means the sum of category counts can exceed the number of unique jobs. A broad engineering label and a narrow development label may overlap. Treating them as independent populations can exaggerate the size of the sample.
Use the distribution to decide where to investigate, then inspect actual roles. A candidate looking for quality assurance work should not conclude that a broad technology category represents the same opportunity set. The chart is a navigation clue, while the vacancy provides the evidence about required work and eligibility.
Location labels are not national labour statistics
A role labelled worldwide, Europe or United States is describing an employer’s hiring boundary, not the number of jobs physically located in each place. Those labels can overlap. An international remote role does not become one separate vacancy for every eligible country, and a country-limited role does not establish the size of that country’s remote economy.
For an individual search, the useful question is simpler: can this employer hire me where I am, and can I meet the schedule? Use the location information to qualify the vacancy rather than turning a region count into a ranking of where a person should move.
Missing salary data changes what can be concluded
The salary-coverage figure indicates whether sampled records include salary information, not whether compensation is good. A category with more missing data can look different from one where employers publish ranges. Wide bands, currencies and pay periods also prevent a simple maximum from becoming a universal benchmark.
If you want to compare pay, create a like-for-like subset and preserve the missing values. Do not replace a blank with zero, and do not treat the top of a range as the expected offer. The salary-transparency guide explains the additional questions needed before the figures become comparable.
Use the sample for decisions it can support
The index can help a reader understand what appears in the current source, find a category to inspect and see where salary context is often missing. It cannot by itself show a hiring boom, prove that one country is better for remote work or predict the candidate’s chance of an interview.
A responsible interpretation names both a useful observation and its limit. For example, a category is prominent in the current sample, so it is worth examining the roles; that does not establish growth in that profession. Keeping those two statements together preserves the value of a live index without asking it to answer questions for which no historical evidence has been collected.
Quick answers
Common questions
Why can category counts add up to more than the number of jobs?
One vacancy can have multiple category labels. Category counts should not be assumed to describe separate, non-overlapping populations.
Can two hourly snapshots establish a hiring trend?
Not by themselves. A trend needs a stable historical collection method, deduplication and a clear policy for closed or missing records.
Sources and review scope
Public sources checked on October 17, 2026. Product claims are attributed to their publishers. We did not purchase subscriptions, submit applications or measure hiring outcomes for this review.