How RemoteTide Research Uses Job and Salary Data

August 31, 2026 · 6 min read

RemoteTide publishes two kinds of numbers: live counts from the job board and summarized salary ranges from the listings that include pay. This article explains where those numbers come from, how they are filtered, and how to read them without overstating what they mean. It is the methodology hub that other data articles on this site link back to.

Where the data comes from

Every job listing on RemoteTide starts as a live posting. The site ingests from multiple sources, including direct employer submissions and structured feeds from applicant-tracking systems and remote-job networks. Each listing is normalized into a common format with fields for title, category, location restrictions, skills, and, when the employer provides it, salary currency, minimum, and maximum.

The database only includes active listings. A listing is considered active if it has not passed its expiry date and has not been removed by the ingest pipeline. Because the inventory changes every day as new posts arrive and old ones expire, every statistical extract is tied to a specific moment. The numbers in this article were extracted on August 31, 2026.

The current snapshot

As of August 31, 2026, RemoteTide held 11,005 active job listings. Of those, 34.5 percent included any salary information, meaning both a minimum and a maximum were present regardless of currency. A smaller share, 22.4 percent, had a usable USD salary that passed all four quality rules described below.

These are the only site-wide statistics this article cites. Any other RemoteTide data article should show sample size and extraction date next to each figure, as done here.

What "usable" means: the four honesty rules

Before a listing contributes to a median or percentile, it must pass four filters. These rules exist to keep the salary numbers honest and comparable.

1. Currency must be null or USD. Remote jobs are posted in many currencies. Mixing USD, EUR, GBP, and others would distort any summary, so only listings with a null currency or an explicit USD currency are counted. Null currency is treated as USD by the same convention used in the live salary pipeline.

2. Both ends of the range must be present. A listing must have both salary_min and salary_max. Partial ranges, single numbers, and free-text pay descriptions are excluded because they cannot be compared mathematically.

3. The midpoint must sit inside a bounded band. For each usable range, we calculate a midpoint: (salary_min + salary_max) / 2. The midpoint must fall between $20,000 and $600,000. This catches obvious data-entry errors, typos, and extreme outliers that would otherwise pull a summary far away from the real market.

4. Role-level medians require at least 15 usable listings. The salary calculator and role-specific reports only show a median when the matching set reaches 15 usable listings. Small samples are too easy to sway with one or two unusual postings. Category-wide snapshots still show the median from the full usable set and flag any sample below 30 as small.

The first three rules and the midpoint method apply to every salary figure RemoteTide publishes, whether in a blog snapshot, a category salary page, or the salary calculator. The 15-listing minimum applies to role-level reports and calculator entries.

How we compute medians and percentiles

Once the four filters are applied, the remaining midpoints are sorted from lowest to highest. All percentile calculations use those sorted midpoints directly.

• Median. If the sample has an odd number of values, the median is the middle value. If the sample has an even number of values, the median is the average of the two middle values, rounded to the nearest dollar.
• 25th percentile (p25). We take the value at the ceiling of 0.25 * n in the sorted list, where n is the sample size. This means we round the index up to the next whole number.
• 75th percentile (p75). We take the value at the ceiling of 0.75 * n in the sorted list, using the same ceiling-index method.

The result is a three-number summary: p25, median, p75. Together, these describe where the middle half of posted midpoint salaries falls. They do not describe the full spread, and they do not predict what any one company will offer.

Why every snapshot carries an extraction date

Job boards are living inventories. A company can post a batch of roles on Monday, let them expire on Friday, or update a range overnight. A statistic that was true on August 31, 2026, may not be true on September 7, 2026.

The extraction date is not a formality. It tells the reader that the number is a photograph, not a permanent fact. When you see a RemoteTide data article, the date next to each statistic is the date the underlying query was run. The live salary pages and calculator refresh from the same dataset on an ongoing basis, but they still reflect the current inventory at the time you load them.

What the numbers can tell you

Used correctly, the salary summaries give you three useful signals.

First, they show the shape of the market for roles that actually disclose pay. If you are comparing offers or setting expectations, the median and percentile band tell you where the transparent portion of the market currently sits.

Second, the transparency rate, the share of listings with any salary info, tells you how much you can rely on posted ranges in a given category. If only a small share of listings include pay, you will need to do more independent research when a post is silent.

Third, the consistency of the methodology lets you compare categories fairly. Because every category uses the same four rules, a higher median in one category is not a counting artifact.

For example, the remote developer salary guide applies the same four rules to engineering listings that this article applies to the full index.

What the numbers cannot tell you

The summaries have clear limits. They do not include total compensation. Equity, bonuses, commissions, benefits, and perks are not part of the posted range and are not captured in the midpoint calculation. A role with a lower base and generous equity may look worse than it is, while a high base with no equity may look better.

They do not adjust for location policy. A globally remote role and a US-only role may both list USD salaries, but the candidate pools and cost assumptions behind them can differ. The numbers treat both as USD midpoints.

They do not adjust for seniority, company size, industry, or contract type unless those dimensions are built into the category filter. A single category can contain entry-level and senior roles side by side, so the percentile band is wide by design.

Finally, they cannot predict your offer. They describe what employers have chosen to publish, not what you will negotiate. Use them as a reference point, not a quote.

How the live salary tools use the same rules

The same four filters and the same percentile method power the live salary reports at /salary/[category] and the salary calculator. Blog snapshots and tools share one methodology.

When you visit a category salary page, the numbers are computed from the current active listings in that category. When you use the salary calculator, the estimate is drawn from the same filtered midpoint distribution. The only difference is presentation: the category pages show a static summary, while the calculator lets you select role, level, and location to narrow the comparison.

Because the rules are identical, a number you read in a blog snapshot should match the logic you see on the live tools. If the live number differs from a published snapshot, the difference is almost always the extraction date, not a change in method.

How to use this methodology in your search

Start with the live inventory. Browse current remote jobs to see which roles include salary ranges today. Then visit the relevant salary page, such as the remote developer salary guide, to see the current midpoint distribution for the category. If you want a more tailored comparison, use the salary calculator to narrow the estimate by role and location.

When a listing has no range, do not assume it pays below the median. Silence just means the employer did not publish a number. Use the category median as your anchor and ask for the company's band directly during the interview process.

The bottom line

RemoteTide's salary numbers come from live job listings, filtered through four consistency rules, and summarized with a simple percentile method. As of August 31, 2026, the index contains 11,005 active listings, of which 34.5 percent include any salary info and 22.4 percent have a usable USD range.

Those figures describe the transparent part of the market at a single point in time. They are useful for orientation, comparison, and negotiation preparation, but they are not a prediction of any individual offer. For the latest numbers, browse remote jobs, check the relevant salary page, or use the salary calculator.

Want a weekly digest of fresh remote opportunities? Subscribe to the Tuesday Drop, or read more guides on the RemoteTide blog.