How we rank tools. And what we leave out.

• Every number has a source and a date. A price, limit, platform list or feature is stored as a value, the URL it was read from and the date it was checked. Where the vendor page shows it, we also keep the exact quote.
• Missing means missing. If we could not verify something, a table shows a dash (–) and a tool page counts it in one line instead of listing it. A dash means unknown, never “no”. We never fill a gap with an estimate, and a criterion with no data is left out of the score instead of being scored as zero.
• Rankings are computed. A ranking is the output of a scorecard applied to the stored facts. The same facts always produce the same order, and one tool's score never depends on which other tools are listed.
• Text cannot contradict the data. In our guides, prices, scores and counts are inserted from the same data as the tool pages. A guide that types a number by hand does not pass our publishing checks.

We read the vendor's own public pages:

• pricing pages (plans, prices, billing periods, published limits, trials);
• product, feature and integration pages, and documentation (supported platforms, features, API, export);
• help-centre articles and changelogs, for limits and for changes after a plan is first listed;
• the vendor's own public statements, such as terms and security pages, where a listing needs them.

We do not scrape or copy review sites. G2, Capterra, AlternativeTo and Software Advice prohibit it, and we do not use their ratings, review counts or text. We do not publish star ratings, user counts or testimonials, because we have no first-party source for them.

• Snapshot. The page is fetched and stored. Facts are read from that stored snapshot, not from memory.
• Quote and source. Each fact keeps the URL it came from, the date it was checked and, where the page states it as text, the exact quote that proves it.
• Grounding check. If a model helped extract a number or limit, the number must appear verbatim in the stored snapshot. A value that does not is rejected, not rounded or guessed.
• Re-verification. Pricing pages are fetched again on a schedule and compared with the stored snapshot. A change updates the check date and is recorded in the changelog, so a price that moved shows both the new value and the date we saw it.
• Dates are shown. Tables show a “Checked” date per tool, and each guide shows the newest check date behind it. A fact we have not re-checked recently keeps its old date; the date is never refreshed without a new read.
• Only positive statements are read from text. Vendors rarely say what they do not offer, so most features are unknown rather than missing. A feature counts as missing only when the vendor's own page says so. When we read a feature, an integration count or an API from a vendor's page without a model, it must be a clear positive sentence on that page (kept as the quote); menu labels, negations (“no API”, “coming soon”) and upgrade prompts are ignored.

Each category has one scorecard: a list of criteria, each with a weight and a plain-language description of what it measures. Every criterion turns the stored facts into a score from 0 to 100, with the number it was based on and its evidence links shown next to it. A tool's total is the weighted average of the criteria that have data:

Total = (sum of weight × score for each criterion with data) ÷ (sum of the weights of those criteria)

Illustration with invented numbers, not data: a criterion weighted 20 where a tool scores 80, and one weighted 10 where it scores 60, give (80 × 20 + 60 × 10) ÷ 30 = 73 on those two. Every tool in a ranking is held to the same criteria and weights. Nobody gets custom weights.

The criteria, weights and what each one measures, per category (weights shown as a share of the whole scorecard):

Every scorecard starts with the same 5 core criteria, worth 75 of its 100 weight points. They are built from facts most vendors publish on their pricing and feature pages, so nearly every tool with a verified pricing page can be scored on them. Each category then adds one to three criteria of its own for the other 25 points (shown as “Category” in the tables above).

• Entry price (25%). Price per month of the cheapest paid plan, against the prices of other social media scheduling tools on the same billing basis (monthly-billed prices against monthly, annual-billed against annual). Log scale. Billed monthly: $9/month or less = 100, $49 or more = 0. Only billed annually: $9/month or less = 100, $58 or more = 0. Only free plans = 100. Usage-priced and one-time plans are not scored.
• Free plan or trial (15%). Whether you can use the tool before paying: a free plan, or failing that a free trial, as shown on the vendor's pricing page. Free plan = 100. No free plan but a free trial = 60. The pricing page lists plans but shows neither = 0. No plans read = not scored.
• Pricing transparency (10%). How much of the pricing is public: the share of plans with a published price ("contact sales" counts against), and whether there are at least 3 published tiers. 80 x (plans with a public price / plans listed) + 20 x (public plans, up to 3) / 3. No plans read = not scored.
• Verified features (15%). How many of the 19 features we track for this category the vendor's own pages verifiably show (posting queue, bulk scheduling (csv or multi-post), calendar view, first-comment scheduling, direct auto-publishing, recurring posts, media library, ai caption assistant, post and account analytics, report exports (pdf/csv), best-time-to-post suggestions, approval workflows, user roles and permissions, separate client workspaces, unified social inbox, white-label reports, link tracking (utm / short links), content repurposing, ugc collection). This counts verified features only: a tool that documents less scores lower here, and only here. Linear: verified features / 10 (half the tracked list) x 100, capped at 100. No tracked feature verified = not scored.
• API, integrations and export (10%). Whether a public API exists, how many integrations the vendor states, and whether the vendor documents a way to export your data. Average of the parts with data: public API yes = 100, no = 0; integrations on a log scale (100 or more = 100); a documented data export = 100. None verified = not scored.

Price anchors. The entry-price and cost-per-user scales are fixed numbers, not live comparisons: for each category and billing basis we took the 10th and the 90th percentile of the verified prices in our data on October 14, 2026 (the cheapest tenth scores 100, the most expensive tenth 0, on a log scale between). A category with fewer than 8 prices on a basis uses the percentiles of all tools, and a range narrower than 4x is widened to 4x. Because the anchors are fixed, adding a tool never changes anyone else's score; changing them is a new scorecard version. The entry-price formula of each category above shows its anchors.

Monthly and annual prices are never mixed. A monthly-billed price is scored against the category's monthly-billed prices. A vendor that publishes only an annual-billed price is scored on that price against the category's annual-billed prices, and the basis line says so.

Verified features measure documentation. The feature criterion counts the category's tracked features that the vendor's own pages verifiably show. A tool that documents less scores lower on that one criterion only; nothing else assumes a missing feature is absent.

Coverage is the share of a scorecard's total weight that has verified data for a tool. When a criterion has no data it is left out and the remaining weights are re-normalised, so a gap neither helps nor hurts a tool directly. Every score shows its coverage and how many criteria it is based on (for example “6 of 8 criteria”).

A tool with under 60% of the weight covered is not ranked. It stays listed, marked “not ranked: insufficient verified data”, and its score is not shown as a rank, because a total built from a small part of the scorecard would not be comparable with the others. A tool with no verified price and fewer than 2 verified features is not ranked whatever its coverage, and is marked “needs a verified price”.

Why 60% and not 70%: the core is 75% of every scorecard, and the category criteria need published limits, platform lists or counts that only a minority of vendors state. A tool fully covered on entry price, free plan or trial, pricing transparency and verified features reaches 65% without an API or export statement, so a 70% rule left many well-documented tools unranked. 60% still needs at least three core criteria with data, and we will not go lower.

• It is not a quality or satisfaction rating. It measures published price, free access, pricing transparency, documented features and the category's published limits, on one scale for every tool. It says nothing about how good the output is, how reliable the service is or how happy customers are.
• It is not a test. Unless a listing carries a dated testing record we have not used the tool; see “What we do not test” below.
• A low score is not a verdict that a tool is bad. It can mean the tool costs more than others in its category, or that its vendor publishes less: the feature criterion counts what is documented, not what exists.
• Small gaps are not meaningful. Two tools a few points apart are close on the published data; read the basis of each criterion rather than the order alone. A score based on fewer criteria (lower coverage) is less certain than one based on all of them.
• It may not fit your team. The reference teams in the formulas (for example 3 accounts or 3 users) may not match yours; the numbers behind each criterion are shown so you can weigh them yourself.
• It is not bought. No vendor, advertiser or affiliate relationship changes a score, a weight or an order.

TrackMedia is a product of the team behind SaaSHunter. It is listed, scored and ranked by exactly the same code, criteria and weights as every other tool, from the same kinds of public sources, and is marked wherever it appears in a ranking.

No score, weight or order is adjusted for our own products, for advertisers or for anyone who contacts us.

Our data is read from vendor pages. Unless a tool's listing carries a dated testing record, we have not used the tool hands-on, and no guide says otherwise. That means we do not measure:

• how reliably a scheduler actually publishes, or how fast an app runs;
• customer support quality, or how easy the interface is to use;
• whether a feature works as well as the vendor describes it;
• the quality of AI output, such as captions, images, video or voices;
• real-world reach or engagement results from using a tool.

A score tells you what a vendor publishes and what it costs, on the same scale for every tool. It does not tell you which tool you will like using. A trial is the way to find that out, and each listing links to the vendor.

If a link on this site earns us a commission, it is marked as sponsored and opens as a sponsored link. Commissions and sponsorships never change a score, a weight, a ranking order or whether a tool is listed.

If a price, limit or feature is out of date or wrong, email maruf@trackrev.io with the page and the vendor URL that shows the correct value. We re-read that page, correct the fact, update its check date and record the change in the changelog. Vendors can ask for corrections the same way; a source link is required from everyone, and a correction is made only when the vendor's public page supports it.

Questions about how a score was computed are welcome at the same address. Every score on the site can be reproduced from the facts and sources shown next to it.