Reviewed November 21, 2026. SE Ranking does not become useful when you only paste a domain. You add a project, then the keywords you actually care about, the search engines and locations, and the competitors you want next to those keywords. Readers comparing it with Semrush often say Semrush will start from a URL and suggest the rest, while SE Ranking asks for that list up front. Plan on that extra setup. It is the product working as designed, not a broken screen.
The free way in is a 14-day trial with no card. Starter limits on that trial include 10 projects, 750 keywords a day, and 20 AI prompts a day. That is enough to test a real site. It is not the paid Core cap of 2,000 keywords and 25 prompts. The plan prices are on the pricing page, and whether the suite fits you is on the review.
What to enter on day one
Start with one site, not every client you might add later. Use the domain you can also see in Search Console, so the tool's positions and Google's clicks describe the same property. Add the keywords you would put in a report, not a dumped list of every variation. Ten accurate terms will tell you if the tracker matches Google. Five hundred unreviewed terms will only give you a noisy average.
Pick the search engine, the country, and the device you report on. If the client asks about phones, add a mobile check for a sample of the list rather than assuming desktop and mobile move together. Add two or three competitors whose pages you would actually study. You can widen the set after the first check looks sane.
What the trial will and will not show
Use the trial on the site and the keyword list you would pay for. A demo project with invented keywords will not tell you whether the positions match what you see in a browser. The starter cap of 750 keywords a day is the limit to respect. If you need 2,000, that is a Core question, not something the trial is supposed to unlock.
AI prompts are a separate counter. The trial lists 20 a day. Core and Growth list 25. That line is not the full AI results tracker. If you need mentions inside AI Overviews or ChatGPT, read LLM SEO tools before you assume the prompt cap includes them. Rank checks themselves are explained on the rank tracker page.
After the first check
Compare a handful of keywords with a normal browser search and with Search Console before you send a report. In late September 2026 the tracker marked keywords as missing while they were still ranking, and SE Ranking later said desktop had been fixed while mobile was still being corrected. One mismatch is a reason to look. A project-wide collapse that Search Console does not show is a reason to wait before you tell a client the site fell apart.
When the positions look right, add the next project. The bill changes with projects, keywords, seats, and add-ons. Core includes 10 projects. If you are about to cross that, price Growth before you promise the eleventh site. The trial page has the signup limits: SE Ranking free trial.
Engines and locations worth adding
Add Google for the country you report, and add a second engine only if a client asks about it in meetings. A project that tracks five engines for every keyword burns the daily cap on numbers nobody reads. The same applies to cities. A national check plus the two cities where the business actually has customers is a clearer report than twenty cities copied from a default list.
Mobile is the check people skip and then regret. The October 2026 fix was announced for desktop first, with mobile still being corrected. If the client lives on a phone, include a mobile engine for the keywords you would be embarrassed to get wrong. You do not need mobile on the entire list to learn whether that check is trustworthy.
What to ignore on the first day
Skip the white-label report, the scheduled email, and the extra seats until a human has looked at the first positions. A branded PDF of bad data is still bad data. Seats matter when a second person needs to work in the project at the same time. The product notes that one active session per login is the normal pattern, so a shared password is a poor substitute for a seat.
Also skip adding every suggested keyword. Suggestions are a research list. The tracked list should be the queries you would mention if someone asked how the site is doing. You can move a suggestion into tracking after you decide the page that should rank for it. That decision is the same one covered in keyword cannibalization: one URL per query.
A sane first week
Day one is the project, the keyword list, and one engine. Day two is the hand check against a browser and Search Console. Day three is the competitors, once you trust the positions. Reports, extra seats, and a second site wait until that hand check looks right. Rushing the branded report is how a trial gets spent on formatting.
Write down the cap you are under. On the trial that is 750 keywords a day and 20 AI prompts. On Core it is 2,000 keywords and 25 prompts, across 10 projects. If the list you want does not fit, the decision is to cut the list or to price the next plan. It is not to paste the full list and hope the extra keywords are tracked anyway.
Questions people ask
Can I skip keywords and let it discover them?
You can research keywords in the tool, and you should still choose the list you track. An unedited suggestion list is how a report fills up with queries you do not offer.
Do I need a credit card for the trial?
No. The public trial is 14 days without a card. Prefer a work email if the form asks for one.
How many keywords should the first project have?
Enough to cover the pages you would discuss in a meeting, and few enough that you can check several of them by hand. Expand after that check, inside the trial cap.