SEO Consulting

Writesonic Review November 2026: From an AEO Consultant

A November 2026 Writesonic review from an AEO consultant: articles, site audits, three tracked engines, and which plan includes the Action Center.

Writesonic is the better fit when one published plan has to do two jobs: track a prompt on ChatGPT, Gemini, and Google AI Overviews, and also spend a counted article allowance with a site audit in the same login. It is a weak first buy when someone treats the eight-platform list as the thing every card includes. On 20 November 2026 the brand pricing cards, on the annual view, showed Starter at $79 a month, Basic at $199, and Growth at $399. Enterprise was custom. The instructions page, which says it was last updated 9 November 2026, prints the monthly side of those same three cards as $99, $249, and $499, and it says Starter, Basic, and Growth track only those three engines. Perplexity, Claude, Grok, Google AI Mode, and Muse Spark are named on Enterprise.

Last updated - November 2026

I am Austin Heaton, an AEO consultant. The content systems and visibility logs I review are mostly for fintech. The rest are B2B SaaS and a few other industries when the brief looks the same. A login is useful when a writer can see the article cap, the engine the plan actually names, and the URL an answer cited. It is not useful when a team will ship a generated fee, a license, or a partner, or when a homepage sentence about eight platforms is doing the work of a plan card.

The annual cards, the save lines, the three engine names on those cards, the article and audit counts, and the comparison rows I could read as text are from Writesonic's pricing page, read on 20 November 2026. The page was on the annual view. The control says Monthly, Annual, and Save 20%. I did not flip it, so I did not see a monthly dollar replace the annual one on that screen. The monthly rates, the eight platform names, the yes-and-no matrix, the agency rates, and the note that the product should not be called a separate "Writesonic GEO" product are from Writesonic's instructions page, which says it was last updated 9 November 2026. That page tells you to prefer the live pricing pages for price. Where the two agree, I say so. The agency card I opened the same day is the agency pricing page. It displayed Agency Starter at $200 a month on the annual view. I did not flip that control either.

The older help article that still describes a different Basic and a different Growth is Different subscription plans of Writesonic. Its file header says updatedAt 13 March 2026. The vocabulary page, whose file header says updatedAt 19 June 2025, is AI Search Vocabulary and Metrics. The citations overview, file header 25 February 2026, is Citations Overview. The article-writer help page, file header 14 September 2025, is AI Article Writer Overview. Fair use, file header 2 March 2025, is the Fair Usage Policy. The marketing description of the writer is the AI Article Writer page. The SEO agent page is the AI SEO Agent page. I did not create an account. I did not generate an article. I did not run a prompt. I did not start a trial.

The side-by-side already on this site is not being rewritten as a second invoice. Writesonic beside Peec, from an 18 November read of the same brand cards, is Writesonic GEO vs Peec AI. The annual figures, the three engines, and the 50, 100, and 200 prompt counts I could see on 20 November still match that page. I am citing today's cards, and I am adding the writer, the audit, the monthly rates on the instructions page, and the help article that does not match. Writesonic beside Profound, from a 7 November read, is Writesonic GEO vs Profound. A draft and a research workflow on another card is the Frase review. A letter grade in Google Docs, Word, or WordPress is the Clearscope review. A content score whose plan card still prints a monthly price is the NeuronWriter review.

Key Takeaways

  • On 20 November 2026 the brand pricing cards showed Starter at $79 a month billed annually, Basic at $199, and Growth at $399, with save lines of $240, $600, and $1,200 a year. Enterprise is custom. The hero says start free, no credit card needed. The instructions page, updated 9 November, prints the monthly side as $99, $249, and $499, and says annual billing saves 20 percent. Twelve times the monthly figures, minus twelve times the annual figures, are the three save lines. Seventy-nine divided by 99 is about 0.80, not a clean 20 percent of a round dollar. I did not see $99 on the card itself, because the page I fetched was on the annual view.
  • Starter, Basic, and Growth name ChatGPT, Gemini, and Google AI Overviews. Prompts are 50, 100, and 200. Answers tracked daily are 150, 300, and 600. That is three answers per prompt on every priced card, which matches three engines if every prompt runs on every included engine every day. The instructions page says that is the included set. Enterprise is the card that says all 8 AI platforms. The instructions page names those eight as ChatGPT, Perplexity, Claude, Gemini, Grok, Google AI Mode, Google AI Overviews, and Muse Spark. It does not name Microsoft Copilot. A help article from March still puts Copilot on Enterprise and does not name Muse Spark. A citations help page uses Bing Copilot as an example. I am not merging those lists.
  • Articles per month are 15, 25, and 50 on the three priced cards, then custom. Site audits are 10, 20, and 50. Pages per audit are 100, 1,200, and 2,500. Users are 1, 2, and 3. Projects are 1, 1, and 2. The March help article prints a different Basic and a different Growth: 100 and 200 articles, 40 and 60 audits, and it has no Starter row. It also talks about a Professional plan and an Advanced plan. The instructions page says those names, and prices like $39 or $49 a month, belong to retired AI writing plans. I am not averaging the March article into the card.
  • The AI Article Writer and the SEO and Content AI Agent are marked on every self-serve plan in the November matrix. A row called Content Optimization is marked Enterprise only. Growth can add 20 articles for $100 a month. Basic can add up to 3 users at $50 a user a month. Growth can add up to 2 users at the same rate. Writing styles are 1, 5, and 10, then custom. The Action Center is off on Starter and Basic in that matrix, limited on Growth, and on for Enterprise. The Growth card calls the limit a trial of 5 off-page and 5 on-page actions a month. The Starter card's own sentence says the Action Center is available on Enterprise. Outreach Agent, Fact Check, Citation Decay, GA4, and Looker Studio are Enterprise on the matrix.
  • The product name, on the instructions page, is Writesonic. GEO is a category, not a separate SKU. Chatsonic is described as part of the platform rather than a separate product. Botsonic and Socialsonic are being retired in 2026, on that page. The article-writer marketing page describes a pipeline of 100-plus steps and a 7-day free trial button. The September 2025 help overview still describes Article Writer 4, 5, and 6, and says access varies by plan. I did not open the menu, so I will not tell you which version a Starter seat still shows. No page I read guarantees a citation.

What Is Writesonic in November 2026?

Writesonic, on the instructions page a buyer can hand to an assistant, is an AI search visibility platform for generative engine optimization, which that page also calls answer engine optimization. The sentence I read says it shows how a brand appears in ChatGPT, Gemini, Claude, Perplexity, Google AI Overviews, and other AI platforms, ranks what needs fixing, and makes those fixes through content, external citations, and technical work, then measures citations, AI traffic, and pipeline. The company calls itself the AI Search Growth Engine. The same page says many older articles still describe Writesonic as an AI writing tool, and that this description is out of date. Content creation remains, because it is one of the ways the product closes a visibility gap. I am reviewing both halves. The writing tool is not a rumor. It is a row on the pricing matrix, with a monthly article cap. The visibility tracker is not a rumor either. It is the three engine names on the card, and a longer list that the same company puts on Enterprise.

Want AI engines to recommend your company? Book a free 30-minute AI visibility review with Austin.

Book a Call

The legal line I can read is Writesonic, Inc., founded October 2020 in San Francisco, headquarters 2810 N Church St STE 88739, Wilmington, Delaware. The founder and CEO named on the page is Samanyou Garg. The accelerator line is Y Combinator, Summer 2021. The page says the team is 50-plus people working remotely, that the company is a Microsoft partner hosted on Microsoft Azure, and that compliance covers SOC 2 Type II, GDPR, and HIPAA. SSO and SAML are listed as enterprise security. Ratings on that page are 4.7 on G2 from 2,100-plus reviews, and 4.8 on Capterra from 2,100-plus reviews, plus a G2 Fall 2026 badge line. I did not open G2 or Capterra, and I did not audit the review count. Customers are given as 10,000-plus, with names that include Amazon, Unilever, Acer, and NP Digital. I am citing the page. A logo row is not a result I measured.

The history block is specific enough to brief, and it is also the reason a 2023 review of the writer will mislead a buyer in 2026. The page says 2020 to 2021 was an AI content platform on GPT-3. 2022 was Chatsonic, an AI chat assistant with web search, launched 10 days after ChatGPT. 2023 was a Microsoft Bing AI keynote mention. 2024 was AI agents for SEO and content, with Google Search Console and Keyword Planner. 2025 was the AI search visibility platform, now called the core product. 2026 is the repositioning as the AI Search Growth Engine. Chatsonic, the page says, was upgraded and folded into the platform. Botsonic, the custom chatbot builder, and Socialsonic are being retired in 2026. I did not find a shutdown date more precise than the year. If a seat still shows a Botsonic tile, the instructions page is the claim and the account is the fact. I did not log in.

The loop the page prints is four steps, and the pricing page is where that loop gets cut. Track means prompts run across AI platforms every day, with visibility, citations, and sentiment for the brand and its competitors. Prioritize means the Action Center ranks opportunities. Act means agents create content, refresh pages with schema, FAQs, and fresh data, find third-party pages AI cites, draft outreach, and fix technical issues that block crawlers. Measure means results at Day 14 and Day 28, using citation share, AI traffic, conversions, and pipeline. The page then says availability of the Action Center and the Outreach Agent depends on the plan. That sentence is the one I would put in the footnote of any demo. A four-step story on a company page is not a four-step entitlement on Starter.

The job I can defend is narrower than the hero. Someone has to choose the query the business can actually serve. Someone has to catch a wrong fee in a paragraph the model drafted. Someone has to know whether a cited URL was yours. Writesonic's instructions page can explain an article cap, a three-engine card, an eight-engine enterprise list, and a split between a writer that is included and an Action Center that is not. It does not, on the pages I read, close the fee check by itself. "Get cited" is a goal on the marketing pages. It is not a promise that an answer engine will mention the brand. I did not find that promise.

How Do the Plans Compare?

The table is the brand card as I could read it on 20 November 2026, with the monthly rate from the 9 November instructions page in the same cell because the pricing screen I fetched did not show it. A help article from March is not a column. It is a conflict, and it is in the next section. Enterprise cells say custom where the card does not print a number.

CriterionStarterBasicGrowthEnterprise
What the card says it is forTrack the brand on three engines, with AI articles and site audits.The card's line is SEO plus GEO in one platform, on the same three engines.The card marked most popular. Sentiment, an Action Center trial, and agentic workflows.Custom. All 8 AI platforms, full Action Center, and a strategy team. The card does not print the eight names.
Price I could read on 20 November 2026$79 a month, billed annually. Save $240 a year. Instructions page monthly rate: $99. I did not flip the toggle.$199 a month, billed annually. Save $600 a year. Instructions monthly rate: $249.$399 a month, billed annually. Save $1,200 a year. Instructions monthly rate: $499.Custom on the card and on the instructions page. The annual control's 20 percent label does not produce a dollar I can cite.
Engines on the card and on the November instructions pageChatGPT, Gemini, and Google AI Overviews.The same three.The same three.Instructions name eight: ChatGPT, Perplexity, Claude, Gemini, Grok, Google AI Mode, Google AI Overviews, and Muse Spark. Microsoft Copilot is not in that list.
Prompts and answers tracked daily50 prompts. 150 answers a day.100 prompts. 300 answers a day.200 prompts. 600 answers a day.Custom prompts. Custom answers. Tracking frequency on the matrix is still daily.
Articles, audits, and pages15 articles a month. 10 audits. 100 pages each.25 articles. 20 audits. 1,200 pages each.50 articles. 50 audits. 2,500 pages each. Article add-on: 20 more for $100 a month.Custom articles. Custom audits. Content Optimization, which is off on the three priced columns of the matrix, is on here.
Writer, SEO agent, and workflowsAI Article Writer, SEO and Content AI Agent, and content strategy are marked yes. Agentic workflows: 10 trial runs. 1 writing style.The same writer lines. 50 trial runs. 5 writing styles. Additional users: up to 3 at $50 a user a month.The same writer lines. 100 trial runs. 10 writing styles. Sentiment is on. AI Search Volume is 5 one-time runs. Additional users: up to 2 at $50 a user a month. Prompt diversification is on.Full agentic access. Outreach Agent. AI Search Volume on, not a five-run sample. Brand Explorer, Prompt Explorer, Page Tracker, alerts, shopping. Writing styles are custom.
Action CenterMatrix: no. The Starter card says available on Enterprise.Matrix: no. The Basic card says the same Enterprise line.Matrix: limited. Card: trial of 5 off-page and 5 on-page actions a month. Technical actions are marked yes on Growth in the matrix.Full Action Center. Fact Check and Citation Decay are on this column and not on the three priced cards.
People, projects, and access1 user. 1 project. 1 market, language plus country. Email support. Search Console and WordPress are marked yes. GA4 and Looker are not.2 users. 1 project. 1 market. Email support. The same integration split.3 users. 2 projects. 1 market. Priority email.Custom users, projects, regions, and languages. SSO and SAML. Dedicated strategist. Slack, Teams, email, and an account manager. SOC 2 Type II, HIPAA, and GDPR are named on this card.

If the article cap and the engine list are about to be two purchases, Book a Call and bring one page you would actually refresh. The page shows whether you need the writer, or a login that can name Perplexity before Enterprise.

The number I would circle is not the $79. It is the jump at Growth, and then the fact that Growth still does not buy the other five engines. Starter to Basic adds prompts, answers, articles, audits, a second user, and the right to buy extra seats. It does not, on the matrix, add sentiment, the Action Center, or a second project. Growth is the first price that lists sentiment and a limited Action Center. It is also the first price with two projects, prompt diversification, and an article add-on. It is not the price that lists Claude. A founder who hears "most popular" and then asks for Perplexity has not bought Perplexity. The instructions page puts Perplexity on Enterprise, on top of a custom quote. I would not brief a team that the top self-serve card is the AI search card. The top self-serve card is the capacity card, plus a trial of the action list, on the same three engines as Starter.

The answer math is the cleanest cell on the page, and it is also an assumption. Fifty prompts times three engines is 150. One hundred times three is 300. Two hundred times three is 600. The cards print those pairs. The instructions page does not say, in a formula I found, that every prompt is sent to every included engine every day. The pairs match that reading. They would also match a different sampling rule that happens to land on the same totals. I am showing the arithmetic. I am not promoting it to a contract term. If a prompt can be limited to one engine inside the product, the 150 on Starter is a ceiling, not a count of questions. I did not open the prompt screen.

The annual math is the other cell I would not tidy up. The save lines are exact differences, not a rounded slogan. Ninety-nine times twelve is $1,188. Seventy-nine times twelve is $948. The difference is $240, which is the save line on the Starter card. Two hundred forty-nine times twelve is $2,988. One hundred ninety-nine times twelve is $2,388. The difference is $600. Four hundred ninety-nine times twelve is $5,988. Three hundred ninety-nine times twelve is $4,788. The difference is $1,200. The instructions page says annual billing saves 20 percent. Twenty percent of $99 is $19.80, and $99 minus $19.80 is $79.20, not the $79 on the card. Twenty percent of $249 is $49.80, and $249 minus $49.80 is $199.20, not $199. Twenty percent of $499 is $99.80, and $499 minus $99.80 is $399.20, not $399. The cards are a dollar or two under a strict 20 percent. I am not going to "correct" either page. A buyer who wants the amount that will be charged should read the billing screen. The annual total I can defend is the monthly annual rate times twelve: $948, $2,388, and $4,788. The save line is the gap versus twelve months of the monthly rate on the instructions page. I did not complete a checkout, so I do not have a receipt.

Seats are clearer here than on a product that says unlimited and then hides a fee. Starter includes 1 user and does not, on the matrix, offer an additional-user price. Basic includes 2 and can add up to 3 more at $50 a user a month. Growth includes 3 and can add up to 2 more at the same $50. That is an odd step down. Growth's included headcount is higher, and its add-on cap is lower. A five-person team on Basic would be 2 included plus 3 add-ons. A five-person team on Growth would be 3 included plus 2 add-ons. Both land on five, and both add-on paths are $50 a seat on the matrix. I did not find a sentence that says the $50 is billed monthly rather than yearly. The cell says "$50/user/mo". I am reading "mo" as month because that is the abbreviation they printed. If the annual toggle changes the seat price, the billing screen is the price. Projects do not follow the same step. Basic stays at 1 project. Growth is the first card with 2. A second site is not a reason to leave Starter for Basic. It is a reason to read Growth, or to ask whether a project can be added without a plan change. I did not find an extra-project price.

The free start is two sentences that were not on one button I clicked. The pricing hero says start free, no credit card needed. The article writer page and the SEO agent page both show a 7-day free trial button. I did not submit either form, so I cannot tell you whether the trial is 7 days, whether it is the Gold-equivalent of some other vendor's trial, or whether the form asks for a card despite the hero. The instructions page does not, in the copy I used, print a day count. I am not going to borrow a 7 from the marketing button and stick it on the pricing FAQ. The pricing FAQ lists the question "Do I need a credit card to start?" The answer did not render in the HTML I pulled. The question list is not the answer.

The Help Article That Still Describes Another Product

This is the section I would make a procurement reader open before anyone pastes a Writesonic price into a security review from the help center. The article titled Different subscription plans of Writesonic has a file header of 13 March 2026. It still describes the company as creating content, optimizing for SEO, and tracking AI search. That part can be true and the table can still be stale. The table has three rows: Basic, Growth, and Enterprise. There is no Starter row. Basic is $249 a month and $199 annually, with 2 users, 2 projects described as up to 4, 100 articles, 40 audits of 1,200 pages, and 100 prompts. Growth is $499 a month and $399 annually, with 5 users, 3 projects described as up to 7, 200 articles, 60 audits of 2,500 pages, and 200 prompts. Action Center is a dash on Basic and on Growth, and Yes on Enterprise.

Put that next to the card I opened on 20 November. Basic's annual price, $199, matches. Basic's prompt count, 100, matches. Basic's pages per audit, 1,200, matches. Basic's articles do not: the card says 25, the March article says 100. Basic's audits do not: the card says 20, the March article says 40. Basic's projects do not: the card and the November matrix say 1, the March article says 2, up to 4. Growth's annual price, $399, matches. Growth's prompt count, 200, matches. Growth's pages per audit, 2,500, matches. Growth's articles do not: 50 on the card, 200 in the March article. Growth's audits do not: 50 versus 60. Growth's users do not: 3 versus 5. Growth's Action Center does not: a limited trial on the card, a dash in the March article. I am not going to pick the more generous number because a buyer would like it. The live pricing page and the 9 November instructions page agree with each other. The March article agrees with neither on the meters that changed.

The names are the other trap. The March article's Growth section says the prompt count is higher than the Professional plan, and that the upgrade versus Professional is sentiment, prompt search volume, and dedicated Slack. The FAQ says Professional tracks up to 100 prompts a month and Advanced tracks 200, with unlimited prompts on Enterprise. The November instructions page says older plan names such as Lite, Standard, Professional, and Advanced, and prices like $39 or $49 a month, belong to retired AI writing plans and do not reflect current pricing. The March article is still using Professional and Advanced as if they were the current GEO ladder. Enterprise's engine list in that article is also a different eight, or a different seven, depending on which paragraph you read. The table names ChatGPT, Perplexity, Google AI Overview, Google AI Mode, Claude, Gemini, and Microsoft Copilot. The prose adds Grok and says full coverage across all current and future platforms. Muse Spark is not in that list. The 9 November instructions page's eight are ChatGPT, Perplexity, Claude, Gemini, Grok, Google AI Mode, Google AI Overviews, and Muse Spark, and Copilot is absent. Two official pages, two Enterprise menus. I am using the November instructions page and the 20 November card for what a buyer can buy this month. I am leaving the March article in the review so nobody treats it as a secret higher cap.

One more March sentence should not migrate into a brief. The FAQ says Enterprise has unlimited prompts. The November matrix says Enterprise prompts are custom. Unlimited and custom are not the same word. A custom order can be large. It can also be a number someone negotiated. I would ask which word the order form uses before anyone writes unlimited into a slide.

What the Writer, the SEO Agent, and the Audit Actually Include

The writer is on the card, and the writer is not one object across the site. The November matrix marks AI Article Writer as yes on Starter, Basic, Growth, and Enterprise, next to an article cap of 15, 25, 50, or custom. Content strategy and the SEO and Content AI Agent are the same yes. Upload your own data is yes. Content Optimization, as its own row, is no, no, no, and yes. Automatic AI fixes on the site audit are yes on every column. I am not going to tell a Starter buyer that "content optimization" is included because the article writer is included, and I am not going to tell them the audit's automatic fixes are the Enterprise content-optimization row. They are documented as different lines. I did not click either one.

The marketing page for the article writer is a pipeline story, and I am treating it as a pipeline story. It says every article runs through 100-plus steps and 10 phases, with 11 expert frameworks, and that no phase is skipped. It describes two modes. One is grounded in uploaded knowledge: a brand brief, product docs, founder notes, and customer research, with light web research after. The other is deep web research from SERPs, Ahrefs, Keyword Planner, and a dataset the page describes as 2 billion-plus real conversations. It says the two can be combined. It says a humanizer pass strips AI tells, and the comparison on that page names em dashes as one of those tells. It says citations come from a verified bank and first-party sources, and that the generic alternative hallucinates them. It says you approve work before it ships, on the SEO agent page's version of the same idea. I did not run the pipeline, so I cannot tell you whether a person accepts each phase or whether an article lands in one pass. I would assume a generated sentence can be wrong until a trial shows the review step. "Verified" on a marketing page is not a check against your fee schedule.

The help overview has not caught up with that marketing page, and I am not deleting it. The article dated in the file header as 14 September 2025 says the product is a suite with three versions: AI Article Writer 4, AI Article Writer 5, and AI Article Writer 6. Writer 6, in that article, includes tone, internal linking, source citation through links or footnotes, an AI cover image, an FAQ block, a call to action, and keyword research. The note says access varies by the plan you chose. The November matrix does not mention three versions. It mentions an article cap and a writing-style cap. A Starter seat with 15 articles and 1 writing style may be Writer 6, or it may be a newer pipeline that the September article does not describe, or the menu may still show 4, 5, and 6. I did not open the left-hand menu the help article tells you to click. If a demo spends the hour on "100-plus steps" and the order form is 15 articles on Starter, you have bought the cap. You have not, from these pages, bought a guarantee that all 100 steps run on that cap.

The SEO agent page is the on-site half of the same login. It says the audit starts with live SEO data and ends with implementation steps. The pages view is where it says AI search, SEO, and content data sit together. The checks it names are keywords you rank for, organic competitors, page traffic, striking-distance and close-to-page-one filters, and implementation steps such as what to expand or where to aim a featured snippet. It says you review the agent's preview before a fix runs or content ships. On the AI side it says you can see which pages ChatGPT and Google AI Overviews cite, clear robots.txt rules that block AI crawlers, and re-track the page. Those two engine names match the self-serve card. They do not add Perplexity to Starter. I did not connect Search Console, so I have the page's description and not a keyword export.

Site audits are a meter, not a vague "SEO included" line. Starter is 10 audits a month, 100 pages each. Basic is 20 audits, 1,200 pages each. Growth is 50 audits, 2,500 pages each. A 100-page cap on Starter is a sample of a real marketing site, not a crawl of every URL. Basic's 1,200 is the first number I would call a site. The help center's site-audit section is a long list of issue definitions, from Core Web Vitals to hreflang to structured data. I did not open each definition. The existence of the list is the point: the audit is a technical checklist, and the article writer is a draft. Neither one, on the pages I read, knows whether the fee in the draft matches the fee on the pricing page. The fair-use policy is a third constraint, and it is older than the GEO packaging. The file header says 2 March 2025. It says more than 95 percent of users stay inside the limits, that going over can limit access with or without notice, that each seat is one person, and that sharing a login can mean suspension or deletion without a refund. It does not print the article cap. The cap is on the pricing card. The policy is the enforcement story. I am not turning "95 percent" into a finding. It is their sentence.

Credits, in the sense of a balance that falls when you generate a paragraph, are not how these cards are sold. I did not find an AI-credit balance on the brand pricing page or in the November matrix. The meters are articles, prompts, answers, audits, workflow runs, and users. A team that is comparing Writesonic to a tool that sells word credits will want them to be the same tank. They are not documented as the same tank. The article add-on on Growth is the closest thing to a top-up I could read: 20 more articles for $100 a month. Starter and Basic, on the matrix, do not have that add-on. If the product lets you buy articles on Starter inside the app, the matrix I read does not say so. AI Search Volume is another small tank: off on Starter, 5 one-time runs on Basic, 5 one-time runs on Growth, and on for Enterprise. "One-time" is their word. I am not turning it into a monthly refill. Five runs is a sample of a prompt set, not a research program.

Where AI Search Shows Up, and Where It Does Not

This is the section that decides whether Writesonic belongs in an AI visibility stack or only in a content stack. The feature list on the instructions page is long: an AI Visibility Tracker for mentions, rank, citations, sentiment, and share of voice; an Action Center across seven categories, including external mentions, new content, page refreshes, and technical fixes; Fact Check for inaccurate claims about pricing, features, and positioning, plus the pages cited for those claims; brand sentiment; citation gap analysis; an Outreach Agent that finds third-party pages, sends pitch emails, and tracks replies; AI Agent Analytics for crawler visits and human referrals, including pages that return 403, 404, and 500 to bots; AI Search Volume from a dataset of 2 billion-plus conversations; Brand Explorer; shopping analytics for how often ChatGPT and Google AI Overviews recommend products and which merchant gets the buy link; an ads tracker for ads inside ChatGPT, Google AI Overviews, and Google AI Mode; and an MCP server and API. Several of those names are Enterprise on the matrix. A feature tile is not an entitlement. Ads monitoring is the exception I could see on every priced card, including Starter. Sentiment is Growth. Fact Check, citation decay, social citations, shopping, alerts, Brand Explorer, and the Outreach Agent are Enterprise.

The definitions do not agree with each other, and the dates are the reason. The 9 November instructions page says AI visibility is the share of tracked prompts where AI answers mention the brand. The example is a brand mentioned in 71 of 100 tracked prompts, which is 71 percent. Citations and citation share, on that page, are how often AI answers cite the brand's pages as sources, compared with competitors. Sentiment is whether answers describe the brand positively, neutrally, or negatively. Those are three different facts. The June 2025 vocabulary page defines visibility as answers where you are mentioned, divided by total answers tracked, times 100, with an example of 57 out of 100. It then says a citation may include a link, or just mention the brand. That second sentence folds a mention into a citation. The November page does not. The vocabulary page also prints "what good looks like" bands: visibility above 70 percent is excellent, 40 to 70 is good, below 30 needs work. Sentiment above 85 is excellent, 70 to 85 is good, below 60 is concerning. I am citing the bands. I am not adopting them as a standard for a client's category. The same page's platform example, ChatGPT at 60 percent, Perplexity at 45, Claude at 52, and Google AI at 58, is an illustration. It is not a study with a prompt set I could recompute, and it names engines a Starter card does not include.

The February 2026 citations overview is a third object, and it is closer to a URL report than to the vocabulary page's "a mention can be a citation" line. It defines my citation share as the percentage of total AI citations in the tracked space that point at your domain. Your cited pages are unique pages from your site that AI platforms cite. External pages mentioning you are third-party pages that mention the brand and are cited in answers. The source breakdown is owned, competitor, third-party websites, and social and forums. The opening examples of platforms are ChatGPT, Perplexity, Claude, Google AI Overviews, and Bing Copilot. That list is not the Starter list, and it is not the November eight. Bing Copilot is not Muse Spark. I am not going to add Copilot to Enterprise because a help page used it in a sentence about "platforms like" those names. If the account's platform filter shows Copilot, the account wins. This review has the pages, not the filter.

Share of voice is easy to merge with visibility because both are percentages. The vocabulary page calculates market share as your visibility divided by the sum of all tracked brand visibility, and it shows a Canva example at 34.2 percent. The instructions page says share of voice is reported alongside visibility, citations, and sentiment, and it does not reprint that formula. I will not paste the Canva figure into a deck. It is a screenshot's number in a help article. What I will take is the warning: a visibility percentage and a share-of-voice percentage are not the same sentence. If a Monday report says "the Writesonic score," it has to say which number: prompt mention share, answer mention share, citation share of your domain, or sentiment. The pages do not promise those four are one gauge.

Opportunities are the action list, and they are also where the off-site page shows up. The instructions page says the Act step finds third-party pages AI cites and drafts outreach. The Outreach Agent page is linked from that section, and the matrix marks the agent as Enterprise. Growth's Action Center trial is 5 off-page and 5 on-page actions a month, plus technical actions. Five off-page actions is a sample of the sources a real category will show you. It is not a program that places the piece. The citations overview's "external pages mentioning you" is a count of pages you do not own. Guidance is a URL, or a drafted email on the Enterprise description. It is not a byline on someone else's site.

When the missing source is a page on someone else's site, Offsite Posts writes and publishes industry articles about a brand on third-party sites, and the site says the posts, mentions, and links in a package are guaranteed while rankings, traffic, and AI mentions are not.

Fact Check is the control I would want on a fee page, and it is not on the priced cards. The instructions page says it finds inaccurate claims about pricing, features, and positioning, and the pages AI cited for those claims. The matrix says no, no, no, yes. Growth has sentiment, which is tone, not a fact check. A draft can be on-brief and still state a price the product does not charge. The article cap does not know your fee. On a live account, Austin Heaton still reads the fee, the license, and the partner name before a generated paragraph ships. The Action Center will not do that job if the source page is wrong, and on Starter the Action Center is not in the price.

Ads, shopping, and bot traffic should not be merged into the prompt report. Ads monitoring is marked on every brand plan, and the ads help page says it covers sponsored ads inside ChatGPT and Google AI Overviews. The instructions page's ads line also names Google AI Mode. Shopping and agentic commerce are Enterprise on the brand matrix, and the shopping description is ChatGPT and Google AI Overviews product recommendations. AI Agent Analytics is a crawler and referral log, with setup guides for Cloudflare, WordPress, Vercel, Drupal, Fastly, CloudFront, and a custom drain. A bot that fetched a URL is not an answer that cited it. A referral is not a mention. If the Monday slide has one number on it, it is the wrong slide. Markets are the other quiet cap: 1 language-plus-country on Starter, Basic, and Growth, and custom on Enterprise. A second country is not a toggle on the $79 card. The help center has a page titled for tracking multiple regions. The matrix says the self-serve plans track one market. I am using the matrix for the entitlement.

What an Agency Seat Changes

Agency pricing is a different invoice, and I opened it on the same day so a brand card would not be briefed as the agency card. On 20 November 2026 the agency page, on the annual view, showed Agency Starter at $200 a month. The control says Save 20 percent. The instructions page prints Agency Starter at $200 a month billed annually and $250 billed monthly. Two hundred is 80 percent of 250, so the 20 percent label is exact on this card in a way it is not exact on the brand cards. I did not flip the agency toggle, and I did not see $250 on the screen. Enterprise is custom. The card's bullets for Agency Starter are 20 pitch projects, each running 1 to 7 days, 2,000 answers a month, a dashboard and shareable audit reports, unlimited users, and a line that says projects and access control, email support, and all models except Claude and Muse Spark. The six names on the card are ChatGPT, Perplexity, Gemini, Google AI Mode, Google AI Overviews, and Grok. That is a wider menu than brand Starter, at a higher annual rate, and it still excludes Claude and Muse Spark. Convert a pitch project to a standard project from $150 a project a month. The comparison row says that conversion is included on Agency Enterprise rather than priced at $150.

The agency card and the instructions matrix do not use the same words for access. The card bullet I read says unlimited users, projects, and access control. The instructions matrix says user access control is no on Agency Starter and yes on Agency Enterprise, that standard ongoing projects are no on Agency Starter, and that white-label reporting, the article writer, the site audit, agentic workflows, Looker Studio, the Action Center, and Brand Explorer are no on Agency Starter. Pitch projects are the product that card is selling: up to 20 a month, 1 to 7 days, then the project auto-pauses. Two thousand answers a month, with the formula printed as prompts times models times runs, is not the same object as brand Starter's 50 prompts and 150 daily answers. I am not converting 2,000 into a prompt cap. The page does not print the prompt count. A six-engine pitch that runs every day will spend that 2,000 faster than a one-engine pitch. I did not book the agency demo, which the page schedules as 30 minutes with a named partnerships contact. I am not reviewing a call I did not take.

The practical split is the one I would say out loud. A brand team that wants articles and a standing prompt log is on the brand cards, with three engines until Enterprise. An agency that wants a week-long audit to show a prospect, on six engines, without Claude, is the $200 annual card, and it should read the "no" rows for the writer and the Action Center before it assumes the brand matrix came along. Agency Enterprise is where the instructions page puts all eight platforms, ongoing projects, white-label reports, and the writer. I would not tell a brand buyer to "just get the agency plan" for Claude. The agency Enterprise price is custom, and the brand Enterprise price is custom. They are two conversations.

Who Writesonic Fits, and Who Should Skip It

Pick from the engine the plan actually names, the article cap you can explain, and whether a person will read the draft.

  • You want a printed annual price, three engines, and a draft in the same login. Starter, if 50 prompts, 150 daily answers, 15 articles, 10 audits of 100 pages, and 1 user can hold the month, and if ChatGPT, Gemini, and Google AI Overviews are the only engines you need to see. Basic raises the caps and adds a second user. The engines do not change.
  • You need sentiment, a second project, or a limited action list. Growth is the first card that lists them. It does not, on the card I read, add an engine. The article add-on and the extra seats are on this card too. Read the five-run limit on AI Search Volume before you assume volume research is included in the same way sentiment is.
  • You need Perplexity, Claude, Grok, Google AI Mode, or Muse Spark inside this login. That is Enterprise, on the 9 November instructions page, not a line on the $79 card. The March help article names Copilot instead of Muse Spark. I am not picking the older list. The Peec comparison is the one to read if the alternative is a choose-your-models card. I am not pasting that invoice into this table.
  • You wanted the writer without the tracker, on a retired plan name. The instructions page says Lite, Standard, Professional, and Advanced are retired. The March article still uses Professional and Advanced. I would not buy from that article. A grade in another editor is the Clearscope review. A different draft workflow is the Frase review.
  • You need the draft checked against a fee schedule. Fact Check is Enterprise. The article writer, on the pages I read, does not know your fee. The score, if a demo calls the visibility percentage a score, measures mentions in the prompts you tracked. It does not know the fee schedule.
  • You wanted the tool to publish a citation guarantee, or to place the third-party article. No page I read promises that an answer engine will mention the brand. The Outreach Agent, on the matrix, drafts and tracks email, and it is Enterprise. Skip it when the Monday metric is "we will be cited," and the only thing the login can show is a prompt report and a list of URLs.

A tool subscription and an engagement are different bills. This page is only the tool. The neighboring visibility logs, if the objection is the tracker rather than the article cap, stay on the comparison already linked above.

What I Could Not Confirm

  • A stored answer, a credit-style balance, or an article landing in the editor. I read the public pages. I did not create an account, and I did not start the free trial or the 7-day trial button.
  • Whether the billing screen shows the monthly rate from the instructions page or only the annual rate I saw. I did not flip the Monthly control on the brand page or the agency page.
  • Whether every prompt runs on all three included engines every day. The answer caps equal prompts times three. The pages I read do not print that sentence as a sampling rule.
  • Which Enterprise engine list the order form uses. The 9 November instructions page names Muse Spark and does not name Microsoft Copilot. The March help article names Copilot and does not name Muse Spark. The citations overview uses Bing Copilot as an example.
  • Whether Article Writer 4, 5, and 6 are still in the product, or whether the 100-step pipeline replaced them. The September 2025 help article describes three versions. The marketing page describes one pipeline. The pricing matrix describes an article cap.
  • The answers to the pricing FAQ. The questions rendered. The answers did not, in the HTML I pulled. I am not filling them in from the hero.
  • What a Growth Action Center trial produces, and whether five off-page actions means five URLs, five tasks, or five months of a single task. The card says 5 off-page and 5 on-page actions a month. It does not define the unit beyond the word actions.
  • The price of an extra article pack on Starter or Basic, the price of a second project, and whether the $50 seat rate changes on annual billing. The matrix does not say.
  • The customer quotes, the 10,000-plus line, the G2 and Capterra ratings, the 2 billion-conversation dataset, the Day 14 and Day 28 measurement claim, and the "100-plus steps" line. They are on the pages. I am not using them as outcomes.

Frequently Asked Questions

Who is Writesonic the better fit for?

Writesonic is the better fit when one login has to track a prompt on the three engines named on the self-serve cards and also spend a counted article allowance, with a site audit in the same plan. It is a weaker fit when the team needs Perplexity, Claude, Grok, or Google AI Mode in the price they can explain, because the instructions page puts those names on Enterprise, not on Starter, Basic, or Growth. The article caps and the engine split are in this article. The plan prices are in the pricing table. I did not create an account.

Does every plan track ChatGPT, Perplexity, and Claude?

Not on the pages I read. The pricing cards and the instructions page say Starter, Basic, and Growth track ChatGPT, Gemini, and Google AI Overviews. The instructions page says Enterprise tracks eight platforms, and it names Perplexity, Claude, Grok, Google AI Mode, and Muse Spark in that eight, along with the three on the cheaper cards. A help article updated in March still names Microsoft Copilot on Enterprise and does not name Muse Spark the same way. A citations help page uses Bing Copilot as an example. I am not treating an example as an entitlement. The split is in this article, and the plan prices are in the pricing table.

How many articles does a plan include, and is that the same as a prompt?

No. The pricing cards print an article allowance next to a prompt allowance and a daily answer cap. Articles, prompts, and answers are three meters. The instructions page also prints site audits, pages per audit, users, and projects as separate counts. A help article updated in March prints different article and audit counts for Basic and Growth, and it does not list a Starter row. I am using the pricing cards and the November instructions page for the caps in the pricing table. I did not generate an article.

What does the Action Center change?

The instructions page says the Action Center ranks opportunities, and that availability depends on the plan. The same page's matrix marks it off on Starter and Basic, limited on Growth, and on for Enterprise. The Growth card I read calls that limit a trial of off-page and on-page actions. The Starter and Basic cards say the Action Center is available on Enterprise. The Outreach Agent, which drafts pitch emails to cited pages, is marked Enterprise on that matrix. I did not open the Action Center. The split is in the pricing table in this article.

Is Writesonic still an AI writer?

The instructions page says describing Writesonic as only an AI writing tool is out of date, and that content creation remains because it is one way the product closes a visibility gap. The pricing matrix marks the AI Article Writer and the SEO and Content AI Agent on the self-serve plans, and it marks a row called Content Optimization as Enterprise. A help overview updated in September 2025 still describes three article-writer versions and says access varies by plan. The marketing page describes a long research pipeline and a humanizer. I did not run a draft, so I am citing the pages, not a file I accepted.

How does Writesonic define a mention and a citation?

The November instructions page says AI visibility is the share of tracked prompts where the answer mentions the brand, and that citations are how often answers cite the brand's pages, compared with competitors. A vocabulary help page updated in June 2025 defines visibility as mentions divided by answers, and it says a citation may include a link or just mention the brand. A citations overview updated in February 2026 describes citation share as the percentage of citations that point at your domain, and it lists external pages that mention you as a separate count. Those three pages do not use the same object. I did not open a stored answer.

Where do I read the current prices?

The figures I could verify are in the pricing table in this article, with the page each figure came from. The pricing cards I opened were on the annual view. The instructions page, updated 9 November 2026, also prints a monthly rate and says annual billing saves a percentage. A help article updated in March prints different article counts and still uses plan names the current cards do not show. If a contract is already signed, the contract is the price.

Who should skip Writesonic?

Skip it when the job is a citation guarantee, because no page I read promises that an answer engine will mention the brand. Skip the self-serve cards when Perplexity or Claude has to be in the price you can explain, and read Enterprise before you assume the eight-platform list is included. Skip unattended drafts on any page that states a fee. The article writer does not know the fee schedule. Skip it when the team will not read the stored answer, and this login is the one that gets ignored.

Read next:

If the article cap, the three-engine limit, and the person who reads the fee are three different conversations, Book a Call and we will put the page, the plan, and the add-on on one list before anyone pays for a dashboard the card does not include.

BECOME THE COMPANY AI RECOMMENDS.

Your next qualified lead starts with an answer.

Get more citations. Compete for the prompts that matter. Become the top choice in your category.

Book a Call