Scrunch AI vs Profound for AI Visibility Tracking in 2026

Scrunch is the better fit when Perplexity has to be in a price you can read, the same invoice has to include a thin site audit and a look at AI bot traffic, and one country, one language, and five competitors are enough. Profound is the better fit when the buying process is a security review, Gemini and Google AI Overviews have to be in the first week, and someone can live with a quote after a trial that will not let them edit the prompts. On 14 November 2026 neither entry offer was a nine-engine monitor. Scrunch Core was $250 a month. Profound's brand page was still a free trial and a custom Enterprise plan. I did not find a paid self-serve brand price on Profound's pricing URL.

Last updated - November 2026

I am Austin Heaton, an AEO consultant. The visibility logs I set up are mostly for fintech. The rest are B2B SaaS and a few other industries when the buyer question is the same. A dashboard is useful when I can open one stored answer, see whether the brand was named, see which URL was attached, and tell a writer which page to change. A trial that will not let me edit the prompt is a storage test, not a measurement program. A paid plan that omits the engine the board asked about is a storage test of a different kind.

Scrunch's Core and Enterprise cells are from Scrunch's pricing page, read on 14 November 2026. The agency prices are from Scrunch's pricing FAQ. The trial length and the credit-card rule are from the trial FAQ. Position, sentiment, and citations are from the FAQ on what Scrunch tracks. The observe, optimize, and deliver description, including the Agent Experience Platform, is from Scrunch's products FAQ. A request to scrunchai.com on the same day returned a 301 to scrunch.com, so I am treating scrunch.com as the site. Profound's Trial and Enterprise cells are from Profound's pricing page, including the comparison table and the FAQ. The model list that does not match the "up to 10" line, and the front-end capture claim, are from Answer Engine Insights. Citation categories are from the citations page. Agents, including the approval step, are from Profound's Agents page. I did not create an account at either vendor. I did not start a trial. I did not open a stored answer.

The single-product reviews already on this site are not being rewritten here. Scrunch's own caps, from a 10 November read that still matches the dollars I saw on 14 November, are the Scrunch AI review. Profound's trial, including the pages that used to say nine engines where the pricing card now says ten, is the Profound review. I am not pasting an older engine count forward as if the pricing page had not been re-read. Peec, which does publish a monthly brand price and a pick-three model menu, is the Peec AI review. Peec beside Scrunch is Peec AI vs Scrunch AI. If the Profound quote is the problem, the shortlist is Profound alternatives. Otterly beside Peec is Otterly.ai vs Peec AI. The four-tool version is Profound vs Peec AI vs Otterly vs Semrush. Where a cell below is a vendor's marketing sentence rather than a cap I could count, I say so.

Key Takeaways

• On 14 November 2026 Scrunch's brand pricing page showed Core at $250 a month: 125 unique prompts, 5,000 responses a month, 5 site audits, 1 brand workspace, and 5 user licenses. The four engines on that card are ChatGPT, Perplexity, Google AI Overviews, and Microsoft Copilot. The comparison table limits Core to 1 country, 1 persona, 1 language, and 5 competitors. The sitemap is 25 pages. Page optimizations are 1 a month. Content generation is marked Basic. Product observability and the Agent Experience Platform are not on Core. Enterprise is custom and names nine engines, adding Claude, Gemini, Meta AI, Google AI Mode, and Grok.
• The pricing FAQ adds a second pair of plans: Agency Core at $500 a month, and Agency Enterprise as custom. That FAQ repeats the brand split, Core at $250 and Enterprise custom, and it does not print an agency prompt count, response cap, or workspace count. I am not going to invent the $500 plan's contents from the brand card.
• Profound's brand pricing page the same day showed a free Trial and a custom Enterprise plan. The trial is 50 unique prompts, run daily for 7 days, on ChatGPT, Gemini, and Google AI Overviews. One language. One region. ChatGPT Shopping is off. History, exports, and API are none. Support is none. SSO is no. Seats are unlimited. The FAQ says trial users receive a recommended prompt set and cannot customize it. I did not find a dollar next to Enterprise, and I did not find a paid self-serve brand card between the trial and the quote.
• Profound's Enterprise card says up to 10 answer engines. The capability list names ChatGPT, Perplexity, Google AI Mode, Gemini, DeepSeek, Anthropic Claude, Google AI Overviews, Exa Search, Muse Spark, and Qwen. I counted ten. The Answer Engine Insights FAQ, opened the same day, says Enterprise runs daily across up to 9 answer engines, and it also names a ten-model set that includes Qwen. I am not dropping a name to make the pages agree. The pricing page is the invoice. A quote should repeat the names.
• Scrunch's trial FAQ says the trial is 7 days of Core, a credit card is required, and the subscription starts unless you cancel. It says a trial user can audit five webpages. Profound's pricing page does not say the trial requires a card, and I did not click Get Started. Scrunch Core includes Perplexity and does not include Gemini or Google AI Mode. Profound's trial includes Gemini and Google AI Overviews and does not include Perplexity. Neither entry offer is both. Neither plan guarantees a citation.

What Job Are These Two Tools Doing?

Both products sell a view of how AI answers talk about a brand. They do not sell the same week of work. Scrunch's products FAQ splits the work into three verbs. Observe: what AI answers say, and what bots do on the site. Optimize: recommendations at the domain and at the page. Deliver: approved content served to agents at the edge, with the human-facing site left unchanged. Profound's pricing page splits the work into a prompt log, a shopping row, an on-site analytics row, and a writing layer it calls AI Marketer and Agents. The pricing pages are narrower than those stories, and the pricing pages are the invoices.

Scrunch's own post from June 2026, Scrunch is now a Sitecore company, says Scrunch continues as a standalone platform. I am citing that sentence because a buyer will hear the acquisition in a demo. I am not treating a parent company as a feature on the Core card. The Core card is still four engines, 125 prompts, and a 25-page sitemap. A later integration does not move a checkmark that the pricing table left off.

Profound's Answer Engine Insights page says the capture is the consumer front end, not an API response: what you see in Profound is what a customer sees. I did not watch a browser session. I am citing Profound's method claim, not a capture I recorded. Scrunch's public pages that I opened do not, in a sentence I could quote, make the same UI-versus-API claim. I will not invent one so the two methods match. If a security review asks which surface was queried, Profound has a sentence to put in the questionnaire, and it still has to be repeated on the order form. Scrunch has a pricing table. Ask Scrunch the same question in writing.

The job I can defend in a Monday meeting is the same job I defend for any other log. Someone has to own the prompts the business can actually serve, someone has to catch a wrong fee, and someone has to know whether the cited URL was yours. A tracker can do the third job only when it stores the answer and the link. A posting tool can do the first draft of a page. It does not, on these pricing pages, close the measurement. The measurement notes live in measuring AI visibility. If the company already pays for a suite, the specialist log still has to beat that suite on the stored answer. That comparison is Ahrefs vs SEMrush for AI search reporting.

How Do Scrunch and Profound Compare?

The table uses each vendor's labels as I could read them on 14 November 2026. "Google AIO" is Scrunch's label on the pricing page. Scrunch's pricing FAQ writes the same engine as Google AI Overviews. I am not expanding Google AIO into AI Overviews plus AI Mode, because Scrunch lists Google AI Mode only on Enterprise. Profound lists Google AI Overviews on the trial and Google AI Mode on the Enterprise capability list, as separate names, which is the right way to print them.

If the engines in the quarterly deck are not the engines on the entry card, Book a Call and bring one prompt. The prompt shows which vendor is even eligible, before anyone argues about agents or audits.

What Scrunch Puts on a Published Price

Core is the only brand plan with a dollar, and the pricing page describes it as the plan for a team that wants to evaluate Scrunch, set a benchmark, and prove impact before it scales. That sentence matters. The card is an evaluation plan that happens to renew, not a quiet way to buy the nine-engine list. Enterprise is where the model list, the countries, the API, the fuller SSO, the dedicated account team, and AXP sit. If a demo spends the hour on AXP and the order form is Core, you have not bought the demo.

Core has two meters, and they do not obviously describe the same month. The card says 125 unique prompts and 5,000 responses a month. Scrunch does not, on the pages I read, publish the formula that turns a prompt into a response. I am not going to borrow another vendor's multiplication and pin it on Scrunch. The arithmetic I can do without guessing the sampling rule is still worth putting on a slide, because it shows which stories are impossible. One hundred twenty-five prompts, once a day, for 30 days, is 3,750 responses if each prompt returns one response. That fits under 5,000. One hundred twenty-five prompts, on all four Core engines, every day for 30 days, is 15,000 responses. That does not fit. Five thousand divided by 125 prompts is 40 responses per prompt per month, if every prompt is used. None of those sentences is Scrunch's sampling policy. They are the boundaries of the two numbers on the card. A sales sentence that says "125 prompts across four engines, tracked daily" has not yet reconciled the response cap. I would ask, in writing, whether a response is one engine's answer to one prompt, whether all four engines run, and what happens at the cap.

The pricing page has a question, "What happens if I exceed my plan limits?" The extract I got did not include the answer. The pricing FAQ page is titled as a pricing FAQ and then answers with the plan list. It does not say what happens when you exceed the cap. I am not filling it in. The trial FAQ says you can add up to 125 unique prompts. It does not say you can swap them, and it does not say a prompt is locked for the month. I would test the editor with one real buyer question before I reported on a set someone else chose.

Competitors are capped at five on Core. That is a hard gate for a category with a long comparison set. Category queries in payments and payroll routinely name more than five companies. A cap of five means the share-of-voice math is the share against the five you remembered to add, not against the set the model actually listed. Personas and languages are also one. A US English persona is the whole Core configuration. A second country is an Enterprise conversation, which is the situation a published price was supposed to avoid.

The tracking FAQ is more specific than the pricing grid about what the rows mean. It names brand mentions, competitive presence, share of voice, answer position and rank, sentiment, citations, AI bot traffic, AI referrals, and AI search trends. Position, in that FAQ, is whether the brand is in the top, middle, or bottom of the response, and in what order. That is not a numbered rank you can drop next to another vendor's average position without a footnote. Sentiment is positive, neutral, or negative. I did not find a numeric sentiment scale on the pages I opened. Citations, in the same list, are whether the answer links to the brand, a competitor, or a third party. I did not find a Scrunch page, among the ones I opened, that separated a URL the model retrieved from a URL it linked. The pricing row is "Citations / Sources" as one line. I would ask to see a single answer with the link before I treated those words as two metrics.

On-site measurement is the part Scrunch puts on the Core card that a pure prompt log often sells later. AI bot traffic is marked on Core. The tracking FAQ describes referrals as people clicking through from an AI answer. Site audits are 5 a month, and the sitemap on Core is 25 pages. The trial FAQ says a trial user can audit five webpages. Those sentences are not the same object as a full-site audit, and Enterprise is where the table says full site audits. Page optimizations are 1 a month. Content generation is Basic on Core and Advanced on Enterprise. I would not put Scrunch in a statement of work as the system that drafts the blog. The products FAQ describes a content editing surface and an approval step inside the deliver workflow. It also uses an example in which a brand appears in 45% of responses and competitors appear in 78%, and then closes a gap within a few weeks. I am not using those percentages. They are an illustration in an FAQ, not a customer result I measured.

AXP, the Agent Experience Platform, is the delivery claim, and the pricing table gates it. The products FAQ says Scrunch can detect agent traffic and serve optimized content to agents while the human-facing site stays unchanged. The Core column does not include AXP. The Enterprise cell says custom, and it says to get a demo. Product observability is the same shape: off on Core, on for Enterprise. A commerce team should not buy Core to learn whether shopping answers tile their products. The pricing table is the gate.

Google single sign-on is marked on Core. SAML or OIDC is Enterprise. Support on Core is email, and onboarding is self-serve. Enterprise adds Slack, onboarding help, strategy calls, a dedicated account team, and exec or board reporting. Looker Studio, the query API, and advanced API are Enterprise. MCP is custom. The CLI cell on Enterprise, in the extract, read as "Email + Slack," which is also the support line. I could not cleanly separate a CLI product from that cell, so I am not describing a command-line tool I did not see documented there.

The trial is the part I would calendar before I would celebrate. Seven days, Core's features, a credit card, and a conversion to paid unless you cancel. The trial FAQ says the trial includes 125 prompts and that you can audit five webpages. I would set a reminder on day six. That is a different trial from a no-card week. Agency Core at $500 is a real price, and it is also an incomplete card. The FAQ does not say how many workspaces, prompts, or countries that $500 buys. I would make Scrunch put those caps in the order email before I compared Agency Core with a brand plan that is half the price and explicitly one workspace.

What Profound Puts on the Trial

The trial is a week of someone else's prompts. Fifty unique prompts, daily, for seven days, on three engines, in one language and one region. The FAQ says you cannot customize that set. It does not say you can swap one prompt. It says the set is recommended from the industry. I would not tell a board that the trial measured the questions the sales team hears. I would tell them it measured Profound's recommended questions, on ChatGPT, Gemini, and Google AI Overviews, for a week.

The page does not say whether each of the 50 prompts runs on all three engines every day. I am not multiplying 50 by 3 by 7. That product is 1,050 answers if every prompt hits every engine every day, and it is 350 answers if each prompt hits one engine a day. Profound does not print the formula. A trial report that says "50 prompts across three engines" has not yet said how many stored answers you will see. Ask, in writing, whether a prompt is one engine or three, and what the screen shows on day eight.

History is none. Exports are none. API is none. That is the part I would underline for anyone comparing it with Scrunch Core. Scrunch's Core price is the plan you keep, and the trial becomes that plan if you do not cancel. Profound's trial is a week of answers the pricing table says you cannot download. If the point of the trial is to see whether Profound stores the response text, you can look at the screen. If the point is to hand a file to a writer on Monday, the trial column says you cannot. Enterprise is where history becomes all-time and exports become CSV and JSON. I did not click an export, so I cannot tell you whether the CSV contains the answer text, the citation URL, or a score. The cell names the formats. It does not name the columns.

Seats are unlimited on the trial, and support is none. A large team can log in. Nobody from Profound is assigned to explain the screen, on the trial column. The FAQ says setup is instant. Instant setup is not onboarding. Enterprise support is a menu: email, dedicated Slack, or a dedicated specialist with a 24 hour SLA. I am quoting the cell. I did not test the SLA. A 24 hour line on a pricing table is not a signed service level.

ChatGPT Shopping is a separate gate, and I could read it on the comparison table: No on the trial, Yes on Enterprise. Scrunch's matching gate is product observability, off on Core and on for Enterprise. A commerce team does not get a shopping view from either entry offer. Profound's trial will not show it. Scrunch's paid Core plan will not show it either. The difference is that Scrunch will charge the Core price for a plan that still excludes it, and Profound will show you a week of non-shopping answers before the quote.

Prompt Volumes on the trial is 5 searches. The cell does not define a search. Enterprise says custom. I would not budget a prompt-research program on the trial cell. Agent Analytics is the on-site row, and it is less stingy than the prompt log. The trial says domains tracked are unlimited, and it lists integrations: Akamai, AWS, Cloudflare, Fastly, Google Analytics, Google Cloud Platform, Netlify, Vercel, and WordPress. Enterprise says custom, with options for that same list. I did not connect a CDN. I cannot tell you which bot names Profound recognizes. The pricing page says the row exists. Unlimited domains are not a 25-page sitemap, and they are not a five-page audit. Scrunch prints those audit limits. Profound's trial does not print a page cap for a content score. I am not going to treat "unlimited domains" as a contractual crawl of a large documentation site.

The engine disagreement inside Profound is why I re-read the pages on the same day. The trial card names three. The Enterprise capability list names ten: ChatGPT, Perplexity, Google AI Mode, Gemini, DeepSeek, Anthropic Claude, Google AI Overviews, Exa Search, Muse Spark, and Qwen. The card's own words are "up to 10." The Answer Engine Insights FAQ names the same ten in a different order: ChatGPT, Perplexity, Claude, Google AI Overviews, Google AI Mode, Google Gemini, Exa, Muse Spark, Qwen, and DeepSeek. A different answer on that FAQ says Enterprise plans run across up to 9 answer engines, and it tells you to see the pricing page for current plan details. The pricing page I opened says 10, and the capability list has ten names. The FAQ answer that says 9 is stale relative to the pricing page, or the pricing page is ahead of the FAQ. I will not pick the shorter number to make a chart tidy. Perplexity, Google AI Mode, Claude, DeepSeek, Exa, Muse Spark, and Qwen are not on the trial. If the board's sentence is Gemini, the trial can show a Gemini answer on Profound's prompts. If the board's sentence is Perplexity, the trial is the wrong week, and Scrunch Core is the entry card that already names Perplexity.

Regions are the same kind of split. The AEI FAQ says Profound supports 30+ languages and 150+ regions. The trial is 1 language and 1 region. Enterprise is custom. I am not going to tell a two-country brand that the FAQ's 150 regions are included because the product is capable of them. Capable and included are different rows. Scrunch is stricter in a way I prefer on a pricing page: Core says 1 country, Enterprise says unlimited. Neither entry offer covers a US product and a UK product together.

The AEI FAQ also says you can modify or add prompts at any time, and that prompting works three ways: auto-generate from the brand and topic, upload your own, or pull queries from Prompt Volumes. The pricing FAQ says trial users cannot customize prompts. I am reading those sentences as two products. The AEI FAQ describes the configured workspace. The pricing FAQ describes the trial. If a demo spends the hour on uploaded prompts and the thing you can click today is the trial, you have not bought the demo. Scrunch's trial, by contrast, tells you to add up to 125 of your own prompts. That is the practical difference between a recommended set and a prompt the founder heard on a sales call. Scrunch lets you type it during the week you are being charged toward. Profound's trial does not, on the FAQ I read.

Citations are the row I actually want from Profound, and the citations page is more careful than a logo wall. It says every cited source gets a category: Owned, Competitor, Earned Media, PR Wire, Social, or Institution. Owned, Competitor, and Custom are yours to define. The rest are auto-assigned and can be overridden. Citation share can be cut by platform, topic, or prompt. Watched pages track how citation volume on a URL changes over time. Export is CSV or JSON, or a handoff into Agents. I am putting the export sentence next to the pricing table. The trial column says exports are none. The citations page describes the product. It does not say the trial includes the file. I would not promise a writer a CSV from a trial login. Scrunch's citation definition is narrower and, on Core, it is on the plan you can pay for without a quote: a link in the answer to you, a competitor, or a third party. Profound's category system is a better taxonomy. Scrunch's Core plan is a taxonomy you can buy this week, if a link is enough and a retrieved-but-not-shown URL is not required.

The citations page also says you can build outreach lists from the publishers and authors who get cited. That is a useful object if the names are real. I did not open a list. The category system is still the right shape. A mention of the brand and a citation of a URL are different objects. If a weekly report adds them together, it has mixed two things.

Both of those citation views stop at the URL that already exists. Neither pricing page is a placement service. When the cited source is a third-party article, and the work is to earn more of those mentions on other sites, Offsite Posts writes and publishes industry articles about a brand on other sites, and it says the posts, mentions, and links in a package are guaranteed while rankings, traffic, and AI mentions are not.

Agents are the part of Profound that is not a prompt log. The Agents page describes templates for a content refresh, an FAQ, competitive research, and a new article. It describes a drag-and-drop builder with steps for scraping, prompting, Answer Engine Insights lookups, and API calls. It says workflows pull from a brand kit and a knowledge base. It says Sheets can run many inputs, with each column as an agent. It says every run includes an approval step before anything publishes, and that agents can connect to an existing CMS. It does not, on the page I read, name the CMS products. I will not add WordPress to that sentence just because WordPress appears on the Agent Analytics integration list. Those are different rows.

One sentence on the Agents page should stay attributed, because I cannot audit it from outside the product. Profound says each agent queries 16 reasoning models and Profound's user prompt data before drafting. I counted the words. I did not count 16 models in a run. The approval step is the part I agree with. On a payments, lending, payroll, or compliance page, a model that cites an agent variant can repeat a fee the legal team never approved. Leave the approval on. The page says the approval is part of the run. I did not see a screenshot of a forced approval that cannot be turned off.

Credits are the meter, and the meter is not an article count. The pricing FAQ says credits are consumed each time an agent runs, relative to complexity. Simpler agents use fewer credits. Multi-step agents use more. The platform shows an estimate before a run and the actual spend after. The trial includes limited credits. The FAQ does not say how many. The self-serve Agency Growth plan includes 400 credits a month per client workspace. Additional thresholds are an Enterprise package. I did not find a dollar next to Agency Growth. I will not invent one, and I will not convert 400 credits into a post count. The Agents page also prints an 800% visibility line from a customer story. I am not using it. The page does not show the prompt set, the date range, or the baseline next to that headline.

FactCheck is on the Answer Engine Insights page and not on the pricing table. The page says it shows inaccurate claims and their sources. The illustration uses a fee example. I am citing the illustration as an illustration. I am not saying FactCheck is on the trial. The pricing table does not have a FactCheck row. If the reason you want Profound is a check on whether the model invents a fee, ask whether FactCheck is in the trial, in Enterprise, or in neither until a package is written down. Scrunch's sentiment row, positive, neutral, or negative, is not that check. Someone still reads the answer.

Security is the strongest reason to take Profound's Enterprise conversation, and it is easy to over-read from a checkmark. The Answer Engine Insights page says Profound is SOC 2 Type II, with SSO using SAML or OIDC, role-based permissions, and daily backups retained for one week. The pricing card puts SSO and SOC 2 on Enterprise, and it puts SSO at No on the trial. I did not download the report. A page that says the audit exists is not the PDF. Scrunch's products FAQ says enterprise features include SOC 2 Type II, SAML and OAuth-based SSO, and role-based access. I am citing that FAQ as Scrunch's claim, not as a report I read, and I am not moving those controls onto the Core card. Core's security cell that I could verify is Google single sign-on.

Which Limits Actually Change the Decision?

Five limits decide this pair more often than the feature count.

The first is the engine you can name before a demo. Scrunch Core includes Perplexity and Copilot, and it does not include Gemini or Google AI Mode. Profound's trial includes Gemini and Google AI Overviews, and it does not include Perplexity or Copilot. A team that has been reporting Perplexity will feel that on day one of a Profound trial. A team that has been reporting Gemini will feel it on Scrunch Core, because Gemini is an Enterprise name there. If the weekly report has to contain both Perplexity and Gemini, neither entry offer does it. Scrunch Enterprise is a quote that adds Gemini. Profound Enterprise is a quote that adds Perplexity. A published menu that lets you pick among both, on a page with a dollar, is a different product. That review is the Peec AI review. I am not pasting Peec's prices into this table.

The second is whether you can type the prompt. Scrunch's trial FAQ tells you to add your own prompts, up to the Core allowance. Profound's pricing FAQ says the trial set is recommended and cannot be customized. For a fintech, that is often the whole test. The question a buyer asked on a sales call is the question worth storing. A recommended industry prompt can still be informative. It is not the same evidence. I would not let a Profound trial slide replace a prompt the founder wrote, and I would not let a Scrunch prompt count imply the four engines ran every day. Those are separate corrections.

The third is the file. Profound's trial has no export. Scrunch Core is a paid plan, and the pages I read do not say the export formats in a cell as clear as Profound's "CSV, JSON." Scrunch does mark a reporting dashboard on Core, and it holds Looker and the query API for Enterprise. If the writer needs a file this week, the Profound trial fails in a sentence. If the writer needs Looker, both entry offers fail, and you are in two different quotes.

The fourth is country. A brand with a US product and a UK product cannot represent both on Scrunch Core or on the Profound trial. Scrunch does not say unlimited countries until Enterprise. Profound does not say a second region until the Enterprise cell, which is custom. If the company is single-market, this limit is free. If it is not, it is the whole purchase, and neither vendor lets you buy your way out of it on a self-serve card.

The fifth is evidence versus a score. Scrunch will tell you, on the FAQ, whether the answer linked to you. Profound will describe, on the citations page, a category for that link, and then the pricing table will tell the trial user they cannot export it. Buy the row you will open. A share-of-voice movement without the sentence is how a wrong APR gets into a slide. The audit limits are the other evidence. Scrunch tells you five audits and 25 pages on the plan with a dollar. Profound tells you unlimited domains and does not tell you a page cap. I trust the number I can see. I do not trust "unlimited" to mean the whole documentation site until someone asks.

Who Each Tool Fits

Pick from the engine, the prompt you are allowed to type, and whether anyone can leave with a file.

• One country, Perplexity is mandatory, and you want bot traffic beside the prompt log. Scrunch Core, if five competitors and five seats are enough and you can live without Gemini and Google AI Mode. Put the day-six cancellation reminder on the calendar. The trial takes a card.
• You need Gemini or Google AI Overviews this week, you can accept someone else's prompts, and a quote is a normal buying step. Profound's trial. Do not brief it as Perplexity, as a CSV, or as ChatGPT Shopping. Those are Enterprise cells.
• You need Perplexity and Gemini in the same weekly report. Do not assume either entry offer does this. Price Scrunch Enterprise against Profound Enterprise, and read the Peec alternatives and the Otterly vs Peec page before you call either quote the only path.
• You need to type the prompt the sales team heard. Scrunch's trial says you can add prompts. Profound's trial FAQ says you cannot customize them. If that prompt is the measurement, Profound's trial is the wrong week even when the engine matches.
• You need a file on Monday, or Looker, or an API. Profound's trial excludes exports and the API. Scrunch holds Looker and the query API for Enterprise. A screen you cannot download is a demo.
• You sell in two countries. Both entry offers are one country or one region. The upgrade is a quote at both vendors.
• An agency with a roster. Scrunch publishes Agency Core at $500 a month and does not, on the FAQ I read, publish the allowances. Profound's FAQ describes 400 AI Marketer credits a month per client workspace on a self-serve Agency Growth plan and does not publish a dollar. I would make each vendor put the missing half in writing before I compared them. A single-brand team does not need either agency plan.
• A commerce brand that needs shopping visibility before a quote. Neither entry offer includes it. Scrunch puts product observability on Enterprise. Profound puts ChatGPT Shopping on Enterprise.
• You need the model fact-checked against a fee schedule. Profound's FactCheck illustration is not a pricing-table row. Scrunch's sentiment row is positive, neutral, or negative. Neither one is a compliance review. On a live account, Austin Heaton still reads the stored sentence before it goes in a deck.
• You wanted the tool to publish the site. Profound's Agents page describes an approval step and a CMS connection, metered in credits, with no article count and no public brand price. Scrunch's AXP cell says to get a demo, and it is not on Core. A posting schedule with a published allowance is a different category of product. I would not let either Enterprise conversation become that schedule by adjective.

The consultant and agency version of this decision, if the dashboard is not the thing you want to operate, sits on the comparison index. A tool subscription and an engagement are different bills. This page is only the tool.

What I Could Not Confirm

• A paid self-serve brand price for Profound. The pricing page shows a free trial and custom Enterprise.
• A dollar for Profound's Agency Growth credit pack. The FAQ says 400 credits a month per client workspace. It does not price the pack.
• How many Profound trial prompts run on all three engines. The card says 50 unique prompts, daily, for 7 days, on three engines. It does not print the multiplication.
• Whether the Profound trial requires a credit card. I did not click Get Started.
• Which of Profound's "up to 10" and "up to 9" sentences the order form will repeat. I counted ten names on the pricing capability list.
• Whether FactCheck is on the Profound trial. It is on the feature page. It is not a pricing-table row.
• Scrunch's run frequency on Core. The card says 5,000 responses a month. It does not say daily.
• What Scrunch does at the response cap. The pricing page asks the question. The extract did not include the answer.
• A Scrunch definition that splits a retrieved URL from a linked citation.
• The allowances inside Scrunch Agency Core. The FAQ prices it. It does not list prompts, workspaces, or countries for that plan.
• Whether Scrunch's CLI cell is a product. The Enterprise extract said "Email + Slack."
• Either vendor's collection method against a live chat I ran myself on 14 November. Profound's feature page says the capture is the consumer front end. I did not watch one. This comparison is the public pricing pages, FAQs, and feature pages. It is not a logged-in bake-off.

Frequently Asked Questions

Which is the better fit, Scrunch or Profound?

Scrunch is the better fit when Perplexity has to be on a plan with a published monthly price, you want bot traffic and a small audit allowance in the same login, and one country plus a capped competitor list is enough. Profound is the better fit when Gemini and Google AI Overviews have to appear in the first week, the buying process can absorb a quote, and a recommended prompt set is acceptable for that week. Neither entry offer includes Perplexity and Gemini together. Neither one guarantees a citation.

Does either vendor publish a paid self-serve price?

Scrunch publishes a brand plan and, on its pricing FAQ, an agency plan, and it leaves Enterprise and Agency Enterprise as custom quotes. Profound's brand pricing page shows a free trial and a custom Enterprise plan. I did not find a paid self-serve brand price there. The Profound FAQ describes an agency credit pack and does not attach a public price to it. Another vendor's page is not a substitute for either pricing URL.

Which engines can I name before a sales call?

On Scrunch Core the four engines are ChatGPT, Perplexity, Google AI Overviews, and Copilot. Gemini, Claude, Google AI Mode, Grok, and Meta AI are on the Enterprise list of nine, not on Core. On the Profound trial the three engines are ChatGPT, Gemini, and Google AI Overviews. Perplexity, Google AI Mode, Claude, and the rest of the Enterprise capability list are not on that trial. The Profound pricing card says up to 10 engines on Enterprise. One Answer Engine Insights answer still says up to 9. I would make the order form repeat the names.

Can I edit the prompts before I pay?

Scrunch's trial FAQ says you can add prompts, up to the Core allowance, during the trial. Profound's pricing FAQ says trial users receive a recommended set based on their industry and cannot customize it. Editing, disabling, and adding prompts are described on Profound Enterprise. A different Profound FAQ says prompts can be modified at any time. This comparison treats that sentence as the configured product, not as the trial.

Do they show the cited URL, or only a score?

Scrunch's tracking FAQ describes a citation as a link in the answer to the brand, a competitor, or a third party, and the pricing page marks Citations / Sources on Core. I did not find a Scrunch page that documented a retrieved-but-not-linked URL as its own metric. Profound's citations page describes categories, citation share, and an export. The pricing table says the trial includes no export. I would not promise a file from a Profound trial login.

What is on the trial, and does it need a card?

Scrunch's trial FAQ says the trial lasts seven days, includes Core features, requires a credit card, and converts to a paid subscription unless you cancel. It also says a trial user can audit five webpages. Profound's trial is a short daily run of a fixed prompt set on three engines, in one language and one region, with no export and no prompt editing. I did not see a card requirement on Profound's pricing page, and I did not start either trial.

Who should skip both entry offers?

Skip both when the weekly report has to include Perplexity and Gemini, when you sell in two countries, or when you need shopping visibility before a quote. Skip the Profound trial when you need prompts you wrote, a file this week, or Perplexity. Skip Scrunch Core when the buyer question lives in Gemini or Google AI Mode, or when five competitors and a small sitemap are too small. Use a tool whose self-serve list already includes the missing engine. Take either Enterprise conversation only when the order form repeats the engines, the prompt count, the regions, and whether the delivery product is included. A citation is still not guaranteed.

Read next:

• Peec AI Review, November 2026
• Scrunch AI Review, November 2026
• Profound Review, November 2026
• Peec AI vs Scrunch AI
• Profound Alternatives for AI Visibility Tracking
• Profound vs Peec AI vs Otterly vs Semrush

If the entry engines, the prompt you are allowed to type, and the export are three different purchases, Book a Call and we will put the prompt, the engine list, and the file on one page before the trial card is charged or the quote is booked.