Profound vs Scrunch: We Tracked the Same 15 Prompts

Profound and Scrunch both track how brands appear across AI search, but they organize the work differently. We tested the same 15 prompts in both tools for six days across ChatGPT, Perplexity, and Google AI Overviews, then compared the results and the workflows behind them.

The biggest differences were not limited to visibility scores. Profound surfaced a broader competitive field and deeper answer-level analysis, while Scrunch put more emphasis on configurable tracking, source segmentation, site diagnostics, and its Enterprise AXP layer.

Quick comparison

We Tracked the Same 15 Prompts in Both Tools for Six Days

We built the same 15-prompt test in Profound and Scrunch, tracked it for six days, and used the same three AI platforms: ChatGPT, Perplexity, and Google AI Overviews. At the end, we compared the brand-level visibility results that each tool reported for every brand.

The setup was deliberately kept identical on the variables we controlled: the prompts, six-day tracking period, and AI platforms. We did not normalize the platforms' scoring systems or convert one metric into the other.

The 15 prompts were non-branded discovery queries around AI app builders, vibe-coding tools, and AI code generators. We deliberately mixed broad category searches with more specific use cases and audience needs, rather than testing variations of one generic “best tools” query.

Examples included:

• Broad category: “What are the best AI app builders?”
• Use case: “What are the best AI app builders for SaaS products?”
• Audience: “What are the best vibe coding tools for non-technical founders?”
• Task: “What is the best AI app builder for building an MVP?”
• Adjacent category: prompts around AI website builders and AI coding assistants for creating web apps.

For the comparison, we used Profound's top 10 brands as the reference set and matched those brands against Scrunch where Scrunch had a result. A brand that was not included in the Scrunch tracking setup is marked Not tracked, rather than being assigned a zero.

The metric names matter. Profound calls its percentage Visibility Score. It calculates the percentage of responses containing the brand among responses containing at least one brand. Scrunch calls its metric Mentions Rate, which measures the percentage of AI responses whose text names the brand.

Five of the six brands measured by both tools were relatively close:

• Lovable differed by 1.2 percentage points
• Replit by 4.1 points
• Cursor by 2.7 points
• Bubble by 2.5 points
• Base44 by just 0.4 points

Bolt was the clear outlier: 50.6% in Profound versus 28% in Scrunch, a 22.6-point difference.

Both tools still placed Lovable first and Replit second among the brands they had in common. The larger difference was coverage: Wix, Vercel, Framer, and GitHub appeared in Profound's top 10 but were not tracked in Scrunch.

Why Profound and Scrunch Show Different Results for the Same Prompts

The largest difference between Profound and Scrunch was not the prompts or AI platforms, which we kept identical. It was competitive coverage. Profound surfaced a much broader brand universe automatically, while our Scrunch Core setup could track only five configured competitors at a time.

Our Scrunch account allowed six brands in total. Profound's expanded Visibility Score ranking, by comparison, contained 131 brands. Scrunch Core's current pricing also lists a five-competitor limit.

That difference materially changes what appears in the ranking.

• Wix ranked #5 in Profound at 26.9%, but Scrunch did not track it.
• Vercel ranked #7, but was not tracked in Scrunch.
• Framer ranked #8, but was not tracked.
• GitHub ranked #9, but was not tracked.
• Scrunch surfaced Wix as a suggested entity, but it was not part of the active comparison set.

This matters because a predefined competitor list can miss brands that AI answers place above the competitors you selected manually.

There was also an entity setup difference worth noting. Profound separated Bolt and Bolt.new, with Bolt at 50.6% and Bolt.new appearing lower in the ranking. Scrunch contained one Bolt entity and no alternative names in the captured configuration.

That difference could affect measurement, but the screenshots do not prove it caused the 22.6-point Bolt gap. It should therefore be treated as a setup difference, not an explanation for the discrepancy.

The narrower field also shifts rank order. Cursor was third in Scrunch but fourth in Profound, because Profound placed Bolt above it.

A fixed competitor list works when you already know exactly whom you want to benchmark. Broader automatic discovery is more useful for uncovering brands you did not think to add. In this test, four of Profound's top 10 fell outside the Scrunch comparison set.

Which Tool Gets You from Setup to Competitive Insight Faster?

In our six-day test, Profound got us from prompt setup to competitive analysis faster. Its rankings populated a broad brand field automatically and kept visibility, prompts, platforms, citations, and fanouts close together, while Scrunch required more configuration before the same analysis became equally useful.

The difference was most obvious in competitive discovery. Profound populated rankings across 131 brands without requiring us to decide in advance which names should make the comparison.

Profound was simpler in three practical ways:

• Competitors appeared automatically in the ranking. We did not have to build the competitive field first.
• The analytical views were tightly connected. Visibility, prompts, platforms, query fanouts, citations, and answer-level records sat inside the same analytics area.
• Less classification work was required before the data became useful. The main competitive picture appeared before we added more segmentation.

Profound's Sentiment profile was still building when we captured the account, with the interface indicating that processing could take up to a week. We therefore did not use sentiment to judge the two interfaces.

Scrunch takes a more configurable approach. Its setup gives more importance to competitors, topics, personas, stages, tags, and prompt organization.

That showed up in our test:

• competitors had to be confirmed and configured
• prompts could carry topics, stages, personas, and tags
• 8 of the 15 prompts were under "Unknown" topic because topics had not been assigned across the full set
• personas added another configuration layer
• the interface showed the response budget before prompts were deployed

Scrunch's flexibility becomes more useful after that setup. Explorer supports custom metrics, filters, breakdowns, and reusable views, while prompt templates can be organized around the team's own structure.

The difference is therefore less about whether both platforms can segment data and more about when the complexity appears. Profound gave us a useful competitive landscape first and let us drill down afterward. Scrunch required more structure upfront, then provided more control over how the analysis was organized.

How Deep Can You Dig Into Individual AI Answers?

Profound gave us the deeper answer-level audit trail in this test. We could open individual responses, see the full answer text, check brand position, engine, and region, then move into query fanouts. Scrunch exposed strong prompt-level breakdowns, but our captures showed less answer-level detail.

Profound lets the user move below the aggregate Visibility Score into the records behind an individual prompt. Our screenshots exposed:

• the full AI-generated answer
• the AI platform
• the region
• brand position
• citations
• individual answer records

That makes it possible to investigate what an aggregate percentage is hiding. A brand can have a high Visibility Score while appearing low in individual recommendations, or two engines can produce materially different answers for the same question.

Query fanouts add another layer

Profound also shows the searches generated underneath a prompt.

In our dataset, "What are the best AI app builders?" produced 40 query fanout variations. This makes it possible to see how one broad prompt expands into narrower searches before the final answer is produced. The captured fanout analysis covered ChatGPT and Perplexity.

Scrunch's prompt pages approach the investigation differently. The supplied screens show:

• platform-level breakdowns
• competitor presence
• citations
• prompt variants
• topic and organizational context

Scrunch's pricing currently lists Query Fan-out as Limited on Core and fully available on Enterprise.

One evidence limit matters here: our Scrunch captures do not show a full answer-text view equivalent to the Profound screen. That does not prove Scrunch lacks one, so we would not describe the feature as missing.

The distinction supported by our test is narrower: Profound made it easier to trace a high-level visibility metric down to the individual AI answer and the searches behind it. Scrunch gave us strong prompt and platform breakdowns, but not the same captured answer-level audit trail.

Profound Prompt Volumes vs Scrunch AI Search Demand Tools

Profound and Scrunch now both try to answer a harder question than simple visibility tracking: what are people actually asking AI systems? Profound approaches it through Prompt Volumes, while Scrunch combines AI Search Trends with a newer Keyword Volume Estimator built around estimated demand.

This comparison sits outside the six-day visibility test. We did not run a matched demand-volume experiment, so the useful comparison is what each research layer currently provides.

Profound Prompt Volumes

Profound separates the prompts you choose to monitor from its Prompt Volumes research dataset.

Profound says Prompt Volumes is built from licensed, anonymized consumer-panel conversations with AI assistants. The US dataset includes ChatGPT, Gemini, Claude, and Perplexity, and the data is updated weekly.

The research environment includes:

• prompt volume and historical trends
• phrase and exact-match research
• bulk topic analysis
• informational, commercial, generative, and conversational intent
• demographic breakdowns
• related terms
• samples of real prompts
• Prompt Research Reports that compare market demand against prompts already being monitored

That last layer is useful because it can expose a different problem from poor visibility: the team may be monitoring questions that do not reflect what people are actually asking AI assistants.

Scrunch AI Search Trends and Keyword Volume Estimator

Scrunch Core includes three AI Search Trends topics, giving users a smaller watchlist for monitoring estimated demand.

Scrunch's Keyword Volume Estimator adds another research layer around AI demand, including estimated weekly US volume, commercial intent, topic themes, and trend direction.

The two approaches therefore serve slightly different research workflows. Profound Prompt Volumes is the deeper research environment, particularly for discovery, prompt samples, intent, demographics, and validating what should be tracked. Scrunch provides a simpler demand layer that fits naturally alongside its topic-based monitoring workflow.

Neither type of dataset should be interpreted as a literal census of every prompt submitted to every AI system. The more useful signal is comparative demand, trend direction, intent, and whether the prompts being monitored reflect real user behavior.

Which Tool Shows You Where AI Answers Get Their Sources?

Both tools go beyond telling you whether a brand appeared in AI answers. Profound organizes citations around share, domains, pages, and page-level performance, while Scrunch emphasizes citation ownership, domain influence, and response-level citation rates. The difference is what each makes easiest to act on.

Profound's Citations area exposed several levels of source analysis in our test:

• Citation Share
• Citation Rank
• citation categories
• Top Citation Domains
• Top Citation Pages
• prompt-level citation data
• platform-level citation data
• page-level citation performance

youtube.com led the captured domain ranking with 7.2% Citation Share, while lovable.dev ranked fifth. Profound also showed the individual pages receiving citations, rather than stopping at the domain level.

Profound defines Citation Share as the percentage of all citations pointing to a given domain, page, or citation category.

Its page-level citation workflow is useful when the goal is to move from:

Which domains influence AI answers?

to:

Which exact URLs are earning that influence?

The interface also displayed a Mentioned on Page field beside individual cited URLs. We do not extend that field beyond what the captured interface supports.

Scrunch emphasizes who owns the source

Scrunch organized the same problem differently.

Its citation view split sources into:

• Brand
• Competitor
• Third party

Third-party sources represented 83.5% in the captured citation-owner view. Scrunch also ranked cited domains, with zapier.com leading our captured dataset, and exposed citation performance at prompt level.

Scrunch's citation tools can therefore answer questions such as:

• Is AI relying primarily on brand sources?
• Are competitors supplying more of the evidence?
• Are third-party publishers and communities dominating?
• Which prompts produce citations?
• Which domains repeatedly influence answers?

The metrics should not be compared number for number.

Profound Citation Share measures a source's proportion of all citations. Scrunch Citations Rate measures the share of responses that cite a registered brand URL, while Scrunch also provides separate citation-owner and source metrics.

A 7% value in one platform therefore does not automatically mean the same thing as 7% in the other.

The clearest difference is how the source intelligence is organized. Profound made it easier to move from citation share to domains, individual pages, and page performance. Scrunch made the brand, competitor, and third-party split more explicit and gave the user more ways to segment source influence.

Profound Agents vs Scrunch AXP: What Happens After You Find a Problem?

Profound and Scrunch take very different approaches once monitoring identifies a problem. Profound uses Agents to automate research, content generation, optimization, and publishing workflows, while Scrunch's AXP changes what AI agents receive from a site by serving them a separate, machine-optimized experience.

These are not equivalent features.

Profound Agents is a workflow automation system. Scrunch AXP is an infrastructure and content-delivery layer.

Profound turns the finding into a workflow

Profound describes Agents as a node-based workflow builder that can combine its AEO data, AI models, external tools, and publishing actions.

An Agent can be built to:

• pull visibility and citation data
• identify content gaps
• research a topic
• generate or optimize content
• apply conditional logic
• iterate across multiple inputs
• send outputs to Slack
• connect with external APIs
• create or update content in connected publishing systems

Profound Growth includes 400 Agent credits per month, while Starter includes 100. Credits are consumed according to workflow complexity.

The important point is that the finding can become an automated process rather than another dashboard item.

Scrunch AXP acts at the retrieval layer

AXP takes a different route.

Scrunch describes AXP as an infrastructure layer that creates a parallel AI-optimized version of a site. AI retrieval agents can receive structured, machine-oriented content while human visitors continue receiving the normal website experience.

That can involve:

• simplifying content for machine retrieval
• reducing dependence on JavaScript-heavy presentation
• adding clearer structured information
• serving optimized content specifically to AI agents
• keeping the normal human-facing experience unchanged

Scrunch Core includes monitoring, audits, recommendations, and one page optimization per month, but AXP is reserved for Enterprise.

So the action layers point in different directions:

• Profound Agents: automate what the marketing or content workflow does next.
• Scrunch AXP: change how approved site content is delivered to AI retrieval agents.

We did not run an Agent workflow and AXP deployment as part of the six-day test, so there is no basis for claiming that one generated a larger visibility improvement.

The meaningful difference is architectural. Profound turns diagnosis into automated AEO work. Scrunch can intervene closer to the retrieval layer by changing the version of site content AI agents receive.

What $250 vs $399 Actually Buys You

At the self-serve level, Scrunch Core costs $250 per month while Profound Growth costs $399 per month. The price gap matters, but the plans are not equivalent: Scrunch includes more prompts and one extra AI platform, while Profound bundles deeper analytics and Agent credits.

Profound does have a cheaper $99 Starter tier, but it tracks ChatGPT only. For a multi-platform comparison, Growth is the more relevant Profound plan.

Scrunch Core: $250/month

Core currently includes:

• 125 unique prompts
• 4 AI platforms: ChatGPT, Perplexity, Google AI Overviews, Copilot
• 5 competitors
• 5 users
• 1 brand workspace
• 5 site audits per month
• 3 AI Search Trends topics
• Limited Query Fan-out
• one page optimization per month
• basic content generation
• Agent Traffic monitoring

That is a broad self-serve package for $250. Its main restrictions are also clear: the competitor set is capped, Query Fan-out is limited, and AXP sits on Enterprise.

Profound Growth: $399/month

Growth includes:

• 100 prompts
• 3 AI platforms: ChatGPT, Perplexity, Google AI Overviews
• 400 Agent credits per month
• full access to the broader Profound analytics workflow shown in our test, including visibility rankings, prompt analysis, citations, answer-level records, and query fanouts

The extra $149 per month does not buy more prompt capacity or broader engine coverage. Scrunch wins both of those rows.

Profound's premium is easier to justify when the team values the areas where it was stronger in our test: automatic competitive discovery, deeper answer inspection, prompt research, query fanouts, citation-page analysis, and Agent workflows.

Scrunch Core is the stronger value when the priority is more prompts, four AI platforms, five users, site auditing, and configurable monitoring for $250.

Final Verdict

Profound and Scrunch overlap on the core job of monitoring AI visibility, but our test showed two different operating models. Profound prioritizes broad competitive discovery and deeper investigation, while Scrunch emphasizes configurable monitoring, structured segmentation, site diagnostics, and an Enterprise content-delivery layer through AXP.

The six-day test produced broadly similar results for most of the brands both platforms measured, but the competitive coverage difference mattered. Profound surfaced brands automatically that were outside Scrunch's configured competitor set, including four names inside Profound's top 10.

Profound also gave us the stronger investigation workflow once a result needed explaining. Answer-level records, query fanouts, Prompt Volumes, citation pages, and Agents create a deeper path from visibility score to diagnosis to action.

Scrunch's case is different. At $250/month, Core includes 125 prompts, four AI platforms, five users, site audits, configurable segmentation, and a wider self-serve monitoring package than Profound's $399 Growth tier on several capacity measures. AXP then gives Enterprise customers a distinctly different way to act on AI visibility by serving optimized content directly to retrieval agents.

The choice comes down to workflow:

• Choose Profound when deeper competitive discovery, answer-level research, prompt intelligence, and automated AEO workflows matter most.
• Choose Scrunch when lower-cost multi-platform monitoring, configurable analysis, site diagnostics, and the option to move into AXP matter more.