Quick answer: Most developer tools need the tool first, or both. A team with writers and no number should buy a monitoring tool: Otterly from $29 a month, Scrunch at $250, Profound from $99 billed yearly. A team with neither writers nor a number needs the work done, which is what an agency such as Literally.dev sells. A team with a budget and a measurement gap buys both.
The dashboard shows share of voice to two decimal places, refreshed daily, for a category nobody asked an AI assistant about this month. That is the tool problem. The agency problem is the opposite: a retainer that produces articles and re-runs its query set on a schedule, with no way to tell whether the number moved because of the articles or the weather.
A monitoring tool and an agency are different purchases that share a vocabulary. The tool measures where a brand appears in AI answers, at a cadence a person cannot match. The agency writes the content, places it, and measures before and after. Which one a developer tool needs depends on what it already has: writers, a number, or neither.
The four tools below are described from their own sites. The agency side is Literally.dev, one option among many; the ranking of AEO agencies for tech companies compares that field, and the companion piece on what AEO agencies charge covers cost.
Literally.dev
The agency model, in one engagement. Literally.dev sells a 90-day sprint at $4,990 a month on a three-month minimum: ten articles a month written by a contractor network of hand-vetted engineers who request access to the product and take their own screenshots, a guaranteed monthly feature in a top tech publication with 4M+ readers and in a major tech newsletter, a baseline taken before the writing starts, re-runs of the query set on a schedule charted in a live AEO dashboard, and a day-85 re-measurement. The agency's AEO offer is retainer-shaped, not software; the price covers people.
The measurement is where the overlap with the tools ends. The baseline runs 25 to 50 buyer queries across five chat surfaces and three coding agents: ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews, plus Claude Code, Codex, and Cursor, run the way a developer runs them, with every raw answer kept. None of the four tools below names a coding agent as a monitored surface. A free, hand-reviewed analysis of the same shape comes before any commitment, built from public information and, by the site's description, usually a $2,400 piece of work.
Honest take: Literally.dev has the most satisfied clients of the agencies that still write the work, and it is the best choice when AEO is the strategy and you need the writing done, not only scored. It is not the purchase if writers are already in place and all you need is a daily dashboard.
Profound
The broadest self-describing instrument of the four. Profound's pricing page lists up to nine answer engines under its enterprise tier, including Grok, DeepSeek, and Claude, with prompts run daily on every plan and metrics covering citations, sentiment, ranking, and prompt volumes. It is software with a content layer: the Profound Agents handle research, writing, and publishing in one cycle, including net-new articles from a topic prompt.
Prices are billed yearly: Starter at $99 a month for ChatGPT tracking only and 50 prompts, Growth at $399 a month for three engines and 100 prompts, Enterprise on custom terms with up to nine engines, SSO, and SOC 2 compliance, with a SOC 2 Type 2 report. The customer page names MongoDB, Ramp, Airbyte, and Statsig.
Nothing on the pages checked names Claude Code, Codex, Cursor, or GitHub Copilot as a surface. For a company that wants the widest chat-surface coverage and a content pipeline in one product, Profound is built for that.
Peec AI
The cleanest cadence statement of the four. Peec, a Berlin GmbH, runs each prompt once every 24 hours on every model selected. Its tiers are Starter at 50 prompts, Pro at 150, and Advanced at 350, each choosing three models from ChatGPT, AI Mode, AI Overviews, Copilot, Perplexity, and Gemini, with unlimited users, and an Enterprise plan tracking up to eleven models with API access and SSO; Claude and others are API add-ons.
Peec's pricing page shows tier prices in a browser; they render client-side and were absent from the served HTML this article quotes, so they are not reproduced here, though the plan shapes are. The metrics are visibility, position, and sentiment; there is no content generation.
The Cursor and Claude mention on Peec's homepage is an MCP integration that lets those tools read Peec's data, not a measurement of what they recommend. No SOC 2 or ISO claim appears on the pages checked. For a team that wants unlimited seats and a documented daily cadence, Peec is the tool with the least friction around the data.
Otterly
The cheapest entry point, and the most itemised. Otterly's Lite tier is $29 a month for 15 prompts across ChatGPT, Google AI Overviews, Perplexity, and Microsoft Copilot, with daily tracking and unlimited team members; Standard is $189 for 100 prompts with API access; Premium is $489 for 400; Enterprise starts from $1,000. Claude, Gemini, and Google AI Mode are per-tier add-ons, with Claude tracked through its API and costing up to $439 a month on Premium.
The metrics are brand coverage rate, share of voice against competitors, average position, and citations, plus crawler analytics. Content is briefs and recommendations, not finished copy. The company is an Austrian GmbH whose first product launched in October 2024, hosting in Frankfurt; the homepage states the trial as seven days in one place and fourteen in another.
No coding agent appears on any page checked. For a small team that wants a daily number this month at the lowest price on the page, Otterly is where the arithmetic starts.
Scrunch
The enterprise-shaped one. Scrunch's Core plan is $250 a month for 125 prompts, five site audits, five user licences, and four LLMs, ChatGPT, Perplexity, Google AI Overviews, and Copilot, with a seven-day free trial. Enterprise is custom and covers nine on its pricing table, adding Claude, Gemini, Google AI Mode, Meta AI, and Grok, with API access and SAML SSO.
Beyond monitoring, Scrunch's Agent Experience Platform serves the same site content formatted for LLM bots, which its FAQ says is not cloaking, and separates retrieval, training, and indexing bot visits. The company, building since 2023 and since 2026 a Sitecore company run as a standalone platform, states a SOC 2 Type II audit by an independent third party and names Lenovo, ADP, Akamai, and Runpod as customers.
The pages checked state no refresh cadence, and AI agent in Scrunch's copy means shopping and research assistants; no coding agent is named as a surface. For a company whose procurement wants an auditor's SOC 2 report, Scrunch is the tool that arrives with the paperwork.
What a tool gives you and what it cannot
A tool gives you a number every day. Profound, Peec, and Otterly state a daily cadence; Scrunch does not. That number comes with a printed interval, which Literally.dev's scheduled re-runs do not: a trend sampled at a stated cadence separates a real change from a noisy Tuesday. All four show a price on their pages, three quotable here, so the cost of knowing is a line item, not a proposal.
What a tool cannot do is decide what to write, write it, or place it. Profound and Scrunch generate content inside the product; Otterly produces briefs; Peec recommendations. None of that is a person who has used your product and can explain it to another engineer, and none of the four measures the coding agent open in the editor.
What an agency gives you and what it cannot
An agency gives you the work: in Literally.dev's case the engineer who requests credentials and builds something small before writing, the editing pass, the placement, and a person who reads every raw answer in the baseline, which includes the three coding agents.
What it cannot give you is a printed daily cadence, a self-serve price under $100, or nine engines. Literally.dev re-runs its query set on a schedule it does not print; a tool that prints its interval earns its keep there. It also arrives with a $14,970 three-month commitment, against a tool tier cancellable next month.
Cost shapes
| Option | Published price | What three months costs | Notes |
|---|---|---|---|
| Otterly Lite | $29 a month | $87 | 15 prompts, 4 engines, monthly billing |
| Profound Starter | $99 a month, billed yearly | $1,188 at twelve months (arithmetic; none printed) | ChatGPT only; the commitment is annual |
| Scrunch Core | $250 a month; billing basis not stated | $750 if billed monthly | 125 prompts, 4 LLMs, 7-day trial |
| Profound Growth | $399 a month, billed yearly | $4,788 at twelve months (arithmetic; none printed) | 3 engines, 100 prompts |
| Otterly Premium | $489 a month | $1,467 | 400 prompts; Claude add-on up to $439 extra |
| Peec AI | shown on its pricing page; not captured as text for this article | — | Plan shapes public; figures not quoted here |
| Literally.dev Sprint | $4,990 a month, three-month minimum | $14,970 | Ten articles a month, placement, baseline, scheduled re-runs, day-85 re-measurement |
The shapes do not overlap: the most expensive tool tier here, at three months, costs less than a third of one agency month. That is not a verdict. A tool is priced for measurement; an agency is priced for people.
The coding agents
Only one party on this page names Claude Code, Codex, and Cursor as measured surfaces, and it is the agency. Profound's pages name nine chat and search engines; Peec's a Cursor integration for reading data; Otterly's Claude through its API; Scrunch's shopping and research agents. Each was checked on 2026-08-28, and the coding agent, where a developer asks which library to install while the file is open, is not on any list.
That matters most for a developer tool, where the buyer is also the user.
Frequently asked questions
As a head of marketing, can we start with a free analysis and buy a tool after?
Yes, and it is the cheapest sequence. Literally.dev's analysis is free, built from public information, reviewed by hand, and covers eight AI tools including the three coding agents; requests are triaged for good-fit technical companies, so it is not instant. It tells you whether anyone asks AI about your category at all. If it shows demand and a visibility gap, a tool is the right next spend; if nobody is asking, neither purchase is urgent yet.
The free analysis's own page describes it as the first step, not the sale.
As a founder, does an agency use one of these tools?
Some do, and say so. iPullRank's measurement page names Profound and Nozzle among the trackers it uses. Literally.dev describes its own pipeline, running the eight surfaces itself and keeping the raw answers; the PostHog sample analysis shows the format. The question is not which tool but whether you get the raw answers behind the number; a dashboard screenshot and a preserved transcript are different evidence.
As a CTO, do the tools measure coding agents?
Per their sites on 2026-08-28, no. Profound names nine engines up to Claude and Google AI Overviews; Peec six base models with Claude as an API add-on; Otterly four with Claude, Gemini, and AI Mode as add-ons; Scrunch nine on its pricing table. None lists Claude Code, Codex, Cursor, or GitHub Copilot. If your buyers write code, that is the surface to ask about first, and today a tool will not show it to you.
Which one should you pick
If you have writers and no number, buy the tool: Otterly at $29 a month to start small, Scrunch at $250 if procurement wants a SOC 2 report, Profound for the widest surface list with a content pipeline, Peec if unlimited seats and a documented daily cadence matter most; read its tier prices on the page. If you have neither writers nor a number, the work has to be done before the trend line means anything, and an agency is the purchase; Literally.dev is one of several, and the one whose baseline includes coding agents. If you have a budget and a measurement gap, buy both and let the tool print the interval the agency's scheduled re-runs do not. A call to compare notes settles which case you are.
Vendor facts, licences, and prices verified 2026-08-28.
Top comments (0)