Skip to content

Compare Amazon advertising software

Perpetua alternatives, compared on the same rows

Perpetua's ad engine runs in its browser app against a target you set. Pacvue fits agency teams across many retailers. Teikametrics fits smaller sellers who want published prices and a trial. Quartile fits brands who want a dedicated strategist and channels beyond Amazon. Kuudo fits operators whose own AI runs advertising on Amazon's surface, beside the rest of the job.

Kuudo facts come from the tools reference for the Amazon Ads MCP and the Skills reference. Perpetua, Pacvue, Teikametrics and Quartile facts come from their own sites, with the date each one was checked.

Thomas Spicer Updated

Compared on · Vendor facts checked

Side by side

Where they actually differ.

24 rows. The first 11 follow the Amazon advertising job step by step, from connecting to your account through to a change going live: what Perpetua and each alternative do at each step, and how Kuudo does it. Beyond PPC, you get Amazon Marketing Cloud, seller operations, Vendor Central, image and video generation, ready-made and custom Playbooks, Amazon operating knowledge, your own data lake, and the AI clients you want to use. Each competitor cell links to the page it was checked against and the date.

Perpetua alternatives, compared on the same rows: 24 criteria compared side by side
CriterionPerpetuaPacvueTeikametricsQuartileKuudo
The job Perpetua does, from connecting to your account through to a change going live
Agent access When my AI connects, does it reach Amazon or the vendor's app?No MCP; web app only22MCP to Pacvue, read-only10MCP to Teikametrics, read-only47No MCP; REST API to Quartile39Amazon's APIs, directly1
Live connection Is it connected to my Amazon Ads account?Yes: Ads API and SP-API30Amazon APIs; Vendor Central named13Yes: Ads, Seller, Vendor, Walmart58Ads API and Marketing Stream40Yes: three Amazon MCP servers6
Reads performance Does it pull my campaign, search-term, and placement reports itself?Yes; hourly on Growth and up31Yes13Yes; search-term view unverified60Yes; search-term reports unnamed40Yes, through a Skill8
Structure audit Does it audit campaign structure and portfolio health?Builds its own; no audit31Restructuring; Agent audits14Partial: flags products, not structure61Its own structure; no audit40Yes, audit only8
Search-term mining Does it find negatives and rising terms, and show its reasoning?Yes; thresholds, not reasoning32Yes; Agent explains why14Yes; thresholds shown, not reasoning62Not published40Yes, gated negatives8
Bids and budgets How does it change bids and budgets?Its algorithm to your target33AI models plus rules you set13Its algorithm, within your limits63Yes: hourly AI or your rules40Proposals, guarded on write6
Budget pacing Does it pace budgets to a plan?Daily budgets; tracker only34Yes; pacing and flighting13Monitoring and forecast only64Not published40Yes, as a proposal8
Applies changes Does it write the change to Amazon, or do I?Automatic through the Ads API33Rules auto-apply; Agent on approval14Bids automatic; keywords optional65Writes changes automatically40Yes, guarded by your policy6
Checks the result Does it confirm a change took and report the effect?Status read-back; no outcome report35Change logs with impact15Counts changes; no confirmation66Not published40Read-back, not a step1
Ad products Which Amazon ad products does it cover?SP, SB, SD, DSP; Walmart, Instacart36SP, SB, SD, DSP, AMC, retailers13SP, SB, SD, DSP; Walmart48Sponsored Ads, DSP, 6 channels41SP, SB, SD, STV, DSP, AMC1
Who decides What decides a change: its algorithm, rules I set, or a playbook I wrote?Its algorithm and your rules37AI models, your rules, approved Agent15Its algorithm; general reasons only49Its AI, or rules you set40Your playbook8
Beyond PPC
Amazon Marketing Cloud Can it query Amazon Marketing Cloud?You connect an AMC instance in Perpetua, can write SQL queries against it, build template-based AMC audiences that apply as bid multipliers on Sponsored Products and Brands and to DSP, and view a consumer lifetime-value report without SQL. Custom AMC reporting is listed on the Premium plan.23A drag-and-drop AMC canvas, a vetted SQL query library, Amazon-certified and Pacvue-trained audience models, and pushing AMC audiences into Sponsored Display and DSP. Pacvue Agent converts business questions into AMC-ready SQL.16Pricing lists integrated Amazon Marketing Cloud on Advanced and Enterprise; the help center says AMC is for managed and agency editions, offers prebuilt AMC dashboards and template-built AMC and DSP audiences for accounts with a DSP connection. Custom SQL AMC queries are not documented.50Quartile says it uses Amazon Marketing Cloud data to build personas and tailored audiences and to inform full-funnel strategy. No page describes customers running their own AMC queries or managing AMC audiences themselves.40Yes. Amazon Marketing Cloud queries and audiences run through the Amazon Ads MCP, and the Turn AMC Questions Into Answers and Audiences Skill returns validated SQL, privacy-safe results, and a segment ready to activate in DSP.8
Seller operations Can it run orders, inventory, pricing, and catalog work?The SP-API connection supplies product data such as new ASINs and inventory levels, plus sales and traffic reporting, and rules can pause ads on inventory levels. No page describes editing listings, prices, orders or catalog.30Pacvue tracks inventory, Buy Box, pricing and product content, alerts on sales, inventory and purchase orders, and routes issues to sales, supply chain or marketing. It is not described as managing orders, editing listings or changing prices on Amazon; the only writes it makes are to ad spend.17Beyond ads it reads Seller Central product and sales data, provides inventory demand forecasts and stockout alerts as reporting, generates listing content that is held for review before submission, and offers FBA refunds recovery through a partner at a 25 percent commission. Order management and repricing are not published.51Beyond ads, Quartile manages Amazon promotions (coupons, Lightning Deals, Prime Exclusive Deals, Subscribe and Save) and its API reads listing status, pricing, inventory level, quality scores and FBA inventory. Order handling, repricing and catalog edits are not published.42Yes. 304 Selling Partner operations for orders, inventory, pricing, catalog, and reports, through the Selling Partner MCP.9
Vendor Central Can it run Vendor Central work?Vendors see Amazon Retail Analytics sales data and Rapid Retail Analytics inventory alongside advertising data. No purchase-order or other Vendor Central operations are published.24Revenue Recovery audits up to two years of Vendor Central data for invoice shortages, chargebacks, price claims and co-op overbilling, raises purchase order and invoice mismatch alerts, and identifies and files disputes. Whether disputes are filed inside Vendor Central or prepared for the user is not stated.18Vendor onboarding grants a Teikametrics user Vendor Central report permissions so total sales and inventory flow in; ordered revenue is limited to brand-registered vendors, inventory refreshes weekly and sales lag up to 72 hours. No purchase-order or retail-analytics workflow is published.52No page or doc mentions Vendor Central, purchase orders or retail analytics.43Yes. 40 Vendor Central operations: purchase orders, direct fulfillment, shipments and invoices, and Retail Analytics, through the Amazon Vendor Central MCP.6
Creative Can it generate listing and ad creative, images and video?Perpetua's DSP Creatives Builder assembles ads from assets you bring: you pick up to 20 products, a product image pulled from Amazon or a custom one, plus your logo and headline, and static images and video are uploaded to spec. The builder validates and submits what you supply. No image, video or listing copy generation is published on it.38Pacvue publishes image generation for Sponsored Brands: its GenAI tool creates custom visuals during campaign setup from the product images on your Amazon detail page plus a keyword-aligned text prompt, built on Amazon's image generation. Pacvue reports that advertisers using Image Generator increased campaign submissions by over 20%. The capability it publishes is images; its current platform and AI pages describe agents, automation and reporting, not video or listing copy generation.21Teikametrics generates listing content and images through GenAI Smart Pages: AI-crafted titles, bullet points and descriptions for Amazon and Walmart listings, plus AI-generated product images tailored to each platform's guidelines. The page publishing the image capability is an early access signup. It describes no video generation.67Quartile's Amazon Sponsored Ads page lists flexible campaign structure, automated hourly bidding, Amazon Marketing Cloud signals, rules-based optimization and reports, and claims no creation of images, video or listing content. Its Amazon DSP launch tool is described as component-based creative launches with bulk uploads, which manages creative you already have. Checked the Sponsored Ads page, the homepage and the DSP launch tool page; nothing published covers generating creative.40Yes. Agent Iris generates listing images and product video through Google Gemini, OpenAI image models, and Google Veo 3, audits an image against listing rules, publishes to the listing, and confirms Amazon took it. Video comes back as a signed URL rather than an upload to Seller Central.5
Playbooks Does it come with playbooks, and can I write my own?Readable playbooks exist, such as AMC audiences for sponsored ads and seasonal recommendations and strategies, plus free Ad School courses. User automation is limited to configurable rules, strategies and scheduled changes; no page offers user-written automation logic.25Users can apply rule-based optimizations from Pacvue's template library or build their own rules and thresholds. Pacvue Agent ships fixed skills for data analysis, recommendations, report building, guided actions and AMC queries; no user-readable or user-written playbooks for the Agent or its models are published.19Campaign creator offers basic, auto, manual and custom structures and named goals such as grow, profit, launch and liquidate, and the help center publishes readable strategy guides. User-authored logic is limited to custom filters with performance thresholds and dayparting schedules; no user-written playbooks or scripts.53Customers can write their own bid and budget rules, apply them to campaigns and schedule them. No published library of playbooks or strategies the customer can read exists, and no automation logic beyond rules is described.4012 verified Skills for Amazon Ads and seller work, each publishing what it is not for, and you can write your own in the same format.8
Operating knowledge Does it bring Amazon know-how to the job?Perpetua publishes Amazon strategy content, but its pages describe recommendations as generated from your strategies and goal performance and do not state that Amazon operating rules or know-how are encoded in them.26Pacvue states its recommendations come from proprietary machine learning models built for commerce media, that Pacvue for Amazon is purpose-built for Amazon's complexity, and that in-platform assistance delivers module-specific best practices. The know-how itself is not published or inspectable.15Marketing calls it retail-trained AI; the bidder is said to fold in seasonality, day of week and Prime Day, the recommendation hub pre-fills goals and cost limits from product performance and category trends, and Compass benchmarks aggregate Teikametrics users. Amazon policy or operating rules are not described as inputs.54Amazon know-how is delivered by a dedicated human strategist and channel experts alongside the AI. No page states that Amazon operating rules or best practices are encoded into the recommendations themselves.44Agent Atlas supplies curated Amazon operating playbooks and rules that your agent retrieves before it acts.2
Your data Where does your account data end up?Advertising data lives in Perpetua's platform; custom reporting and dashboards export CSV, and no warehouse destination or public API is published. The terms say customers own their materials but Perpetua may freely use data to improve its services; the privacy policy gives no retention period or hosting location.27Data lives in Pacvue's platform; the DaaS product pushes daily feeds into Snowflake, BigQuery and other warehouses, Console exports reports, and MCP returns CSV or Excel with up to two years of history. No retention period is published; the terms allow Pacvue to use customer data in aggregated form.20Advertising data lives in Teikametrics' platform; users can export tables and Compass data as CSV or Excel, but no warehouse destination or customer API is published. The terms say the company has no obligation to retain customer data after termination and may use it for aggregated metrics.55Advertising data lands on Quartile's US servers and the terms let Quartile retain data derived from the services. Customers can pull campaign, keyword and report data through the REST API; no warehouse destination is offered. Click-tracking data is kept up to about 13 months.45Agent Flow lands your Ads and Selling Partner data in a data lake in your own cloud; your warehouse and BI tools read it downstream.4
Where you use it Which apps can I use it from?Perpetua is used through its browser app. Help-center searches for mobile app and API return no mobile app, public API, or ChatGPT, Claude or MCP integration; the Premium plan lists custom data integrations without detail.22Web Console, Pacvue Agent inside Slack, and a Pacvue MCP endpoint for Claude, ChatGPT, Cursor, VS Code or any MCP client. The MCP is read-only today (reports, data queries, share of voice); a write-capable campaign MCP is described as in development. No mobile app or public API documentation is published.10A web app with an in-app assistant chat, a Teikametrics MCP server for Claude and ChatGPT in early access with twelve read-only tools and no write actions, and daily Slack budget-pacing notifications. No mobile app and no customer-facing API are published.47Customers use the Quartile portal and a documented REST API with OAuth2, a developer portal, batch endpoints and a Postman guide; API subscriptions must be activated by Quartile. No mobile app, ChatGPT or Claude integration, or MCP server is published.39Claude, ChatGPT, Cursor, Copilot, n8n, or any MCP client.7
How it runs, and how you buy it
Where it runs Where does it run?Perpetua is a hosted web application with user credentials issued by Perpetua. No option to deploy in the customer's own cloud account is published.28Pacvue is used through its hosted Console, which the MCP page calls the governance, configuration and operational hub. No option to deploy in the customer's own cloud account is published.10Teikametrics is a hosted web application; the terms define the services as Teikametrics-operated websites, platform and software accessed over the internet. No option to deploy in the customer's own cloud account is published.55Quartile runs its own hosted platform; the privacy policy says it operates in the United States and stores data on its own servers. No option to deploy in the customer's cloud account is published.45Your own AWS, GCP, Azure, or Cloudflare account. You own the deployment, the data, and the perimeter.7
Approval What happens before anything changes on Amazon?Rules and keyword harvesting can be set to manual approval in a pending-approval tab or to apply automatically, and recommendations require approve or dismiss. The engine's own bid changes have no approval step.25Teams can apply approval workflows and guardrails and choose the level of automation. Pacvue Agent actions are approval-based in every case, while rule automations run continuously once enabled.15There is no approval queue for bids: once AI bidding is enabled they post automatically, and manual bid edits are only available when automation is off. Keyword and negative filters default to review mode with accept, snooze and dismiss, and can be switched to automatic.53No review queue, approval step or manual mode is described. The AI updates bids, budgets and placements in real time and customer rules run automatically on schedule.40Read tools return data only. Write tools are guarded and may require your explicit approval before a change reaches Amazon.6
Pricing model How is it priced?Essentials is $695 a month for up to $10,000 of monthly ad spend; Growth is $695 a month plus an undisclosed percentage of ad spend above $10,000; Premium is custom above $500,000. Annual commitment is not published; the terms say contracts auto-renew per the order form.22No pricing page exists; the pricing and plans URLs return 404 and the demo, contact and platform pages carry no prices, plan tiers, percentage-of-spend terms, minimums or contract length. The only route is a demo request.11Essentials is $179 a month, or $149 a month billed annually, for up to $10,000 of monthly ad spend; above that, plans add 3 percent of monthly ad spend over $10,000. The help center also lists Essential Tier 2 at $599, Advanced at $1,439, Enterprise as custom, and a free basic plan without AI.56No prices are published and there is no pricing page. The terms say fees are set in an order form under standard pricing tiers based on aggregate monthly ad spend, number of channels and service level, due in advance and non-refundable, on a one-year auto-renewing term.46Kuudo Pro: $0 a month plus $0.95 per 1,000 Amazon API calls. Kuudo Scale: $249 a month plus $0.55 per 1,000 Amazon API calls. No percentage of ad spend, and the calculator models a month for a brand owner, an agency or aggregator, and an automation-heavy seller.3
Seats and organizations How many people and accounts?There is no limit on team members, and unlimited accounts can be added as long as combined monthly ad spend stays within the plan's spend tier. Manager accounts add administrator and manager roles for agencies.29No user, seat or account counts and no plan tiers appear on the Amazon, automation, orchestration or demo pages; the Amazon page lists user roles only. The MCP page says access scopes to your existing Pacvue account permissions.13Neither the pricing page nor the billing article states user, seat or account limits; plans are tiered by ad spend. Roles are account owner, admin and viewer, and an agency edition manages multiple nested client accounts.57No user, seat or account limits are published for any tier; the solutions pages, terms and API docs refer to a single Quartile account without counts.44Kuudo Pro: up to 25 seats across 5 organizations. Kuudo Scale: up to 50 seats across 10 organizations. Enterprise: unlimited.7
Free tier Can I try it free?The pricing and features pages list no free trial or free tier; the call to action is a demo. The only trial in the help center is a 60-day trial of a paid AMC purchase dataset.22No free trial or free tier is mentioned on the homepage, demo, contact, platform or Amazon pages; the entry point offered is a personalized demo.12The trial is described as risk-free for 30 days for sellers with up to $10,000 of monthly ad spend, converting to $149 or $179 a month. Onboarding states data sync starts after a credit card is entered and a plan selected; cancelling before the trial ends drops the account to the free basic plan.59No free trial or free tier is published. The terms define a 60-day evaluation period in which the client may terminate for convenience through a button in the portal, effective 30 days after clicking; fees are otherwise due in advance and non-refundable.46Kuudo Community is free: one instance of each Amazon MCP server in your own cloud, with no usage charges.7

Their fit

Choose Pacvue if

Real situations where Pacvue is the better buy. If one of them is yours, buy it.

  • You buy media on more retailers than Amazon and want one team working in one place.

    Pacvue names 100 or more retailers including Walmart Connect, Instacart, Target, Kroger, Criteo and TikTok Shop, alongside Sponsored Products, Sponsored Brands including video, Sponsored Display, Amazon DSP and AMC. Sponsored TV appears there only as something measured through AMC. Kuudo's published MCP servers are for Amazon Ads, Selling Partner and Vendor Central, and this page claims no other channel.

  • Money owed back on purchase orders is a job you want done for you.

    Pacvue's Revenue Recovery audits up to two years of Vendor Central data for invoice shortages, chargebacks, price claims and co-op overbilling. It raises purchase order and invoice mismatch alerts, then identifies and files disputes. Kuudo's Vendor Central MCP puts the purchase orders, shipments, invoices and Retail Analytics behind that audit in front of your agent; the filing is Pacvue's service, and the reason to choose it is that you want it done for you.

  • Your analysts and your sign-offs already live in one vendor console.

    Pacvue publishes a drag-and-drop Amazon Marketing Cloud canvas, a vetted SQL query library and Amazon-certified audience models with a path into Sponsored Display and DSP, alongside approval workflows, guardrails and change logs on every Agent action. Those are the vendor's screens for jobs your own agent runs on Amazon's surface; the reason to choose them is a team that already works there.

Their fit

Choose Teikametrics if

Real situations where Teikametrics is the better buy. If one of them is yours, buy it.

  • You want to see the price and try the product before you talk to anyone.

    Teikametrics publishes its plans: Essentials at $179 a month, or $149 a month billed annually, for up to $10,000 of monthly ad spend, then 3 percent of monthly ad spend above that, with Essential Tier 2 at $599 and Advanced at $1,439. The trial is described as risk-free for 30 days for sellers under $10,000 of monthly ad spend, and cancelling before it ends drops the account to a free basic plan without AI.

  • You want an hourly bidder with fences you set yourself.

    Goal-based bidding sets bids hourly on Amazon from a forecast conversion rate and payoff per order, inside a goal, an advertising cost of sales limit, a bid modifier and optional minimum and maximum bids. The bidder does not change campaign budgets. Campaigns built elsewhere can opt into it.

  • You manage to a monthly number and want the reporting to tell you where you stand.

    Budget pacing takes a monthly budget goal per merchant and shows month-to-date spend, a daily breakdown, a full-month forecast, and a daily Slack or email snapshot. It reports rather than adjusts, which is the point if your team wants to make the call.

  • Walmart is a real line in your plan, not an afterthought.

    Walmart ads connect through an API partner integration with write access, plus Retail Link for product data, and coverage runs to Walmart Sponsored Products, Sponsored Brands, video and onsite display, with TikTok Shop as well. Goal-based bidding sets Walmart bids daily.

Their fit

Choose Quartile if

Real situations where Quartile is the better buy. If one of them is yours, buy it.

  • You want a person who owns the account, not another screen for your team to run.

    Quartile delivers Amazon know-how through a dedicated human strategist and channel experts working alongside the AI. Its algorithm changes bids, budgets and placements automatically on an hourly cadence from Amazon Marketing Stream data, so the day-to-day moves without your team in the loop.

  • Amazon is one of six or seven places you buy media.

    Quartile covers Amazon Sponsored Ads and Amazon DSP alongside Walmart, Instacart, Criteo for Target and Best Buy, Google Ads, Meta Ads and Microsoft Ads. Amazon's individual ad products are not named on its pages, so ask which of Sponsored Products, Brands, Display and TV are in scope for your account.

  • You want changes applied continuously, with no queue for anyone to work.

    Quartile applies changes to Amazon itself with no user step described: bids, budgets and placements update in real time from Marketing Stream data. Customers can also write their own bid and budget rules and run them on a schedule, alongside or instead of the AI.

  • Promotions and deals sit in the same plan as the ads.

    Beyond advertising, Quartile manages Amazon promotions: coupons, Lightning Deals, Prime Exclusive Deals and Subscribe and Save. Its documented REST API with OAuth2 also reads listing status, pricing, inventory level, quality scores and FBA inventory, with a developer portal, batch endpoints and a Postman guide.

Our fit

Choose Kuudo if

Where Kuudo is the better buy. A PPC tool does one Amazon job, in the vendor's app, with the vendor's AI. Kuudo does the whole job in your cloud, from the AI you want to use. Beyond PPC, you get Amazon Marketing Cloud, seller operations, Vendor Central, image and video generation, ready-made and custom Playbooks, Amazon operating knowledge, your own data lake, and the AI clients you want to use.

  • The whole Amazon job

    You want the advertising job and everything beyond PPC: ads, seller, vendor and creative, from the same AI.

    Amazon Ads: 711 operations across 53 services, for Sponsored Products, Brands, Display and TV, DSP and Amazon Marketing Cloud. Selling Partner: 304 operations for orders, inventory, pricing, catalog and reports. Vendor Central: 40 operations for purchase orders, shipments, invoices and Retail Analytics. Creative: listing images and product video through Agent Iris. The operating context from Agent Atlas, and your data in a lake you own, all from the AI client you want to use. Perpetua's engine does the advertising slice of this. Kuudo does the whole job.

  • Your AI client

    Your AI client is already chosen, and it is not a vendor's.

    Kuudo connects to Claude, Claude Desktop, Claude Cowork, Claude Code, ChatGPT, Codex, Cursor, Gemini, n8n, Make, Activepieces and any other client that speaks Model Context Protocol. Where a vendor on this page ships an MCP, it points at the vendor's app and exposes the vendor's list; Kuudo's point at Amazon, with no Kuudo app between your agent and the API. Your team keeps the client it runs today, and the agent that handles advertising is the one that handles the rest of the Amazon job.

  • Your rules

    Your rules, your logic, your workflows: read the Skill before it runs, and write the next one.

    12 verified Skills state the job each one does, the inputs it needs, the outputs it returns and the limits it keeps, and you write your own logic, rules and workflows in the same format the structure audit and search-term Skills ship in. The Marketing Cloud Skill is the example: a plain-English question becomes validated SQL, privacy-safe results, and a segment ready to activate in DSP behind your approval. Agent Atlas supplies the Amazon operating knowledge the agent reads before it acts. A vendor's engine runs the vendor's logic; here the logic is yours to read and change.

  • Ownership

    The deployment, the data and the perimeter are yours. Private. Trusted. Owned.

    Kuudo runs in your own AWS, GCP, Azure or Cloudflare account, and Agent Flow lands your Ads and Selling Partner data in a data lake there for your warehouse and BI tools. Your cloud resources connect directly to Amazon, and substantive requests and responses do not route through Kuudo. Kuudo Pro is $0 a month plus $0.95 per 1,000 Amazon API calls, and Kuudo Scale is $249 a month plus $0.55 per 1,000 Amazon API calls: a price for the work, not a share of the budget.

  • Control

    How much you automate, and who does the work, is your call.

    Read tools return data only. Write tools validate the input, apply your workspace policy, and may require a person's approval before a change reaches Amazon, and every call is recorded with its parameters and approval state in a run log. An engine you set and leave is a fine working style; here it is a policy you wrote, not a default you inherited.

Staying put

Stay with Perpetua if

Where Perpetua is still the right answer. Switching costs real time, and these are the cases that do not repay it.

  • Your account runs to a target, and the engine hitting it is the job.

    You set a target advertising cost of sales and a daily budget per goal, and the ad engine recalculates bids daily, or hourly in the Stream-based conversion optimization beta, then shifts budget between that goal's campaigns. Template-based AMC audiences apply as bid multipliers on Sponsored Products and Sponsored Brands and carry through to DSP. Fixed bids, multipliers, minimum and maximum bids, placement multipliers and rules are yours to override with, and an optional Amazon status column shows what Amazon reported back about 15 minutes after a change. Kuudo runs the advertising job as agent tasks on a playbook, on request or on a schedule; it ships no standing bid engine.

  • The console's pending queue is how your team's day already runs.

    Rules and keyword harvesting can be routed to manual approval in a pending-approval tab, recommendations require an approve or dismiss, and an optional Amazon status column shows the status Amazon reports back about 15 minutes after a change, with a red dot on ads Amazon is not serving. Kuudo gates writes on a policy you write and records every call in a run log; if the tab is the habit, the habit is the reason to stay.

  • You add team members and accounts without adding cost.

    There is no limit on team members, and accounts can be added as long as combined monthly ad spend stays inside the plan's spend tier. Manager accounts add administrator and manager roles for agencies. If seat math governs your rollout, that structure is hard to improve on.

The differences that matter

5 questions that decide it.

Each one opens with the answer, then what each product does, then what it means for the person doing the work.

When my AI connects, does it reach Amazon or the vendor's app?

Pacvue and Teikametrics each publish an MCP, and each one is an MCP to that vendor's own platform, read-only today. Perpetua and Quartile publish none. Kuudo's MCP servers expose Amazon's own API operations from your cloud.

Perpetua

Perpetua is used through its browser app. Help-center searches return no public API, mobile app, or ChatGPT, Claude or MCP integration; the Premium plan lists custom data integrations without detail. Whatever your agent does, it does through Perpetua's screens.

Pacvue

An MCP to Pacvue: report, data query and share-of-voice services, read-only, with a 50,000-row cap per report, two years of history and a 180-day maximum per pull. A campaign-management MCP with human confirmation is described as in development. Your agent reaches Pacvue, and Pacvue reaches Amazon.

Teikametrics

An MCP to Teikametrics, in early access, giving Claude and ChatGPT twelve read-only tools over Teikametrics' own data, with no write or campaign-management actions. Your agent reaches Teikametrics, and Teikametrics reaches Amazon.

Quartile

No MCP is published. Customers reach Quartile's platform through a documented REST API with OAuth2 that Quartile must activate; no endpoint for setting bids or budgets directly was found.

Kuudo

Amazon. Kuudo's MCP servers expose Amazon's own API operations as tools, 711 of them for Ads across 53 services, run from your own cloud. There is no Kuudo app between your agent and Amazon; the tools are the operations Amazon publishes.

What it means for you. If your agent's job is to read what the vendor already collected, an MCP to that vendor covers it, inside their row caps and their read-only tool lists. If the job is to reach an operation Amazon publishes that the vendor did not expose, this is not a feature gap, it is a question of who holds the menu.

What decides a change: an algorithm, rules I set, or a playbook I wrote?

Perpetua, Teikametrics and Quartile decide with their own algorithm inside limits you set. Pacvue adds an agent whose recommendations your team approves. On Kuudo the playbook is a Skill you can read, or one you wrote.

Perpetua

The engine decides bids against your target; rules you configure for negatives, organic rank, competitor signals and inventory, and recommendations you approve or dismiss, add user-driven changes. A history dashboard logs when rule changes applied, but no page shows why the engine chose a given bid.

Pacvue

Three mechanisms are published: Pacvue's proprietary machine learning models optimizing toward defined targets, rules and thresholds the team defines, and Agent recommendations the team reviews and approves. The Agent explains the reason for each change and every action is logged.

Teikametrics

Teikametrics' own goal-based bidding algorithm decides bids inside the user's goal, cost limit and minimum and maximum; user-written logic is limited to keyword custom filters and dayparting schedules. The FAQ explains the general drivers of bid moves, without a per-change explanation.

Quartile

Quartile's proprietary, patented AI decides changes toward a target the customer sets; customers can add or substitute their own scheduled rules. No page says the customer can see why a specific change was made.

Kuudo

Your playbook: 12 verified Skills, each stating the job it does, the inputs it needs, the outputs it returns, and the limits it keeps, or Skills you write in the same format. Your AI runs them where you can read every step.

What it means for you. An algorithm you cannot read works while your account matches the pattern it was built for, and it stays behind the day you switch tools. A playbook you can read, change and take with you costs more attention up front and leaves your team owning the judgment.

What happens before anything changes on Amazon?

Each of the four applies at least some changes to Amazon with no step from you. Pacvue's agent actions, Perpetua's rules and harvesting, and Teikametrics' keyword filters can be routed through approval instead. On Kuudo, a write may require your explicit approval first.

Perpetua

Rules and keyword harvesting can be set to manual approval in a pending-approval tab or to apply automatically, and recommendations require approve or dismiss. The engine's own bid changes have no approval step.

Pacvue

Teams can apply approval workflows and guardrails and choose the level of automation. Pacvue Agent actions are approval-based in every case, while rule automations run continuously once enabled.

Teikametrics

There is no approval queue for bids: once AI bidding is enabled they post automatically, and manual bid edits are only available when automation is off. Keyword and negative filters default to review mode with accept, snooze and dismiss, and can be switched to automatic.

Quartile

No review queue, approval step or manual mode is described. The AI updates bids, budgets and placements in real time and customer rules run automatically on schedule.

Kuudo

Read tools return data only. Write tools are guarded and may require your explicit approval before a change reaches Amazon.

What it means for you. Decide which changes you want to see before they reach Amazon, then ask where the queue for them lives. At Perpetua, Pacvue and Teikametrics the answer differs by change type, so it is a set of per-feature settings rather than one switch, and at Quartile no queue is described at all.

Where does my account data end up?

All four hold your advertising data in their own platform, and Pacvue is the one that pushes daily feeds to a warehouse you own. Agent Flow lands your Ads and Selling Partner data in a data lake in your own cloud.

Perpetua

Advertising data lives in Perpetua's platform; custom reporting and dashboards export CSV, and no warehouse destination or public API is published. The terms say customers own their materials but Perpetua may freely use data to improve its services; the privacy policy gives no retention period or hosting location.

Pacvue

Data lives in Pacvue's platform; the DaaS product pushes daily feeds into Snowflake, BigQuery and other warehouses, Console exports reports, and MCP returns CSV or Excel with up to two years of history. No retention period is published; the terms allow Pacvue to use customer data in aggregated form.

Teikametrics

Advertising data lives in Teikametrics' platform; users can export tables and Compass data as CSV or Excel, but no warehouse destination or customer API is published. The terms say the company has no obligation to retain customer data after termination.

Quartile

Advertising data lands on Quartile's US servers and the terms let Quartile retain data derived from the services. Customers can pull campaign, keyword and report data through the REST API; no warehouse destination is offered. Click-tracking data is kept up to about 13 months.

Kuudo

Agent Flow lands your Ads and Selling Partner data in a data lake in your own cloud; your warehouse and BI tools read it downstream.

What it means for you. If the vendor's screens and a CSV export are where your team answers questions, this row changes little. It matters when a warehouse or a BI layer is the destination, because then you are choosing between a feed the vendor pushes, an API you pull, and a data lake in your own cloud.

How is it priced as my ad spend grows?

Perpetua and Teikametrics publish a base price plus a percentage of ad spend above $10,000 a month, Perpetua's percentage undisclosed. Pacvue and Quartile publish no prices at all. Kuudo prices per 1,000 Amazon API calls, with no percentage of ad spend.

Perpetua

Essentials is $695 a month for up to $10,000 of monthly ad spend; Growth is $695 a month plus an undisclosed percentage of ad spend above $10,000; Premium is custom above $500,000. Annual commitment is not published; the terms say contracts auto-renew per the order form.

Pacvue

No pricing page exists; the pricing and plans URLs return 404 and the demo, contact and platform pages carry no prices, plan tiers, percentage-of-spend terms, minimums or contract length. The only route is a demo request.

Teikametrics

Essentials is $179 a month, or $149 a month billed annually, for up to $10,000 of monthly ad spend; above that, plans add 3 percent of monthly ad spend over $10,000. The help center also lists Essential Tier 2 at $599, Advanced at $1,439, Enterprise as custom, and a free basic plan without AI.

Quartile

No prices are published and there is no pricing page. The terms say fees are set in an order form under standard pricing tiers based on aggregate monthly ad spend, number of channels and service level, due in advance and non-refundable, on a one-year auto-renewing term.

Kuudo

Kuudo Pro: $0 a month plus $0.95 per 1,000 Amazon API calls. Kuudo Scale: $249 a month plus $0.55 per 1,000 Amazon API calls. No percentage of ad spend, and the calculator models a month for a brand owner, an agency or aggregator, and an automation-heavy seller.

What it means for you. A percentage of ad spend ties the bill to how much media you buy, so it moves with the plan rather than with the work. Per-call pricing ties it to how much your AI actually does, which is worth modelling against the month you expect before comparing headline numbers.

Running both

Keep Perpetua on the bids, and let your AI read Amazon directly.

The two connect to Amazon separately. Kuudo's MCP servers run in your own cloud on your own Amazon credentials, so your AI can pull the reports from Amazon itself, run the Selling Partner and Vendor Central work Perpetua does not describe, and land the data in your own lake, while the ad engine goes on doing what you hired it for. One thing to settle first: Perpetua's engine recalculates bids on its own schedule and overrides bid edits made in the Amazon console, so decide which system owns bids before your AI writes one.

  1. Step 1

    Deploy the Amazon Ads and Selling Partner MCP servers in your own cloud and connect your Amazon accounts.

  2. Step 2

    Ask your AI for the campaign, search-term and placement reports from Amazon itself, and read the account back as Amazon has it, whatever moved it.

  3. Step 3

    Run the work the ad engine does not cover: Vendor Central purchase orders, shipments, invoices and Retail Analytics, plus Selling Partner orders, inventory, pricing and catalog.

  4. Step 4

    Land the Ads and Selling Partner data in a data lake in your own cloud with Agent Flow, and leave bid writes with Perpetua's engine until you decide otherwise.

Methodology

How we compared

Every fact on this page links to the page that states it, with the date it was checked. What we could not verify is listed, not guessed.

Competitor facts come from pages the competitor publishes. Kuudo facts come from Amazon Ads MCP tools reference, Amazon Agent Atlas documentation, Kuudo pricing calculator, Amazon Agent Flow documentation, Amazon Agent Iris documentation, Kuudo MCP servers reference, Kuudo pricing, Amazon Agent Skills reference, Amazon Selling Partner MCP tools reference. Planned work is cited to the roadmap and worded as planned. This page compares tools. For the job itself, see the amazon advertising software solution; for Kuudo's side of it, see Amazon Agent Skills and the Amazon Ads MCP.

Sources

  1. 1 Amazon Ads MCP tools reference checked
  2. 2 Amazon Agent Atlas documentation checked
  3. 3 Kuudo pricing calculator checked
  4. 4 Amazon Agent Flow documentation checked
  5. 5 Amazon Agent Iris documentation checked
  6. 6 Kuudo MCP servers reference checked
  7. 7 Kuudo pricing checked
  8. 8 Amazon Agent Skills reference checked
  9. 9 Amazon Selling Partner MCP tools reference checked
  10. 10 Pacvue MCP checked
  11. 11 Pacvue book a demo checked
  12. 12 Pacvue contact checked
  13. 13 Pacvue for Amazon checked
  14. 14 Pacvue real-time automation and optimization checked
  15. 15 Pacvue Agent checked
  16. 16 Pacvue Amazon Marketing Cloud checked
  17. 17 Pacvue unified commerce operations checked
  18. 18 Pacvue revenue recovery checked
  19. 19 Pacvue Amazon DSP checked
  20. 20 Pacvue Data-as-a-Service checked
  21. 21 Pacvue blog: changes to Amazon Sponsored Brands images requirement checked
  22. 22 Perpetua pricing checked
  23. 23 Perpetua help: DSP and AMC audiences checked
  24. 24 Perpetua help: sales and traffic for vendors checked
  25. 25 Perpetua help: rules for negative matches checked
  26. 26 Perpetua help: recommendations and strategies checked
  27. 27 Perpetua privacy policy checked
  28. 28 Perpetua terms of service checked
  29. 29 Perpetua help: team members checked
  30. 30 Perpetua help: account connection guide checked
  31. 31 Perpetua help: Sponsored Products goal structure checked
  32. 32 Perpetua Amazon PPC software page checked
  33. 33 Perpetua help: what the ad engine is doing checked
  34. 34 Perpetua help: monthly budget tracker checked
  35. 35 Perpetua help: campaign status columns checked
  36. 36 Perpetua for agencies checked
  37. 37 Perpetua help: rules on organic rank checked
  38. 38 Perpetua help: DSP Creatives Builder checked
  39. 39 Quartile API docs checked
  40. 40 Quartile for Amazon Sponsored Ads checked
  41. 41 Quartile homepage checked
  42. 42 Quartile Amazon promotion management checked
  43. 43 Quartile for Mid-Market and Enterprise checked
  44. 44 Quartile for SMB checked
  45. 45 Quartile privacy policy checked
  46. 46 Quartile terms and conditions checked
  47. 47 Teikametrics help: MCP overview checked
  48. 48 Teikametrics help: ad types and levels in Ads Manager checked
  49. 49 Teikametrics help: bidding algorithm FAQ checked
  50. 50 Teikametrics help: Amazon audiences checked
  51. 51 Teikametrics help: ARI catalog recommendations checked
  52. 52 Teikametrics help: Amazon 1P onboarding checked
  53. 53 Teikametrics help: keyword and targeting actions checked
  54. 54 Teikametrics help: intro to recommendations checked
  55. 55 Teikametrics terms and conditions checked
  56. 56 Teikametrics pricing checked
  57. 57 Teikametrics help: the viewer role checked
  58. 58 Teikametrics help: connecting your Amazon account checked
  59. 59 Teikametrics free trial signup checked
  60. 60 Teikametrics help: performance trends and metrics checked
  61. 61 Teikametrics help: products not in smart campaigns checked
  62. 62 Teikametrics help: keyword recommendations checked
  63. 63 Teikametrics help: decoding predictive AI bidding checked
  64. 64 Teikametrics help: budget pacing checked
  65. 65 Teikametrics help: intro to the bidder checked
  66. 66 Teikametrics help: Compass and AI value checked
  67. 67 Teikametrics GenAI Smart Pages early access checked

What we could not verify

  • Pacvue's image generation is documented in a Pacvue blog post that describes the GenAI tool as set to launch in Q3, and reports Image Generator results after the fact. Pacvue's current platform, AI and automation pages do not describe it, so the live state of the feature is stated by that post rather than by a product page.
  • Perpetua's marketing blog announced a Sponsored Brands video creator that builds a video for any ASIN from preset themes. The current help center article for launching a Sponsored Brands video goal documents uploading an MP4 or MOV to spec with no such option, and the Sponsored Brand Goals collection lists no video creator article, so that tool could not be confirmed as live today.
  • Teikametrics publishes AI-generated product images only on the GenAI Smart Pages early access landing page. The main www.teikametrics.com/genai-smart-pages/ URL describes Gen AI that enhances titles and descriptions and does not mention images, so image generation may be limited to the pilot.
  • quartile-2 (https://www.quartile.com/amazon-ppc) and scaleinsights-15 (https://scaleinsights.com/) were both re-verified today, 2026-09-08; their checkedOn values in comparePages.js still read 2026-09-02 and should be bumped if this row ships reusing them.
  • The exact percentage of ad spend Perpetua charges on Growth and Premium; its invoice article says to review the contract signed at onboarding.
  • Perpetua's annual commitment, minimum term or cancellation notice period; its terms defer these to the order form.
  • Whether Perpetua supports Sponsored TV; no site or help-center page mentions it.
  • Whether Perpetua explains why its engine chose a specific bid; the ad-engine and bid-adjustment articles do not say.
  • Perpetua's cloud provider, hosting region or data retention period; its privacy policy gives none.
  • A named Seller Central connection for Pacvue: its pages say Amazon APIs and its terms say your Amazon account; only Vendor Central is named.
  • Whether Pacvue manages Sponsored TV campaigns: its DSP page only says Sponsored TV is measured through AMC.
  • Whether Pacvue confirms a bid or budget change actually landed in Amazon, as opposed to logging the change and its impact.
  • Whether Pacvue's approval workflows are on by default for rule automations, given its FAQ says automations run without manual intervention.
  • Pacvue's data retention period and hosting provider: its terms name none.
  • Whether Pacvue's Revenue Recovery files disputes inside Vendor Central or prepares them for the user.
  • Whether Pacvue Agent's skills or its models' logic can be read or edited by the user; only rule templates and custom rules are described as editable.
  • Whether Quartile covers Sponsored Products, Brands, Display and TV individually; its pages say only Amazon Sponsored Ads.
  • Whether Quartile connects to Seller Central or Vendor Central; its privacy policy references an Amazon MWS API token and neither SP-API nor either console is named.
  • Quartile's search-term harvesting, negative keywords, budget pacing, approval step and per-change verification: none is published.
  • Quartile's actual prices, minimums or percentage-of-spend rate; its terms describe tiering by ad spend, channel count and service level only.
  • Teikametrics' search-term and placement report views: not found in Ads Manager or help searches; the only search-term mention tells users to download the report from Amazon.
  • Whether Teikametrics supports Sponsored TV; it is not listed among supported ad types.
  • Whether Teikametrics' predictive bidding applies to Amazon DSP; its DSP article documents manual bid edits only.
  • Teikametrics' data retention period and hosting region; neither its privacy policy nor its terms state them.
  • Which Teikametrics plan includes its MCP; the overview says a paid plan without naming a tier.

If you work at Perpetua, Pacvue, Teikametrics or Quartile and something here is wrong or out of date, tell us and we will fix it. Contact us.

Frequently asked questions

Questions sellers ask when they outgrow Perpetua.

What each alternative is actually for, what it costs, and where Kuudo fits.

What are the best alternatives to Perpetua?
Pacvue, Teikametrics, Quartile and Kuudo. Pacvue suits enterprise and agency teams running Amazon plus 100 or more other retailers, with approval workflows and change logs on every agent action. Teikametrics publishes its prices and a 30-day trial and suits smaller sellers who want an hourly bidder they can fence. Quartile pairs a dedicated human strategist with hourly automatic optimization across Amazon, Walmart, Instacart, Google, Meta and Microsoft. Kuudo is the one where your AI reaches Amazon's own API operations from your own cloud rather than a vendor's app.
Which Perpetua alternative publishes its pricing?
Teikametrics and Kuudo. Teikametrics lists Essentials at $179 a month, or $149 billed annually, for up to $10,000 of monthly ad spend, then 3 percent of monthly ad spend above that, with Essential Tier 2 at $599 and Advanced at $1,439. Kuudo lists Pro at $0 a month plus $0.95 per 1,000 Amazon API calls, and Scale at $249 a month plus $0.55 per 1,000. Pacvue and Quartile publish no prices: Pacvue's pricing and plans URLs return 404, and Quartile sets fees in an order form on a one-year auto-renewing term.
Can my AI assistant connect to any of these?
Pacvue and Teikametrics each publish an MCP, and each one is an MCP to that vendor's own platform. Pacvue's is read-only today, covering report, data query and share-of-voice services with a 50,000-row cap per report and 180 days per pull; a campaign-management MCP with human confirmation is described as in development. Teikametrics' is in early access with twelve read-only tools for Claude and ChatGPT. Perpetua publishes no API or MCP and is used through its browser app. Quartile publishes no MCP, only a REST API it must activate. Kuudo's MCP servers expose Amazon's own API operations, 711 of them for Ads across 53 services, from your own cloud.
Is Kuudo a replacement for Perpetua's ad engine?
No. Kuudo has no bidding algorithm. Your AI proposes bid and budget changes from live Amazon reports, and a write tool validates the input, applies your workspace policy, and may require your explicit approval before the change reaches Amazon. If what you want is a target you set and an engine that moves bids toward it on its own, Perpetua, Quartile and Teikametrics each publish one.
Can I run Kuudo alongside Perpetua?
Yes. Kuudo's Amazon MCP servers run in your own cloud on your own Amazon credentials, so your AI can pull reports straight from Amazon and run Selling Partner and Vendor Central work while Perpetua's engine manages bids. One caution: Perpetua's engine recalculates bids on its own schedule and overrides bid edits made in the Amazon console, so decide which system owns bids before your AI writes one.
Which of these paces budgets to a monthly plan?
None of the four adjusts spend on its own to land a monthly number. Teikametrics comes closest as reporting: a monthly budget goal per merchant with month-to-date spend, a daily breakdown, a full-month forecast and a daily Slack or email snapshot. Perpetua's monthly budget tracker shows projected spend against a target without pacing or capping it. Pacvue manages pacing, flighting and out-of-budget controls, though pacing to a monthly plan is not stated in those words. Quartile publishes nothing on it. Kuudo's pacing Skill returns a plan-versus-actual report, a rest-of-month forecast and a campaign-level budget change set that remains a proposal until approved.
Which alternatives cover Vendor Central?
Pacvue's Revenue Recovery audits up to two years of Vendor Central data for invoice shortages, chargebacks, price claims and co-op overbilling, raises purchase order and invoice mismatch alerts, and identifies and files disputes. Teikametrics grants a service user Vendor Central report permissions so total sales and inventory flow in, with inventory refreshing weekly and sales lagging up to 72 hours. Quartile publishes nothing on Vendor Central. Perpetua shows Amazon Retail Analytics sales and Rapid Retail Analytics inventory beside advertising data. Kuudo's Amazon Vendor Central MCP carries 40 operations: purchase orders, direct fulfillment, shipments and invoices, and Retail Analytics.

The bottom line

Point your AI at Amazon, not at another app.

Your AI runs Amazon Ads, AMC, seller and vendor work through Kuudo's MCP servers in your own cloud, on Skills you can read or write yourself, with reads that change nothing and writes guarded by the approval policy you set. Kuudo Pro starts at $0 a month.