The AI Tool Stack for Product Managers in 2027

Whiteboard wall with user-journey sketches behind a laptop showing an app prototype under cool blue light

The AI tool stack for product managers in 2027: customer evidence, specs, prototypes, delivery, analytics and decision records, with verified prices per tool.

AI / AI tool stacks

The AI tool stack product managers use in 2027 runs from Dovetail and Productboard for customer evidence, through ChatPRD and Claude for specs, to Figma Make, v0 or Lovable for prototypes, Linear for planning and delivery, Amplitude for adoption data, and Granola and Notion for meetings and decisions, with Perplexity for competitive research. We grouped 12 tools by the six steps a feature goes through, from the first customer signal to the decision record after launch, and checked every price on the vendor's pricing page on 19 November 2026. The biggest shift for PMs is the prototype step: a working, clickable version of an idea now costs an afternoon instead of a design sprint, which changes how teams decide what to build.

Key takeaways

  1. Customer evidence platforms now use agents to classify feedback and answer questions across interviews, tickets and surveys. Dovetail and Productboard both lead with that capability.
  2. Specs are faster to write and easier to get wrong. ChatPRD and Claude draft a PRD in minutes, so the PM's job shifts to judgment: which problem, for whom, and what success looks like.
  3. Prototyping tools change product reviews. A PM can bring a working Figma Make, v0 or Lovable prototype to a review and test it with five customers before engineering commits.
  4. Analytics is close to free at the start. Amplitude's plans start at $0 and early-stage startups can get a free year of its Growth plan.

Which AI tools should a product manager use at each step?

Ranked by how much a practitioner or technology leader gets out of each one. "Checked" names the source behind each number and when we looked.

01 Gather customer evidence

  1. Bronze $4,800/yr for the platform and 5 users, plus credits at 15 cents each; Silver and Gold on request · web, API · self-serve (card) and sales-led

    For: Product, research and design teams that want interviews, tickets, NPS and reviews analyzed in one place.

    Dovetail pulls customer feedback from many channels, classifies it into themes, and lets a PM ask questions across all of it in chat. Its newest feature builds digital twins of customer segments from real data, so a PM can pressure-test an idea before scheduling interviews. Treat twin answers as hypotheses to verify with real customers, not as evidence.

    Checked: dovetail.com/pricing and product menu (Bronze, credits, Digital Twins), 19 Nov 2026

  2. Business and Enterprise plans on request; AI credit top-ups $5 per 50 (monthly) or $60/yr per 600 (annual) · web · sales-led

    For: Product organizations that want feedback, prioritization, roadmaps and specs connected in one system.

    Productboard's Spark agent works across customer feedback, product context and roadmaps, drafting specs and surfacing opportunities with knowledge of your product. The Business plan includes unlimited feedback and roadmaps, product usage integrations with Amplitude and Mixpanel, and customer portals. It suits teams with several PMs who need a shared prioritization system.

    Checked: productboard.com/pricing (Business and Enterprise inclusions, credit top-ups FAQ), 19 Nov 2026

  3. Free; Pro $17/mo billed yearly; Max $167/mo billed yearly · web, mobile, Comet browser · self-serve

    For: PMs tracking competitors, pricing changes and market context for a roadmap decision.

    Perplexity is the quickest way to get a cited summary of what competitors shipped, how they price and what reviewers say, which is the market half of the evidence a PM brings to planning. Pro adds deep research and premium data sources. Verify competitor claims on the competitor's own pages before they reach a strategy doc.

    Checked: perplexity.ai pricing page (Free, Pro, Max), 19 Nov 2026

02 Write the spec

  1. Free (3 chats); Pro $15/mo billed yearly ($179/yr); Teams $29 per seat/mo billed yearly ($349/yr) · web · self-serve

    For: PMs who want a purpose-built assistant for PRDs, user stories and product strategy documents.

    ChatPRD is built only for product documents: it drafts PRDs, user stories and strategy documents from a short brief, and can hand documents straight to v0, Cursor, Lovable or Replit for prototyping. That handoff is what sets it apart from a general assistant. The Pro plan is priced below most general-purpose AI subscriptions.

    Checked: chatprd.ai/pricing (Free, Pro, Teams, FAQ), 19 Nov 2026

  2. 5

    Claude

    Free; Pro $20/mo ($17 yearly); Team $25 per seat/mo ($20 yearly) · web, desktop app, mobile · self-serve

    For: PMs who write specs, strategy memos and stakeholder updates and also want help analyzing data and research notes.

    Claude handles the full range of PM writing, from a one-page problem statement to a launch FAQ, and its Projects keep product context and past decisions available across sessions. It can also read exported interview notes or a CSV of feature usage and summarize what changed. PMs who already pay for Claude may not need ChatPRD; those who write PRDs daily often prefer the specialist.

    Checked: claude.com/pricing (Individual and Team plans), 19 Nov 2026

03 Prototype before engineering commits

  1. Included with Figma plans; Professional Full seat $16/mo with 3,000 AI credits; free Starter with limited AI credits · web, desktop · self-serve

    For: PMs at companies where design already lives in Figma and prototypes should match the design system.

    Figma Make turns a prompt or an existing design into working code inside Figma, so a prototype uses the team's real components and stays next to the design files. That keeps designers involved instead of routing around them. Our comparison of Claude Design, Figma Make and Google Stitch covers how it stacks up against newer design agents.

    Checked: figma.com/pricing (Starter, Professional, Organization, Enterprise seats and AI credits), 19 Nov 2026

  2. 7

    v0

    Free (7 messages a day); Plus from $30 per user/mo with $30 in monthly credits; Business $100 per user/mo · web · self-serve

    For: PMs at web product companies who want a React prototype that engineers can actually build from.

    v0 generates web apps from prompts, deploys them to Vercel and syncs with GitHub, which makes it easy for engineering to pick up a prototype as a starting point. A visual Design Mode lets a PM adjust layouts without prompting for every change. Usage beyond included credits is billed per token by model.

    Checked: v0.app/pricing (Free, Plus, Business, model token rates), 19 Nov 2026

  3. Free; Pro $25/mo (100 monthly credits); Business $50/mo · web · self-serve

    For: PMs who want a full working app, including data and hosting, to put in front of customers for a week.

    Lovable builds and hosts a complete app from a description, which suits customer tests that need real data entry or accounts, not just clickable screens. Unlimited collaborators on every plan make it easy to bring in a designer or an engineer. Our Lovable review and our Lovable versus v0 comparison cover the trade-offs.

    Checked: lovable.dev/pricing (Free, Pro, Business), 19 Nov 2026

04 Plan and ship

  1. 9

    Linear

    Free (2 teams, 250 issues); Basic $10 per user/mo yearly; Business $16 per user/mo yearly; Enterprise custom · web, desktop · self-serve

    For: Product and engineering teams that want fast issue tracking, cycles and roadmaps with agents built in.

    Linear includes its own Linear Agent, an agent platform and MCP access on every plan, so coding agents and the PM's own assistant can create, triage and update issues. For PMs the useful part is less admin: agents draft issues from a spec and keep project updates current. The free plan is enough for a small team to trial it.

    Checked: linear.app/pricing (Free, Basic, Business, Enterprise; agent features table), 19 Nov 2026

05 Measure adoption

  1. Free plan; Plus starts at $0 with the first 2M events a month free, then usage-based; Growth and Enterprise on request; startups under $10M funding and 20 staff get a free Growth year · web · self-serve

    For: PMs who need product analytics, session replay and experiments without waiting on a data team.

    Amplitude includes its AI analytics features, session replay and unlimited seats on every plan, and the Plus tier adds AI agents and an MCP server so a PM can ask usage questions from an assistant. The startup program removes the cost barrier for early teams. Event tracking design still needs care up front; AI cannot fix events that were never instrumented.

    Checked: amplitude.com/pricing (Free, Plus, Growth, startup program), 19 Nov 2026

06 Record meetings and decisions

  1. 11

    Granola

    Basic free (limited history); Business $14 per user/mo; Enterprise $35 per user/mo · desktop, iOS, Android, Apple Watch · self-serve

    For: PMs running customer interviews, stakeholder reviews and planning meetings all day.

    Granola takes bot-free notes from device audio and merges them with the PM's own jottings, which keeps customer interviews conversational. Business adds unlimited history and MCP access, so a PM can ask an assistant what customers said about a feature across a month of calls. Export the key quotes into Dovetail or Productboard so evidence does not stay locked in meeting notes.

    Checked: granola.ai/pricing (Basic, Business, Enterprise), 19 Nov 2026

  2. 12

    Notion

    Free; Plus $10 per member/mo; Business $20 per member/mo with Notion Agent and AI Meeting Notes · web, desktop, mobile · self-serve

    For: Product teams that keep specs, decision logs and launch plans in a shared wiki.

    Notion is where many product teams keep the decision record: why a feature was built, what was traded off and what success meant. The Business plan adds Notion Agent for multi-step tasks and search across connected apps, which makes old decisions findable. Free and Plus include only an AI trial.

    Checked: notion.com/pricing (Free, Plus, Business, AI features), 19 Nov 2026

How did we choose the product management tools?

We followed one feature from the first customer complaint to the post-launch review and asked where a PM's week goes: collecting and making sense of evidence, writing the spec, testing the idea, planning delivery, measuring adoption and recording decisions. For each step we picked the tools we would ask a PM team to trial first, with an alternative where the right choice depends on company size or design culture.

All prices were read on vendor pricing pages on 19 November 2026. Productboard sells its Business plan through demos, and v0's Plus plan shows a price range by credit volume, so we list the starting figure. Enterprise feedback analytics vendors whose pricing pages we could not open are left out.

How do AI prototypes change product reviews?

They move the debate from slides to software. When a PM brings a working Figma Make, v0 or Lovable prototype to a review, stakeholders react to the experience rather than to a description of it, and weak ideas fail faster. The same prototype can go in front of five customers the next day, which turns an opinion contest into a test.

Prototypes also change the PM and engineering relationship. Engineers can review a working prototype for feasibility in minutes, spot the hard parts early and suggest a simpler version before any sprint planning. Teams that invite an engineer to the prototype review, rather than handing over a finished spec, tend to cut scope earlier and ship sooner.

The risk is mistaking a prototype for a plan. Generated prototypes skip error states, permissions, performance and data migration, which is where most engineering effort goes. A good practice is to label every prototype with what it deliberately ignores, and to keep the spec as the source of truth for scope.

What does the stack cost for one PM?

An individual PM can cover research, specs, prototyping and meeting notes for about $75 a month: Perplexity Pro at $17, Claude Pro at $20 or ChatPRD Pro at $15, Lovable Pro at $25 and Granola Business at $14. Linear, Amplitude and Notion are usually company purchases, and all three have free tiers for small teams.

For a product organization of ten PMs, the line items that scale fastest are per-seat assistants and prototyping credits. Agree on one general assistant for the whole team so shared projects and templates work, and give prototyping credits to the PMs who are actively testing ideas that quarter rather than to everyone by default.

Customer evidence platforms are the expensive step. Dovetail's entry plan is $4,800 a year for five users before analysis credits, and Productboard is quoted per customer. Teams below that budget can get far with Granola transcripts, a shared Notion database of tagged quotes and Claude to summarize themes each month.

What should product managers not hand to AI?

The choice of problem and the definition of success. AI tools are good at summarizing what customers said and drafting a spec for a chosen solution, and poor at weighing strategy, timing and the cost of saying no to something else. A PM who outsources those calls to a model ends up shipping plausible features nobody prioritized on purpose.

A useful test is whether the PM can explain a decision without reference to the tool that suggested it. If the rationale in the spec reads as a summary of AI output rather than a judgment about customers, strategy and trade-offs, the work is not finished yet.

Customer trust is the other boundary. Recording interviews, building digital twins of segments and piping feedback into AI systems all require clear consent and data handling. Tell customers how their words will be used, keep sensitive accounts out of shared AI workspaces unless contracts allow it, and check vendor data retention settings before connecting a CRM or support inbox.

Listed here, or should be?

If an entry is out of date or something belongs on this list, email [email protected]. Listings are editorial and free. A featured, labeled Sponsored slot is available on request.

Related questions

What AI tools do product managers use most in 2027?

A general assistant (Claude or ChatGPT), a spec tool such as ChatPRD, a prototyping tool (Figma Make, v0 or Lovable), an issue tracker with agents such as Linear, an analytics platform such as Amplitude, and a meeting notetaker such as Granola.

Is ChatPRD worth it if I already pay for Claude or ChatGPT?

For PMs who write PRDs every week, often yes: it is built around product documents and hands specs to prototyping tools. For occasional spec writing, a general assistant with a good PRD template is enough.

Should PMs use Figma Make, v0 or Lovable for prototypes?

Figma Make fits teams whose design system lives in Figma, v0 fits web teams that want React code engineers can reuse, and Lovable fits PMs who need a full working app with data and hosting for customer tests.

What is the cheapest way to analyze customer feedback with AI?

Record interviews with Granola, tag quotes in a shared Notion database and ask Claude to summarize themes monthly. Move to Dovetail or Productboard once several PMs need shared evidence and reporting.

Explore More

Ready to Find the Right AI Tools?

Browse our data-driven rankings to find the best AI tools for your team.