ProgrammableMarketing/fal.ai (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 6
  • This quarter · 10
  • Big bets · 3
  • Wildcards · 2
  • How we can help
Generated Oct 23, 2026
  1. Programmable
  2. Playbooks
  3. fal
fal logo

growth playbook · Oct 23, 2026

fal.ai (opens in a new tab)

fal growth playbook

Diagnosis

fal is a leading generative media platform with 2.5 million developers and 1,500+ models, and its hiring shows a move upmarket and into raw compute: a hunter AE for Compute, product marketers and a technical content manager for Serverless and Compute, enterprise field marketing and ABM, inbound SDRs and BDRs, community, events and creator partnerships.

The site has not caught up with Compute: there is no product page, the links land on docs that list fewer GPUs than the homepage promises, and custom deals are sent to a support email. Proof is thin for its size (four customer stories, one industry page), headline numbers disagree, the events page is stale and the homepage is slow for real users.

This playbook builds the Compute page and quote flow first, fixes routing, numbers, speed and events, then adds infrastructure content, comparisons, benchmarks, creator programs, usage-based routing and a train-then-serve path.

Growth Score

7/10

Biggest Opportunity

Turn fal's developer reach into infrastructure revenue: a proper Compute front door, usage-based routing to the new AEs, and content that wins people searching for GPUs and inference.

Biggest Risk

GPU clouds (Lambda, Together AI) and multi-model platforms (Replicate, Runware, Modal) all sell both model access and capacity; if fal's Compute offer stays hidden in docs, buyers will compare it on rivals' pages.

01

Start here

The three moves to make first. Each one is written up in full below.

  1. 1

    Give Compute a real product page with a cluster quote form, so the new Compute AE has somewhere to send buyers

    Agree with the Compute team which GPU types and cluster sizes can be promised on the page.

    High impact First result in 1 month
  2. 2

    Make the events page show what is next, not September

    Swap the featured event for the next one.

    Medium impact First result in 2 weeks
  3. 3

    Make every headline number match: models, developers and the Poe share

    Rewrite the homepage meta description.

    Medium impact First result in 1 week

02

Your idea portfolio

21 ideas, grouped by when to run them. Open any idea for how to run it: the steps, who to approach, copy you can use, and how to measure it.

  • 7 ideas are marked Unexpected: plays you are unlikely to have tried.

Every idea placed by the effort it takes against the impact it could have. Quick wins sit top left; select a number to open that idea. The filters below narrow the list by status, timing, category and what Programmable can build.

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 1Big bets · 13Fill-ins · 5Reconsider · 2
Effort →

Showing 21 of 21 ideas

This week 6

Quick wins with a clear first action.

  • 01

    fal is hiring a hunter AE for Compute because customers are "coming to us for raw compute", plus an infrastructure product marketer and a content manager for "landing pages" and "solution pages". But Compute has no product page: every Compute link on the homepage, and fal.ai/compute itself, lands on a documentation introduction. That page lists only 1x and 8x H100 instances, while the homepage promises "interconnected clusters of H100, H200, and B200 GPUs" across "hundreds or thousands of GPUs". The pricing page puts Compute in one table with Serverless and sends custom deals to [email protected]. A neolab founder comparing GPU clouds finds specs and an email address, not a way to ask for a cluster.

    High impactMedium effort First result in 1 month Product Marketer

    First action

    Agree with the Compute team which GPU types and cluster sizes can be promised on the page.

    How to measure it

    Qualified Compute quote requests

    Now
    –
    Target
    40 a month
    By
    90 days

    Steps

    1. 1

      Agree the offer

      GPU types, cluster sizes, terms that can be sold today.

      Product Marketer, Infrastructure 3 days

    2. 2

      Write the page

      Specs, terms, use cases, serving path and proof.

      Technical Content Marketing Manager, Infrastructure 1 week

    3. 3

      Design and build

      Product page in the Serverless style.

      Senior Product Designer 1 week

    4. 4

      Build the quote form

      GPU type, count, start date, term; route to the Compute AE.

      GTM Operations Engineer 2 days

    5. 5

      Repoint links

      Homepage, pricing and nav to /compute; docs stay for setup.

      Senior Product Designer 1 day

    6. 6

      Add proof

      One training customer story or quote.

      Technical Content Marketing Manager, Infrastructure 3 weeks

    7. 7

      Measure

      Quotes, meetings and pipeline from the page.

      GTM Operations Engineer weekly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Product Marketer, Infrastructure

    Smallest test

    Replace the support@ email on the pricing page with a Compute quote form this week.

    Why it works

    Build fal.ai/compute as a product page: the GPU generations and cluster sizes actually offered, on-demand vs reserved terms, interconnect and storage, what training and fine-tuning teams run on it, how Compute connects to Serverless for serving the trained model, and one customer proof point. End it with a short quote form (GPU type, count, start date, term) that goes straight to the Compute AE, not support. Point the homepage Compute links and the pricing page at it, and keep the docs page for setup.

    Buyers of reserved GPU capacity compare providers on a single page and then ask for a quote. A page that says what is available and takes the request turns the traffic fal already has into Compute pipeline for the new AE, and gives the infrastructure team the landing page its SEO and content plans need.

    Ready to use

    ChecklistCompute page outline

    [ ] Headline for training and fine-tuning teams
    [ ] GPU generations and cluster sizes offered now
    [ ] On-demand vs reserved terms
    [ ] Interconnect, storage and regions
    [ ] Train on Compute, serve on Serverless
    [ ] One customer proof point
    [ ] Quote form: GPU, count, start date, term
    [ ] Link to docs for setup

    Risks and how to handle them

    • Promising capacity that is not available

      List only configurations the Compute team signs off, with "contact us" for larger clusters.

    • Cannibalising Serverless

      Explain clearly which workloads fit each product.

    Evidence

    • HiringAI Native Account Executive, Compute (posted October 19, 2026): "More and more of them are coming to us for raw compute as well as inference and model APIs." The AE sells "dedicated GPU clusters, reserved and on-demand capacity, and the inference platform", owns "all net-new logo acquisition across a defined set of AI-native startups, neolabs and research-driven model builders", and builds pipeline "through founder and investor networks, targeted outbound, community and events". "This is a pure hunter role. You won't inherit a book."Source
    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
    • Your websiteCompute documentation: "Dedicated GPU instances for training, fine-tuning"; "Compute runs on NVIDIA H100 SXM GPUs"; two instance types, 1xH100-SXM and 8xH100-SXM, with InfiniBand between 8-GPU instances in the same sector; "billed per-hour at fixed rates". The page is a technical introduction with no contact or quote form.Source
    • Your websitePricing page: a single "Serverless & Compute Pricing" table with list and "as low as" hourly prices (B300 $12.99/h list, as low as $5.99/h; B200 $7.99/h, as low as $5.49/h; H200 $6.00/h, as low as $2.99/h; H100 $4.50/h, as low as $2.49/h); for custom deployments it says "Contact [email protected] to get started". There is no separate section for reserved clusters, multi-node training or commitment terms.Source
    • HiringProduct Marketer, Infrastructure (posted October 5, 2026): owns GTM for "Serverless for deploying and scaling custom AI models across thousands of GPUs, to Compute for training on large, interconnected GPU clusters"; must "Develop our competitive positioning across inference platforms, GPU clouds, hyperscalers, and in-house infrastructure", "Own SEO and AEO strategy for infrastructure, capturing developers actively searching for inference, training, and compute solutions" and create "benchmarks, product launches, solution pages" and "case studies".Source
    • HiringTechnical Content Marketing Manager, Infrastructure (posted October 5, 2026): "Some of the most interesting work happening at fal is deep inside the infrastructure, and most of the market never sees it." Writes "technical blogs, landing pages, guides, tutorials, solution pages" for Serverless and Compute, "benchmarks and competitive content", and "technical case studies".Source
  • 02

    fal is hiring an Event Marketing Manager and a field marketer, and runs meetups, webinars and hackathons. On October 22 the events page still features the GenMedia Conference of September 24, and "All upcoming events" starts with September dates.

    Medium impactLow effort First result in 2 weeks Event Marketing Manager

    First action

    Swap the featured event for the next one.

    How to measure it

    Registrations from /events

    Now
    –
    Target
    +30%
    By
    60 days

    Steps

    1. 1

      Swap the feature

      Next event first.

      Event Marketing Manager 1 hour

    2. 2

      Auto-archive

      Past events move by date.

      Senior Product Designer 2 days

    3. 3

      Add recordings

      Past talks.

      Event Marketing Manager 1 week

    4. 4

      Add subscribe

      Calendar and email.

      GTM Operations Engineer 2 days

    5. 5

      Promote

      Discord and X.

      Community Growth weekly

    6. 6

      Measure

      Registrations.

      Event Marketing Manager monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Event Marketing Manager

    Smallest test

    Change the featured event today.

    Why it works

    Feature the next event automatically, move past events into an archive with recordings, and add a newsletter or calendar subscription for future meetups.

    An up-to-date page turns visits into registrations; a stale one suggests the community is quiet.

    Risks and how to handle them

    • Few events in a month

      Show webinars and community calls too.

    Evidence

    • Your websiteEvents page (checked October 22, 2026): the featured event is still "GenMedia Conference 2026 Registration, September 24, 2026", and "All upcoming events" opens with September 15, 19 and 25 dates before the later October 31, November and March 2027 events.Source
    • HiringEvents and field: an Event Marketing Manager (posted October 6, 2026) for "meetups, workshops, hackathons, dinners", webinars and "ABM activations"; a Field Marketing Manager, Enterprise & ABM (posted October 6, 2026) to "turn field marketing into a meaningful driver of pipeline and revenue" with "executive dinners and roundtables, customer gatherings, account-specific experiences".Source
    • HiringCommunity Growth (posted September 9, 2026): "fal has a large community of developers, creative technologists, hackers, and creators"; the role owns "events, hackathons, creator competitions, webinars, Discord, Reddit, and developer content like demos, prompt guides, and walkthroughs" and a "network of power users... who test new fal tools, provide feedback, and build things we can showcase".Source
  • 03

    Search results show fal's old description ("1,000+ AI models", "1.5M devs") while the homepage says 1,500+ models and 2.5 million developers. The homepage says fal powers 40% of Poe's bots and the Enterprise Hub says 50%. The Grants page stat labels are mismatched.

    Medium impactLow effort First result in 1 week Product Marketer

    First action

    Rewrite the homepage meta description.

    How to measure it

    Click-through

    Now
    –
    Target
    +8%
    By
    60 days

    Steps

    1. 1

      List claims

      Models, developers, uptime, customers.

      Product Marketer, Developer Platform 1 day

    2. 2

      Fix meta

      Homepage and top pages.

      Product Marketer, Developer Platform 1 hour

    3. 3

      Check Poe figure

      With Quora.

      Product Marketer, Developer Platform 2 days

    4. 4

      Fix Grants

      Stat labels.

      Senior Product Designer 1 hour

    5. 5

      Audit

      Top 30 pages.

      Technical Content Marketing Manager, Infrastructure 2 days

    6. 6

      Review

      Monthly with each new number.

      Product Marketer, Developer Platform monthly

    First result
    1 week
    Cost
    $0
    Owner
    Product Marketer, Developer Platform

    Smallest test

    Update the homepage description today.

    Why it works

    Keep one list of approved claims, update the meta descriptions, align the Poe figure with Quora, fix the Grants stats, and check the top 30 pages for older numbers.

    Consistent numbers build trust with enterprise buyers and keep AI assistants from quoting the old ones.

    Risks and how to handle them

    • Numbers change fast

      Pull them from one shared source.

    Evidence

    • Your websiteInconsistent claims: the homepage meta description (shown in search results) still says "Access 1,000+ AI models" and "Trusted by 1.5M devs" while the page says 1,500+ models and 2,500,000 developers; the Enterprise Hub says "fal powers 50% of Poe's image and video generation" while the homepage says 40%; the Grants page stat block pairs "$2m" with "faster response time compared to other providers" and "12" with "companies incubated".Source
    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
    • Your websiteEnterprise Hub: "Custom training for Fortune 500 companies"; four customer highlights (Quora/Poe, Genspark, Pimento, Layer), which are also the only four /customer-case pages; the only industry page is /industry/advertising. Logos and quotes elsewhere include Canva, Adobe, Shopify, Suno, World Labs, Krea and HeyGen.Source
  • 04

    Real Chrome users see the homepage's main content after 3.15 seconds at the 75th percentile, above Google's 2.5 second target and 24% slower than a year ago; the mobile lab test takes 10.4 seconds. The site sells "10x faster inference".

    High impactMedium effort First result in 1 month Senior Product Designer

    First action

    Identify the largest element on a mobile run.

    How to measure it

    LCP p75

    Now
    –
    Target
    under 2.5 s
    By
    60 days

    Steps

    1. 1

      Profile

      Mobile run.

      Senior Product Designer 1 day

    2. 2

      Preload hero

      Main image or text.

      Senior Product Designer 1 day

    3. 3

      Defer media

      Gallery and videos.

      Senior Product Designer 3 days

    4. 4

      Trim scripts

      Remove unused tags.

      GTM Operations Engineer 2 days

    5. 5

      Budget

      Limit for launch sections.

      Senior Product Designer 1 day

    6. 6

      Track

      Chrome UX Report monthly.

      Senior Product Designer monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Senior Product Designer

    Smallest test

    Defer the model gallery and re-test.

    Why it works

    Find the largest element on mobile, preload it, defer the animated model gallery and below-the-fold media, and set a performance budget for new launch sections.

    A faster first view lifts signups and fits the speed promise.

    Risks and how to handle them

    • Launch pages add weight again

      Check the budget at each launch.

    Evidence

    • Google PageSpeedfal.ai homepage: mobile lab performance 59 with Largest Contentful Paint 10.4 s (desktop 85). Real Chrome users see LCP p75 of 3.15 s, above the 2.5 s "good" threshold and 24% slower than a year ago; INP 147 ms and CLS 0.03 pass.Source
    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
  • 05

    fal is hiring inbound SDRs to qualify "website inquiries", yet the pricing page sends custom deployments to [email protected] and the Grants page to an email address. Emails to support are not qualified, scored or measured.

    High impactLow effort First result in 2 weeks Sales Development Representat…

    First action

    Count last month's sales questions in the support inbox.

    How to measure it

    Qualified inbound leads

    Now
    –
    Target
    +25%
    By
    60 days

    Steps

    1. 1

      Count

      Sales emails in support.

      Sales Development Representative (Inbound) 1 day

    2. 2

      Build forms

      Pricing, Grants, Enterprise.

      GTM Operations Engineer 2 days

    3. 3

      Enrich

      Clay rules.

      GTM Operations Engineer 2 days

    4. 4

      Route

      One-hour SLA.

      GTM Operations Engineer 1 day

    5. 5

      Script

      First reply.

      Sales Development Representative (Inbound) 1 day

    6. 6

      Measure

      Leads to meetings.

      GTM Operations Engineer weekly

    First result
    2 weeks
    Cost
    $0
    Owner
    Sales Development Representative (Inbound)

    Smallest test

    Swap the pricing page email for a form.

    Why it works

    Replace those addresses with short forms (workload, volume, timeline) that create a lead in Salesforce, enrich it in Clay and assign it to an SDR within the hour.

    Every buyer who asks about custom deployments is a sales lead; capturing them in the CRM makes the new SDR team measurable.

    Risks and how to handle them

    • Developers want email

      Keep the address as a fallback.

    Builds on

    Evidence

    • Your websitePricing page: a single "Serverless & Compute Pricing" table with list and "as low as" hourly prices (B300 $12.99/h list, as low as $5.99/h; B200 $7.99/h, as low as $5.49/h; H200 $6.00/h, as low as $2.99/h; H100 $4.50/h, as low as $2.49/h); for custom deployments it says "Contact [email protected] to get started". There is no separate section for reserved clusters, multi-node training or commitment terms.Source
    • HiringSales build-out: a Sales Development Representative (Inbound) to "Qualify inbound leads from marketing campaigns and website inquiries" (posted September 29, 2026), a BDR, an Account Executive, Commercial for "creative studios, AI innovation teams, and AI developers" (October 21), an Account Executive, Enterprise (October 22), and a GTM Operations Engineer for "a GTM org of 30" building target lists and enrichment in Clay and Salesforce.Source
    • Your websiteStartup Program: credit tiers (for example "Growth $5,000", which "Unlocks at $2,500/mo on fal"); for teams that are VC-backed or have traction; "Partner-referred applications are prioritized"; "Partner perks... Coming soon". The Research Grants program offers a "$50k average grant check" of free compute for open-source projects.Source
  • 06

    fal shows logos and quotes from Canva, Adobe, Shopify, Suno, World Labs, Krea and HeyGen, but has only four customer stories (Poe, Genspark, Pimento, Layer). The new marketers are asked for "customer and developer stories" and "technical case studies".

    High impactMedium effort First result in 1 quarter Technical Content Marketing M…

    First action

    Ask Canva for a story built around the homepage quote.

    How to measure it

    New stories

    Now
    –
    Target
    5
    By
    1 quarter

    Steps

    1. 1

      Pick five

      Largest logos.

      Technical Content Marketing Manager, Infrastructure 1 day

    2. 2

      Ask

      Through the account team.

      Account Executive, Enterprise 2 weeks

    3. 3

      Interview

      Engineering leads.

      Technical Content Marketing Manager, Infrastructure 1 month

    4. 4

      Write

      Workload, scale, result.

      Technical Content Marketing Manager, Infrastructure 1 month

    5. 5

      Publish

      Enterprise Hub and /customer-case.

      Senior Product Designer 1 week

    6. 6

      Use

      In ABM and sales decks.

      Field Marketing Manager, Enterprise & ABM ongoing

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Technical Content Marketing Manager, Infrastructure

    Smallest test

    Publish one short Canva story.

    Why it works

    Ask the five largest logo customers for a short story each: workload, models, scale, latency or cost result. Start with Canva, whose team is already quoted on the homepage.

    Enterprise buyers trust peers at their scale more than a logo bar.

    Risks and how to handle them

    • Approvals take time

      Start with quotes and short stories.

    Evidence

    • Your websiteEnterprise Hub: "Custom training for Fortune 500 companies"; four customer highlights (Quora/Poe, Genspark, Pimento, Layer), which are also the only four /customer-case pages; the only industry page is /industry/advertising. Logos and quotes elsewhere include Canva, Adobe, Shopify, Suno, World Labs, Krea and HeyGen.Source
    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
    • HiringProduct Marketer, Developer Platform (posted October 5, 2026): owns GTM for "Model APIs for accessing 1,500+ generative media models", "fal Agent" and "fal CLI"; "The winning developer platform won't just have the best models. It'll make them easy to discover"; will "Develop customer and developer stories showcasing what people are building on fal" and "Work with Growth on SEO/AEO, lifecycle, and product-led growth to turn discovery into active users".Source
    • HiringTechnical Content Marketing Manager, Infrastructure (posted October 5, 2026): "Some of the most interesting work happening at fal is deep inside the infrastructure, and most of the market never sees it." Writes "technical blogs, landing pages, guides, tutorials, solution pages" for Serverless and Compute, "benchmarks and competitive content", and "technical case studies".Source

This quarter 10

The core plays for the next three months.

  • 07

    fal has 227 learn articles, nearly all about models; only seven cover infrastructure. The infrastructure marketers are to own SEO and AEO for "inference, training, and compute".

    High impactMedium effort First result in 1 quarter Technical Content Marketing M…

    First action

    List the 20 searches with the most volume.

    How to measure it

    Infra organic visits

    Now
    –
    Target
    +200%
    By
    2 quarters

    Steps

    1. 1

      Research

      Search and AI-assistant queries.

      Technical Content Marketing Manager, Infrastructure 1 week

    2. 2

      Plan

      20 pages.

      Product Marketer, Infrastructure 1 week

    3. 3

      Write

      With engineers.

      Technical Content Marketing Manager, Infrastructure 1 quarter

    4. 4

      Link

      To product pages.

      Senior Product Designer ongoing

    5. 5

      Refresh prices

      Monthly.

      Technical Content Marketing Manager, Infrastructure monthly

    6. 6

      Measure

      Visits and quotes.

      GTM Operations Engineer monthly

    First result
    1 quarter
    Cost
    Writing time
    Owner
    Technical Content Marketing Manager, Infrastructure

    Smallest test

    Publish the H100 rental page.

    Why it works

    Write 20 pages: H100 and H200 rental prices, multi-node training with InfiniBand, fine-tuning a video model on dedicated GPUs, cold starts, autoscaling, and train-then-serve guides, each linking to Compute or Serverless.

    Infrastructure buyers search before they talk to sales; ranking there brings Compute and Serverless leads.

    Risks and how to handle them

    • Price pages go stale

      Date each update.

    Builds on

    Evidence

    • Your websiteSitemaps (2,476 URLs): 1,525 model pages with their own titles and descriptions; 227 /learn articles, mostly model prompt and developer guides plus model-vs-model comparisons; only 7 under /learn/serverless (including "nvidia-b200-price", "reserve-nvidia-b300" and "dedicated-gpus-vs-serverless-inference-break-even"); no pages comparing fal with other inference platforms or GPU clouds, and no multi-node training guides.Source
    • HiringProduct Marketer, Infrastructure (posted October 5, 2026): owns GTM for "Serverless for deploying and scaling custom AI models across thousands of GPUs, to Compute for training on large, interconnected GPU clusters"; must "Develop our competitive positioning across inference platforms, GPU clouds, hyperscalers, and in-house infrastructure", "Own SEO and AEO strategy for infrastructure, capturing developers actively searching for inference, training, and compute solutions" and create "benchmarks, product launches, solution pages" and "case studies".Source
    • HiringTechnical Content Marketing Manager, Infrastructure (posted October 5, 2026): "Some of the most interesting work happening at fal is deep inside the infrastructure, and most of the market never sees it." Writes "technical blogs, landing pages, guides, tutorials, solution pages" for Serverless and Compute, "benchmarks and competitive content", and "technical case studies".Source
    • Your websiteCompute documentation: "Dedicated GPU instances for training, fine-tuning"; "Compute runs on NVIDIA H100 SXM GPUs"; two instance types, 1xH100-SXM and 8xH100-SXM, with InfiniBand between 8-GPU instances in the same sector; "billed per-hour at fixed rates". The page is a technical introduction with no contact or quote form.Source
  • 08

    The infrastructure PMM must build "competitive positioning across inference platforms, GPU clouds, hyperscalers". fal has model-vs-model guides but no page comparing fal with other providers, while rivals such as Runware pitch "400K+ models, plus serverless GPUs and reserved GPU capacity".

    High impactMedium effort First result in 1 quarter Product Marketer

    First action

    Pick the two rivals sales meets most.

    How to measure it

    Comparison-page conversions

    Now
    –
    Target
    100 a month
    By
    2 quarters

    Steps

    1. 1

      Pick rivals

      From lost deals.

      Product Marketer, Infrastructure 1 week

    2. 2

      Check facts

      Current pricing and features.

      Product Marketer, Infrastructure 1 week

    3. 3

      Write

      Fair tables.

      Technical Content Marketing Manager, Infrastructure 2 weeks

    4. 4

      Legal review

      Claims.

      Senior Counsel, GTM 1 week

    5. 5

      Publish

      With migration guides.

      Senior Product Designer 1 week

    6. 6

      Refresh

      Quarterly.

      Product Marketer, Infrastructure quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Product Marketer, Infrastructure

    Smallest test

    Publish one comparison.

    Why it works

    Publish one factual page per rival: model catalogue, speed, pricing model, private deployments, compute options, enterprise features and migration steps.

    Comparison searches come from buyers close to a decision.

    Risks and how to handle them

    • Outdated claims

      Date and source every row.

    Evidence

    • HiringProduct Marketer, Infrastructure (posted October 5, 2026): owns GTM for "Serverless for deploying and scaling custom AI models across thousands of GPUs, to Compute for training on large, interconnected GPU clusters"; must "Develop our competitive positioning across inference platforms, GPU clouds, hyperscalers, and in-house infrastructure", "Own SEO and AEO strategy for infrastructure, capturing developers actively searching for inference, training, and compute solutions" and create "benchmarks, product launches, solution pages" and "case studies".Source
    • Competitor researchRivals' own homepages: Replicate, "Run open-source machine learning models with a cloud API"; Together AI, "The AI Native Cloud"; Modal, "Cloud infrastructure for teams that develop, train, and serve AI applications"; Runware, "One API for 400K+ AI models, plus serverless GPUs and reserved GPU capacity"; Lambda, "Launch on-demand instances or reserve a cluster".Source
    • Your websiteSitemaps (2,476 URLs): 1,525 model pages with their own titles and descriptions; 227 /learn articles, mostly model prompt and developer guides plus model-vs-model comparisons; only 7 under /learn/serverless (including "nvidia-b200-price", "reserve-nvidia-b300" and "dedicated-gpus-vs-serverless-inference-break-even"); no pages comparing fal with other inference platforms or GPU clouds, and no multi-node training guides.Source
  • 09

    The homepage says the fal Inference Engine is "up to 10x faster", and the content role calls for "benchmarks and competitive content". A claim with a method is more convincing to technical buyers.

    High impactMedium effort First result in 1 quarter Technical Content Marketing M…

    First action

    Pick five models to benchmark.

    How to measure it

    Benchmark citations

    Now
    –
    Target
    50
    By
    2 quarters

    Steps

    1. 1

      Pick models

      Five popular.

      Product Marketer, Infrastructure 1 week

    2. 2

      Run tests

      With engineers.

      Principal/Senior Forward Deployed Engineer 1 month

    3. 3

      Publish

      Method and script.

      Technical Content Marketing Manager, Infrastructure 2 weeks

    4. 4

      Promote

      X, Reddit, Discord.

      Community Growth 1 week

    5. 5

      Update

      At each launch.

      Technical Content Marketing Manager, Infrastructure monthly

    6. 6

      Use

      In sales.

      Account Executive, Enterprise ongoing

    First result
    1 quarter
    Cost
    Engineer time
    Owner
    Technical Content Marketing Manager, Infrastructure

    Smallest test

    Publish one model's benchmark.

    Why it works

    Publish benchmarks for popular image and video models: latency and throughput per model, hardware, settings and a script others can run. Update it as models launch.

    Engineers trust what they can reproduce, and benchmarks get cited in comparisons and by AI assistants.

    Risks and how to handle them

    • Rivals dispute numbers

      Publish the script and invite reruns.

    Evidence

    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
    • HiringTechnical Content Marketing Manager, Infrastructure (posted October 5, 2026): "Some of the most interesting work happening at fal is deep inside the infrastructure, and most of the market never sees it." Writes "technical blogs, landing pages, guides, tutorials, solution pages" for Serverless and Compute, "benchmarks and competitive content", and "technical case studies".Source
    • HiringProduct Marketer, Infrastructure (posted October 5, 2026): owns GTM for "Serverless for deploying and scaling custom AI models across thousands of GPUs, to Compute for training on large, interconnected GPU clusters"; must "Develop our competitive positioning across inference platforms, GPU clouds, hyperscalers, and in-house infrastructure", "Own SEO and AEO strategy for infrastructure, capturing developers actively searching for inference, training, and compute solutions" and create "benchmarks, product launches, solution pages" and "case studies".Source
  • 10

    fal adds new models constantly and is hiring for creator partnerships "around model and product launches, including early access". fal's own channels (@madeonfal, YouTube with 9,510 subscribers) are modest next to the creators who make AI video popular.

    High impactMedium effort First result in 1 quarter Influencer & Creator Partners…

    First action

    Shortlist 20 creators.

    How to measure it

    Creator-driven signups

    Now
    –
    Target
    5,000 a quarter
    By
    2 quarters

    Steps

    1. 1

      Shortlist

      20 creators.

      Influencer & Creator Partnerships 2 weeks

    2. 2

      Sign ten

      Terms and credits.

      Influencer & Creator Partnerships 1 month

    3. 3

      Early access

      Per launch.

      Product Marketer, Developer Platform per launch

    4. 4

      Publish

      With the launch.

      Influencer & Creator Partnerships per launch

    5. 5

      Track

      Links and codes.

      GTM Operations Engineer ongoing

    6. 6

      Review

      Keep the best.

      Influencer & Creator Partnerships quarterly

    First result
    1 quarter
    Cost
    Creator fees and credits
    Owner
    Influencer & Creator Partnerships

    Smallest test

    Give three creators early access to the next launch.

    Why it works

    Recruit ten creators with a high bar, give them early access a week before each launch, and publish their work with the launch. Track signups by creator link.

    Launch moments are when attention peaks; great early work from trusted creators sets fal as the place to try new models.

    Risks and how to handle them

    • Uneven quality

      Brief clearly and review before posting.

    Evidence

    • HiringInfluencer & Creator Partnerships (posted October 15, 2026): "a curated network of creators, AI filmmakers, designers, and creative technologists"; "Activate creators around model and product launches, including early access, sponsored projects, credits"; "We care deeply about quality over quantity."Source
    • Public profilesChannels: YouTube @fal.ai.official with 9,510 subscribers and 120 videos (874,666 views); Instagram and TikTok as @madeonfal; Discord and Reddit communities; GitHub fal-ai with 75 public repositories; the fal Python package is at version 1.82.1 (October 20, 2026).Source
    • HiringProduct Marketer, Developer Platform (posted October 5, 2026): owns GTM for "Model APIs for accessing 1,500+ generative media models", "fal Agent" and "fal CLI"; "The winning developer platform won't just have the best models. It'll make them easy to discover"; will "Develop customer and developer stories showcasing what people are building on fal" and "Work with Growth on SEO/AEO, lifecycle, and product-led growth to turn discovery into active users".Source
  • 11

    Community Growth owns "creator competitions", hackathons and Discord. Each new model is a natural theme.

    Medium impactMedium effort First result in 1 quarter Community Growth

    First action

    Pick next month's model.

    How to measure it

    Entries per competition

    Now
    –
    Target
    500
    By
    2 quarters

    Steps

    1. 1

      Theme

      Newest model.

      Community Growth monthly

    2. 2

      Judges

      Model partners.

      Community Growth monthly

    3. 3

      Run

      On Discord.

      Community Growth monthly

    4. 4

      Gallery

      Winners page.

      Senior Product Designer 2 weeks

    5. 5

      Templates

      From winners.

      Product Marketer, Developer Platform monthly

    6. 6

      Measure

      Entries and retention.

      Community Growth monthly

    First result
    1 quarter
    Cost
    Credit prizes
    Owner
    Community Growth

    Smallest test

    Run one competition.

    Why it works

    Each month, run a competition on the newest model with credits as prizes, judges from model partners, and a winners gallery. Turn winning workflows into templates.

    Regular competitions keep users returning and produce showcase content and templates.

    Risks and how to handle them

    • Rights to entries

      Clear terms for showcase use.

    Evidence

    • HiringCommunity Growth (posted September 9, 2026): "fal has a large community of developers, creative technologists, hackers, and creators"; the role owns "events, hackathons, creator competitions, webinars, Discord, Reddit, and developer content like demos, prompt guides, and walkthroughs" and a "network of power users... who test new fal tools, provide feedback, and build things we can showcase".Source
    • Your websiteSitemaps (2,476 URLs): 1,525 model pages with their own titles and descriptions; 227 /learn articles, mostly model prompt and developer guides plus model-vs-model comparisons; only 7 under /learn/serverless (including "nvidia-b200-price", "reserve-nvidia-b300" and "dedicated-gpus-vs-serverless-inference-break-even"); no pages comparing fal with other inference platforms or GPU clouds, and no multi-node training guides.Source
    • HiringInfluencer & Creator Partnerships (posted October 15, 2026): "a curated network of creators, AI filmmakers, designers, and creative technologists"; "Activate creators around model and product launches, including early access, sponsored projects, credits"; "We care deeply about quality over quantity."Source
  • 12

    fal has one industry page (advertising), yet its stories include gaming (Layer) and its logos include Shopify and Canva. The field marketer is to run "executive dinners and roundtables" for priority accounts.

    High impactMedium effort First result in 1 quarter Field Marketing Manager

    First action

    Pick the first industry and speaker.

    How to measure it

    Industry pipeline

    Now
    –
    Target
    +30%
    By
    2 quarters

    Steps

    1. 1

      Pick industries

      Three.

      Field Marketing Manager, Enterprise & ABM 1 week

    2. 2

      Write pages

      Use cases and stories.

      Technical Content Marketing Manager, Infrastructure 1 month

    3. 3

      Build

      Pages.

      Senior Product Designer 2 weeks

    4. 4

      Invite

      Target accounts.

      Field Marketing Manager, Enterprise & ABM 1 month

    5. 5

      Host

      Dinners.

      Event Marketing Manager 1 quarter

    6. 6

      Measure

      Pipeline.

      GTM Operations Engineer monthly

    First result
    1 quarter
    Cost
    Event budget
    Owner
    Field Marketing Manager, Enterprise & ABM

    Smallest test

    Publish the gaming page.

    Why it works

    Add pages for gaming, e-commerce and creative tools with use cases, models and a story each, then run one dinner per industry with a customer speaker.

    Buyers look for their own industry, and dinners convert best when a peer speaks.

    Risks and how to handle them

    • Thin proof

      Start with the gaming story.

    Evidence

    • Your websiteEnterprise Hub: "Custom training for Fortune 500 companies"; four customer highlights (Quora/Poe, Genspark, Pimento, Layer), which are also the only four /customer-case pages; the only industry page is /industry/advertising. Logos and quotes elsewhere include Canva, Adobe, Shopify, Suno, World Labs, Krea and HeyGen.Source
    • HiringEvents and field: an Event Marketing Manager (posted October 6, 2026) for "meetups, workshops, hackathons, dinners", webinars and "ABM activations"; a Field Marketing Manager, Enterprise & ABM (posted October 6, 2026) to "turn field marketing into a meaningful driver of pipeline and revenue" with "executive dinners and roundtables, customer gatherings, account-specific experiences".Source
    • HiringProduct Marketer, Infrastructure (posted October 5, 2026): owns GTM for "Serverless for deploying and scaling custom AI models across thousands of GPUs, to Compute for training on large, interconnected GPU clusters"; must "Develop our competitive positioning across inference platforms, GPU clouds, hyperscalers, and in-house infrastructure", "Own SEO and AEO strategy for infrastructure, capturing developers actively searching for inference, training, and compute solutions" and create "benchmarks, product launches, solution pages" and "case studies".Source
  • 13

    fal has 2.5 million developers and a growing sales team. Teams spending more each week are the best leads for reserved pricing, private deployments or Compute.

    High impactMedium effort First result in 1 quarter GTM Operations Engineer

    First action

    List workspaces whose spend doubled last month.

    How to measure it

    Usage-signal pipeline share

    Now
    –
    Target
    30%
    By
    2 quarters

    Steps

    1. 1

      Define signals

      Spend growth, models, seats.

      GTM Operations Engineer 1 week

    2. 2

      Build scoring

      Weekly job.

      GTM Operations Engineer 2 weeks

    3. 3

      Enrich

      Clay.

      GTM Operations Engineer 1 week

    4. 4

      Route

      By segment.

      GTM Operations Engineer 1 week

    5. 5

      Play

      Usage-based outreach.

      Account Executive, Commercial ongoing

    6. 6

      Measure

      Win rate.

      GTM Operations Engineer monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    GTM Operations Engineer

    Smallest test

    Route ten workspaces by hand.

    Why it works

    Score workspaces weekly on spend growth, model mix and team size, enrich them in Clay, and route those above a threshold to the commercial, enterprise or Compute AE with a usage summary.

    Usage-based leads close faster and expand more than cold outbound.

    Risks and how to handle them

    • Feels intrusive

      Offer help and pricing, not surveillance.

    Evidence

    • HiringSales build-out: a Sales Development Representative (Inbound) to "Qualify inbound leads from marketing campaigns and website inquiries" (posted September 29, 2026), a BDR, an Account Executive, Commercial for "creative studios, AI innovation teams, and AI developers" (October 21), an Account Executive, Enterprise (October 22), and a GTM Operations Engineer for "a GTM org of 30" building target lists and enrichment in Clay and Salesforce.Source
    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
    • HiringAI Native Account Executive, Compute (posted October 19, 2026): "More and more of them are coming to us for raw compute as well as inference and model APIs." The AE sells "dedicated GPU clusters, reserved and on-demand capacity, and the inference platform", owns "all net-new logo acquisition across a defined set of AI-native startups, neolabs and research-driven model builders", and builds pipeline "through founder and investor networks, targeted outbound, community and events". "This is a pure hunter role. You won't inherit a book."Source
  • 14

    The Startup Program targets VC-backed teams and gives credits through partner funds, and grants fund open-source projects. The Compute AE hunts "AI-native startups, neolabs" through "founder and investor networks".

    High impactMedium effort First result in 1 quarter AI Native Account Executive

    First action

    List program members already training models.

    How to measure it

    Program-sourced Compute deals

    Now
    –
    Target
    10
    By
    2 quarters

    Steps

    1. 1

      List members

      Training usage.

      GTM Operations Engineer 1 week

    2. 2

      Design offer

      Trial block.

      AI Native Account Executive, Compute 1 week

    3. 3

      Reach out

      Founders.

      AI Native Account Executive, Compute 1 month

    4. 4

      Fund intros

      Partner VCs.

      AI Native Account Executive, Compute 1 quarter

    5. 5

      Case study

      First win.

      Technical Content Marketing Manager, Infrastructure 1 quarter

    6. 6

      Measure

      Deals.

      GTM Operations Engineer monthly

    First result
    1 quarter
    Cost
    Trial compute
    Owner
    AI Native Account Executive, Compute

    Smallest test

    Call five members who train models.

    Why it works

    Flag program members who start fine-tuning or training, offer them a Compute trial block, and ask partner funds to introduce portfolio teams that are planning training runs.

    fal already has relationships with exactly the teams the Compute AE is told to find.

    Risks and how to handle them

    • Credits without conversion

      Time-limit the trial.

    Evidence

    • Your websiteStartup Program: credit tiers (for example "Growth $5,000", which "Unlocks at $2,500/mo on fal"); for teams that are VC-backed or have traction; "Partner-referred applications are prioritized"; "Partner perks... Coming soon". The Research Grants program offers a "$50k average grant check" of free compute for open-source projects.Source
    • HiringAI Native Account Executive, Compute (posted October 19, 2026): "More and more of them are coming to us for raw compute as well as inference and model APIs." The AE sells "dedicated GPU clusters, reserved and on-demand capacity, and the inference platform", owns "all net-new logo acquisition across a defined set of AI-native startups, neolabs and research-driven model builders", and builds pipeline "through founder and investor networks, targeted outbound, community and events". "This is a pure hunter role. You won't inherit a book."Source
    • Your websiteCompute documentation: "Dedicated GPU instances for training, fine-tuning"; "Compute runs on NVIDIA H100 SXM GPUs"; two instance types, 1xH100-SXM and 8xH100-SXM, with InfiniBand between 8-GPU instances in the same sector; "billed per-hour at fixed rates". The page is a technical introduction with no contact or quote form.Source
  • 15

    The developer platform marketer is to "Develop customer and developer stories", and Community Growth to "Find and amplify interesting builders". fal has workflows templates but no gallery of products built on it.

    Medium impactLow effort First result in 1 quarter Product Marketer

    First action

    Collect 20 projects from Discord.

    How to measure it

    Projects listed

    Now
    –
    Target
    200
    By
    2 quarters

    Steps

    1. 1

      Collect

      20 projects.

      Community Growth 2 weeks

    2. 2

      Build

      Gallery page.

      Senior Product Designer 2 weeks

    3. 3

      Form

      Submissions.

      GTM Operations Engineer 2 days

    4. 4

      Feature

      Monthly.

      Product Marketer, Developer Platform monthly

    5. 5

      Video

      YouTube spotlights.

      Product Marketer, Developer Platform monthly

    6. 6

      Measure

      Visits and signups.

      GTM Operations Engineer monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Product Marketer, Developer Platform

    Smallest test

    Publish ten projects.

    Why it works

    Launch a gallery of apps and projects with a short write-up each, a submission form and a monthly feature on X and YouTube.

    Showcasing builders brings new developers who want similar results and rewards existing ones.

    Risks and how to handle them

    • Low-quality entries

      Curate before listing.

    Evidence

    • HiringProduct Marketer, Developer Platform (posted October 5, 2026): owns GTM for "Model APIs for accessing 1,500+ generative media models", "fal Agent" and "fal CLI"; "The winning developer platform won't just have the best models. It'll make them easy to discover"; will "Develop customer and developer stories showcasing what people are building on fal" and "Work with Growth on SEO/AEO, lifecycle, and product-led growth to turn discovery into active users".Source
    • HiringCommunity Growth (posted September 9, 2026): "fal has a large community of developers, creative technologists, hackers, and creators"; the role owns "events, hackathons, creator competitions, webinars, Discord, Reddit, and developer content like demos, prompt guides, and walkthroughs" and a "network of power users... who test new fal tools, provide feedback, and build things we can showcase".Source
    • Public profilesChannels: YouTube @fal.ai.official with 9,510 subscribers and 120 videos (874,666 views); Instagram and TikTok as @madeonfal; Discord and Reddit communities; GitHub fal-ai with 75 public repositories; the fal Python package is at version 1.82.1 (October 20, 2026).Source
  • 16

    fal's YouTube channel has 120 videos and 9,510 subscribers, while its learn hub publishes prompt guides for each model. Developers search YouTube for how to use new models.

    Medium impactMedium effort First result in 1 quarter Product Marketer

    First action

    Record the next launch.

    How to measure it

    YouTube subscribers

    Now
    –
    Target
    25,000
    By
    2 quarters

    Steps

    1. 1

      Template

      Five-minute format.

      Product Marketer, Developer Platform 1 week

    2. 2

      Record

      Per launch.

      Community Growth per launch

    3. 3

      Embed

      On model pages.

      Senior Product Designer per launch

    4. 4

      Clips

      Shorts.

      Influencer & Creator Partnerships per launch

    5. 5

      Track

      Links.

      GTM Operations Engineer ongoing

    6. 6

      Review

      Monthly.

      Product Marketer, Developer Platform monthly

    First result
    1 quarter
    Cost
    Production time
    Owner
    Product Marketer, Developer Platform

    Smallest test

    Record one walkthrough.

    Why it works

    Record a five-minute walkthrough for each major model on launch day (prompts, API call, cost), link it from the model page and the prompt guide.

    Video guides rank in search and AI answers and bring new users at the moment of interest.

    Risks and how to handle them

    • Too many launches

      Only major models get a video.

    Evidence

    • Public profilesChannels: YouTube @fal.ai.official with 9,510 subscribers and 120 videos (874,666 views); Instagram and TikTok as @madeonfal; Discord and Reddit communities; GitHub fal-ai with 75 public repositories; the fal Python package is at version 1.82.1 (October 20, 2026).Source
    • Your websiteSitemaps (2,476 URLs): 1,525 model pages with their own titles and descriptions; 227 /learn articles, mostly model prompt and developer guides plus model-vs-model comparisons; only 7 under /learn/serverless (including "nvidia-b200-price", "reserve-nvidia-b300" and "dedicated-gpus-vs-serverless-inference-break-even"); no pages comparing fal with other inference platforms or GPU clouds, and no multi-node training guides.Source
    • HiringProduct Marketer, Developer Platform (posted October 5, 2026): owns GTM for "Model APIs for accessing 1,500+ generative media models", "fal Agent" and "fal CLI"; "The winning developer platform won't just have the best models. It'll make them easy to discover"; will "Develop customer and developer stories showcasing what people are building on fal" and "Work with Growth on SEO/AEO, lifecycle, and product-led growth to turn discovery into active users".Source

Big bets 3

Larger plays that can change the trajectory.

  • 17

    fal publishes a yearly gated report and sees usage across 1,500+ models. A public quarterly index (most used models by category, price per output, speed trends) would be cited every quarter.

    High impactMedium effort First result in 3 quarters Product Marketer

    First action

    Pick five charts that can be public.

    How to measure it

    Report leads

    Now
    –
    Target
    +50%
    By
    3 quarters

    Steps

    1. 1

      Choose charts

      Five.

      Product Marketer, Developer Platform 2 weeks

    2. 2

      Aggregate

      Privacy-safe.

      Senior Software Engineer, Data Platform 1 month

    3. 3

      Legal

      Review.

      Senior Counsel, GTM 2 weeks

    4. 4

      Publish

      Index page.

      Senior Product Designer 2 weeks

    5. 5

      Press

      Each quarter.

      Product Marketer, Developer Platform quarterly

    6. 6

      Leads

      Report CTA.

      GTM Operations Engineer ongoing

    First result
    3 quarters
    Cost
    Analyst time
    Owner
    Product Marketer, Developer Platform

    Smallest test

    Publish one chart.

    Why it works

    Publish a free quarterly index page with charts from aggregated usage; keep the full report gated for leads.

    Regular, original data makes fal the reference source for generative media and earns links and AI citations.

    Risks and how to handle them

    • Partners dislike rankings

      Share with model partners before publishing.

    Evidence

    • Your websiteThe State of Generative Media Vol. 2: the report page is a form (name, work email, company, role, company size) with "We'll use your details to provide report access and may follow up about fal"; Volume 1 is also in the sitemap.Source
    • Your websiteHomepage (checked October 22, 2026): "Trusted by over 2,500,000 developers"; "Access 1,500+ generative media models"; the Compute section says "Access large, interconnected clusters of H100, H200, and B200 GPUs" to "Scale workloads across hundreds or thousands of GPUs". All four Compute links on the homepage go to the documentation (fal.ai/compute redirects to /docs/documentation/compute); Serverless has its own product page. Quotes from Black Forest Labs, Canva and Quora ("fal powers 40% of Poe's official image and video generation bots").Source
    • Your websiteSitemaps (2,476 URLs): 1,525 model pages with their own titles and descriptions; 227 /learn articles, mostly model prompt and developer guides plus model-vs-model comparisons; only 7 under /learn/serverless (including "nvidia-b200-price", "reserve-nvidia-b300" and "dedicated-gpus-vs-serverless-inference-break-even"); no pages comparing fal with other inference platforms or GPU clouds, and no multi-node training guides.Source
  • 18

    Compute buyers want to know what is available and when. fal's Serverless page already shows live GPU usage; Lambda and Runware sell reserved capacity directly.

    High impactHigh effort First result in 3 quarters AI Native Account Executive

    First action

    Show availability for H100 clusters only.

    How to measure it

    Online reservations

    Now
    –
    Target
    30 a quarter
    By
    3 quarters

    Steps

    1. 1

      Define sizes

      Standard blocks.

      AI Native Account Executive, Compute 1 month

    2. 2

      Availability data

      From fleet.

      Compute Operations 1 quarter

    3. 3

      Build view

      Compute page.

      Senior Product Designer 1 month

    4. 4

      Reserve flow

      Deposit and terms.

      Senior/Staff Software Engineer, Core Product 1 quarter

    5. 5

      Legal

      Terms.

      Lead AI Infrastructure Counsel 1 month

    6. 6

      Measure

      Reservations.

      GTM Operations Engineer monthly

    First result
    3 quarters
    Cost
    Engineering time
    Owner
    AI Native Account Executive, Compute

    Smallest test

    Publish an availability table updated weekly.

    Why it works

    Add an availability view to the Compute page (GPU type, count, earliest start) and a reserve flow for standard sizes with a deposit, with larger deals going to the AE.

    Self-serve reservations shorten sales cycles for smaller teams and free the AE for big deals.

    Risks and how to handle them

    • Availability changes fast

      Show ranges and confirm by email.

    Builds on

    Evidence

    • HiringAI Native Account Executive, Compute (posted October 19, 2026): "More and more of them are coming to us for raw compute as well as inference and model APIs." The AE sells "dedicated GPU clusters, reserved and on-demand capacity, and the inference platform", owns "all net-new logo acquisition across a defined set of AI-native startups, neolabs and research-driven model builders", and builds pipeline "through founder and investor networks, targeted outbound, community and events". "This is a pure hunter role. You won't inherit a book."Source
    • Competitor researchRivals' own homepages: Replicate, "Run open-source machine learning models with a cloud API"; Together AI, "The AI Native Cloud"; Modal, "Cloud infrastructure for teams that develop, train, and serve AI applications"; Runware, "One API for 400K+ AI models, plus serverless GPUs and reserved GPU capacity"; Lambda, "Launch on-demand instances or reserve a cluster".Source
    • Your websiteServerless page: "fal Serverless powers 1,300+ production endpoints across thousands of GPUs", with a live GPU-usage view and an "Explore Demo" link. The footer also links "Verify fal-Generated Content", and the advertising page promises "Compliance from the moment of creation".Source
    • Your websitePricing page: a single "Serverless & Compute Pricing" table with list and "as low as" hourly prices (B300 $12.99/h list, as low as $5.99/h; B200 $7.99/h, as low as $5.49/h; H200 $6.00/h, as low as $2.99/h; H100 $4.50/h, as low as $2.49/h); for custom deployments it says "Contact [email protected] to get started". There is no separate section for reserved clusters, multi-node training or commitment terms.Source
  • 19

    fal offers training on Compute and serving on Serverless with "1,300+ production endpoints", but the docs present them as separate choices. Teams that fine-tune want to deploy in the same place.

    High impactHigh effort First result in 3 quarters Product Marketer

    First action

    Document the current manual steps.

    How to measure it

    Compute + Serverless customers

    Now
    –
    Target
    50
    By
    3 quarters

    Steps

    1. 1

      Map steps

      Current flow.

      Principal/Senior Forward Deployed Engineer 2 weeks

    2. 2

      Build path

      One command.

      Senior/Staff Software Engineer, Core Product 1 quarter

    3. 3

      Example

      Reference repo.

      Principal/Senior Forward Deployed Engineer 1 month

    4. 4

      Case study

      First customer.

      Technical Content Marketing Manager, Infrastructure 1 quarter

    5. 5

      Launch

      Blog and events.

      Product Marketer, Infrastructure 1 month

    6. 6

      Measure

      Cross-product use.

      GTM Operations Engineer quarterly

    First result
    3 quarters
    Cost
    Engineering time
    Owner
    Product Marketer, Infrastructure

    Smallest test

    Publish the manual guide first.

    Why it works

    Build and market a one-command path from a Compute training run to a Serverless endpoint, with a reference example and a customer case study.

    A joined path makes fal the full-stack choice rivals with only one half cannot match.

    Risks and how to handle them

    • Complex workloads

      Start with fine-tuning jobs.

    Evidence

    • Your websiteCompute documentation: "Dedicated GPU instances for training, fine-tuning"; "Compute runs on NVIDIA H100 SXM GPUs"; two instance types, 1xH100-SXM and 8xH100-SXM, with InfiniBand between 8-GPU instances in the same sector; "billed per-hour at fixed rates". The page is a technical introduction with no contact or quote form.Source
    • Your websiteServerless page: "fal Serverless powers 1,300+ production endpoints across thousands of GPUs", with a live GPU-usage view and an "Explore Demo" link. The footer also links "Verify fal-Generated Content", and the advertising page promises "Compliance from the moment of creation".Source
    • HiringProduct Marketer, Infrastructure (posted October 5, 2026): owns GTM for "Serverless for deploying and scaling custom AI models across thousands of GPUs, to Compute for training on large, interconnected GPU clusters"; must "Develop our competitive positioning across inference platforms, GPU clouds, hyperscalers, and in-house infrastructure", "Own SEO and AEO strategy for infrastructure, capturing developers actively searching for inference, training, and compute solutions" and create "benchmarks, product launches, solution pages" and "case studies".Source

Wildcards 2

Unexpected experiments, each with a cheap way to test it.

  • 20

    fal's Research Grants give an average $50k of compute to open-source projects. The projects are not shown on the grants page.

    Medium impactLow effort First result in 1 quarter Community Growth

    First action

    List past grant recipients.

    How to measure it

    Applications

    Now
    –
    Target
    +50%
    By
    2 quarters

    Steps

    1. 1

      List

      Recipients.

      Community Growth 1 week

    2. 2

      Ask

      Permission and credit.

      Community Growth 2 weeks

    3. 3

      Build

      Showcase page.

      Senior Product Designer 1 week

    4. 4

      Promote

      X and Discord.

      Community Growth 1 week

    5. 5

      Apply form

      Replace the email.

      GTM Operations Engineer 2 days

    6. 6

      Measure

      Applications.

      Community Growth quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Community Growth

    Smallest test

    Publish five projects.

    Why it works

    List funded projects with a short description and results, ask maintainers to credit fal, and use the page to attract more strong applicants.

    Public results show the program works and earn credit links from popular repos.

    Risks and how to handle them

    • Projects unwilling

      List only those who agree.

    Evidence

    • Your websiteStartup Program: credit tiers (for example "Growth $5,000", which "Unlocks at $2,500/mo on fal"); for teams that are VC-backed or have traction; "Partner-referred applications are prioritized"; "Partner perks... Coming soon". The Research Grants program offers a "$50k average grant check" of free compute for open-source projects.Source
    • HiringCommunity Growth (posted September 9, 2026): "fal has a large community of developers, creative technologists, hackers, and creators"; the role owns "events, hackathons, creator competitions, webinars, Discord, Reddit, and developer content like demos, prompt guides, and walkthroughs" and a "network of power users... who test new fal tools, provide feedback, and build things we can showcase".Source
    • Your websiteInconsistent claims: the homepage meta description (shown in search results) still says "Access 1,000+ AI models" and "Trusted by 1.5M devs" while the page says 1,500+ models and 2,500,000 developers; the Enterprise Hub says "fal powers 50% of Poe's image and video generation" while the homepage says 40%; the Grants page stat block pairs "$2m" with "faster response time compared to other providers" and "12" with "companies incubated".Source
  • 21

    fal links a "Verify fal-Generated Content" tool in its footer and its advertising page promises "Compliance from the moment of creation". Enterprise brands worry about the provenance of AI-generated media.

    Medium impactLow effort First result in 1 quarter Field Marketing Manager

    First action

    Write the provenance page outline.

    How to measure it

    Deals where provenance helped

    Now
    –
    Target
    10
    By
    2 quarters

    Steps

    1. 1

      Outline

      Page and checklist.

      Field Marketing Manager, Enterprise & ABM 1 week

    2. 2

      Write

      With legal.

      Technical Content Marketing Manager, Infrastructure 2 weeks

    3. 3

      Build

      Page.

      Senior Product Designer 1 week

    4. 4

      Enable

      Sales talk track.

      Account Executive, Enterprise 1 week

    5. 5

      Present

      At dinners.

      Field Marketing Manager, Enterprise & ABM ongoing

    6. 6

      Measure

      Deal notes.

      GTM Operations Engineer quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Field Marketing Manager, Enterprise & ABM

    Smallest test

    Add the checklist to one enterprise proposal.

    Why it works

    Make provenance a named enterprise feature: a page explaining verification, a checklist for brand and legal teams, and a session at each ABM dinner.

    Legal and brand teams are often the last blockers in enterprise media deals; a clear provenance story removes one.

    Risks and how to handle them

    • Overclaiming

      Describe exactly what verification covers.

    Evidence

    • Your websiteServerless page: "fal Serverless powers 1,300+ production endpoints across thousands of GPUs", with a live GPU-usage view and an "Explore Demo" link. The footer also links "Verify fal-Generated Content", and the advertising page promises "Compliance from the moment of creation".Source
    • Your websiteEnterprise Hub: "Custom training for Fortune 500 companies"; four customer highlights (Quora/Poe, Genspark, Pimento, Layer), which are also the only four /customer-case pages; the only industry page is /industry/advertising. Logos and quotes elsewhere include Canva, Adobe, Shopify, Suno, World Labs, Krea and HeyGen.Source
    • HiringEvents and field: an Event Marketing Manager (posted October 6, 2026) for "meetups, workshops, hackathons, dinners", webinars and "ABM activations"; a Field Marketing Manager, Enterprise & ABM (posted October 6, 2026) to "turn field marketing into a meaningful driver of pipeline and revenue" with "executive dinners and roundtables, customer gatherings, account-specific experiences".Source

03

How Programmable can help

Where an idea in this playbook is something Programmable builds or runs, it's listed here with what we'd do. Everything else is yours to run.

ProgrammableTalk

Fill the events in this plan

These ideas are events. Talk gets registrants for them and tells you who's coming and what their company does before the day.

Plan an event (opens programmable.talk in a new tab)
More example playbooks and reviews
ProgrammableDigital

Show fal’s product working

Turn screenshots or the website into an animated product demo in its own brand, for the homepage, launch posts and outreach in this playbook.

Make a demo (opens programmable.digital in a new tab)

More growth playbooks

Other companies we have researched, each read from what they are hiring for.

  • SMV Virtual Consulting growth strategyA no-code and AI systems consultancy that designs and builds custom databases, CRMs, client portals, project-management workspaces and…24 growth ideas
  • General Astronautics growth strategyA Y Combinator (Winter 2026) space robotics startup building autonomous dexterous robotic arms for orbital inspection, repair and…23 growth ideas
  • Maze growth strategyA user and market research platform: participant recruitment from its own panel, unmoderated tests, surveys, moderated and AI-moderated…21 growth ideas
  • The Design Cart growth strategyAn online store for fabrics, beads, buttons, laces, embellishments, jewellery findings and threads for fashion and jewellery makers, with…21 growth ideas
  • OpenArt growth strategyAn AI creative platform for images, video, audio and characters, with 100+ models in one place, Director for chat-based "vibe directing"…22 growth ideas
  • Anchorage Digital growth strategyAn institutional crypto platform with a federally chartered crypto bank: custody, trading and financing, staking, settlement, stablecoin…21 growth ideas
  • HelloFresh growth strategyAmerica's largest meal-kit subscription: weekly boxes of pre-portioned ingredients and step-by-step recipes from 100+ weekly options, part…22 growth ideas
  • BetterHelp growth strategyThe world's largest online therapy service: matching people with licensed therapists for messaging, chat, phone and video sessions, with…21 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on October 23, 2026, and cites them throughout. fal did not commission, review or endorse it, and Programmable is not affiliated with fal. Company and product names are trademarks of their owners, used only to identify them. The ideas are suggestions and opinion, not professional, legal or financial advice, and not statements about how fal runs its business.

Something wrong, or do you represent fal? Email [email protected] and we will correct it or take it down.

30/60/90 roadmap

What to focus on in each month, and the ideas that carry it.

  1. Days 1 to 307 ideas

    Compute page and quote form, events page, consistent numbers, homepage speed, lead forms instead of support@, and the first big-logo stories; list grant projects.

  2. Days 31 to 6011 ideas

    Infrastructure content, comparisons, benchmarks, creator bench and competitions, industry pages and dinners, usage-based routing, Startup Program to Compute, the builder gallery and launch videos; provenance page.

  3. Days 61 to 903 ideas

    Commit to the big bets: a public quarterly index, online cluster reservations and a train-then-serve path.

The research behind the plan

3 research areas, each summarised in a few lines. Open any of them for the full detail.

Customers

What customers say

Customers value new models available on launch day. The most cited problem: keeping up with model releases.

  • Fast inference and reliability at scale
  • GPU cost and availability
  • In their words: “generative media”

Market

Competitive landscape

5 competitors compared: Replicate, Together AI, Modal, Runware and 1 more.

  • Gap to exploit: Generative media training and serving on one platform
  • Replicate: Less focus on generative media speed
  • Together AI: Language-model focus

Hiring signals

fal is moving from developer API to full infrastructure and enterprise sales.

  • AI Native Account Executive, Compute: Customers asking for raw compute; new-logo hunter. Ideas 1, 14, 18.
  • Product Marketer, Infrastructure: GTM, SEO/AEO and positioning for Serverless and Compute. Ideas 1, 7, 8, 19.
  • Technical Content Marketing Manager, Infrastructure: Landing pages, benchmarks and case studies. Ideas 6, 7, 9.

From the team that wrote this playbook

All products
  • ProgrammableAgencyWorking AI systems in days, not quarters (opens programmable.agency in a new tab)
  • ProgrammableInfoAn agent that finds, researches and asks firstNew accounts get 100,000 free enrichment credits with code BETA100K (opens programmable.info in a new tab)
  • ProgrammableDigitalScreenshots in, an animated product demo out (opens programmable.digital in a new tab)
  • ProgrammableWatchKnow the day a supplier falls out of compliance (opens programmable.watch in a new tab)
  • ProgrammableTalkGet registrants for your events, and know who’s coming (opens programmable.talk in a new tab)
  • ProgrammableMediaAn AI note taker, interpreter and webinar host for your calls (opens programmable.media in a new tab)
  • ProgrammablePageText what you need, your customer gets a page that works (opens programmable.page in a new tab)
  • ProgrammableCommunityForums with labelled AI members, never empty (opens programmable.community in a new tab)
ProgrammableMarketing
ExamplesLeaderboardsProductsTermsPrivacyCookiesDPALegalContact

© 2026 AI-Powered Growth Intelligence