01 · Landscape
Landscape snapshot
Three AI coding leaders — Cursor, GitHub Copilot (Microsoft) and Claude Code (Anthropic) — produced ~36 unique creatives across 218 raw impressions in this dataset, spanning single image, video, document and message formats. The combined set covers the full AI-coding category: AI-native IDE (Cursor), IDE extension (Copilot), and CLI/terminal-first agent (Claude Code). Three completely different form factors competing for the same developer wallet — and three completely different paid-media bets.
The headline pattern: each advertiser brands itself differently. Cursor is the only advertiser in the category that brands itself as an AI coding tool — every creative names Cursor and centers the product. GitHub Copilot never names Copilot in any of its in-market creative; it markets GitHub-the-platform around "agentic AI" + "DevOps" instead, leaving the Copilot brand implicit. Anthropic markets Claude as a horizontal platform across multiple verticals (Legal, AI-native Startups, developer multi-agent systems) and product pillars (Claude Code, Claude Cowork, Microsoft Foundry partnership, Opus 4.8 launch). In a category where the buyer is shopping for an AI coding tool, Cursor is the only advertiser actually selling one by name.
Library shape and concentration split sharply. Cursor's 100-item library is dominated by a single $500-Teams-Credits creative running ~82 of 100 impressions; the Emily ABM DMs, Gartner Magic Quadrant badge, Developer Habits Report and Rippling case study all run at low single-digit frequency. GitHub runs ~6 enterprise-IT-positioning ads in a perfectly even rotation. Anthropic's 99-item library is the most diversified of the three — ~20 unique creatives across Opus 4.8 launch, Claude Code features, Microsoft Foundry partnership, a full Legal-vertical track (5 variants), an AI-native Startup playbook track (3 variants), Claude Cowork as a developed product pillar, and a named-Anthropic-employee video testimonial series (Lisa Crofoot, product manager; Chuma Kabaghe, product engineer) running at the highest frequency in the set. Format mix tilts heavily to documents (Anthropic ~50%) and single image (Cursor ~95%, GitHub 100%).
Confidence note: Built from full libraries for Cursor (100 items) and Anthropic (99 items), plus the full GitHub library (19 items) in the 30-day window. Frequency distributions reflect actual dedup counts. HIGH confidence on volume-related dimensions; qualitative findings stable.
02 · Industry trends
Strategic industry trends
Outside-in market context. Each trend is anchored to a credible published source, paired with what it means for Cursor's paid marketing.
| Trend & market signal | Marketing implication for Cursor |
|---|---|
|
01
Cursor was named a Leader in Gartner's Magic Quadrant for AI Code Assistants
Gartner's 2025 Magic Quadrant for AI Code Assistants placed Cursor in the Leaders quadrant — alongside GitHub and ahead of multiple incumbents. Cursor's in-market creative cites this directly across multiple ad variations. Gartner — Magic Quadrant for AI Code Assistants ↗ |
The analyst-validation layer is now table stakes for the AI-coding category, and Cursor is the only advertiser of the three actively running it. Claude Code and Copilot leave third-party validation entirely uncontested. |
|
02
AI coding tools are the fastest-spreading developer technology in history
The 2025 Stack Overflow Developer Survey found 76% of developers are using or planning to use AI tools in their development process — up from 70% in 2024. JetBrains' State of Developer Ecosystem found 49% adoption of AI coding assistants in production code. Stack Overflow — 2025 Developer Survey ↗ |
The category is past the early-majority threshold. The growth engine for category leaders now comes from displacing the buyer's existing tool, not from converting the holdout. The advertiser that owns the switch-cost narrative wins the displacement market. |
|
03
Agentic coding is the dominant 2026 messaging theme
All three advertisers center "agentic" framing in 2026 creative — Cursor's Bugbot + Cursor 3 agents, GitHub's "agentic AI + DevOps" rotation, Claude Code's /goal and /remote-control slash commands for hands-off coding. Gartner names agentic AI the #1 emerging enterprise technology for 2026. Gartner — Top Strategic Technology Trends 2026 ↗ |
When every competitor leads with the same theme, the theme stops being the differentiator — execution and proof become the only edges. Cursor's named enterprise customers (Rippling 150 → 500+ engineers, NVIDIA, Stripe) plus the Developer Habits Report put it furthest along on proof. |
|
04
GenAI is now the default starting point for B2B research
89% of B2B buyers use generative AI for self-guided research; AI determines the consideration set, peers validate the finalist. Buyers entering a CFO or VP Engineering finance-tooling decision have already been shaped by an AI answer. Forrester — The State Of Business Buying, 2026 ↗ |
Win the AI-discovery layer. The advertiser whose product name, category language and benchmark stats are most retrievable by AI assistants gets cited first. Cursor's brand-as-category-language move ("Cursor" = AI coding) is the strongest GEO posture of the three. |
|
05
The enterprise IT buyer is now the highest-value AI-coding ICP
JetBrains 2025 State of Developer Ecosystem reports enterprise IT and engineering leadership now drive AI-coding-tool buying decisions in 64% of large orgs, up from 41% a year earlier. The IC-developer-led PLG flywheel has matured into enterprise procurement. JetBrains — State of Developer Ecosystem 2025 ↗ |
Cursor's enterprise pivot (Cursor Teams + ABM DMs to engineering leaders) is timed correctly. GitHub Copilot speaks to enterprise IT but doesn't name the product. Claude Code is positioned to developers and has no enterprise-leader-targeted creative in the set. |
|
06
AI coding companies sit on a unique paid-media paradox
Most successful AI-coding companies grew without paid media — Cursor itself crossed $500M ARR primarily via developer Twitter, GitHub stars and word-of-mouth. The shift to paid professional-social advertising in late 2025 / early 2026 signals the category has matured into enterprise procurement requiring sales-assisted buying. Cursor — $500M ARR Announcement ↗ |
The category's marketing playbook is being written right now. The advertiser that establishes the dominant paid-media playbook in 2026 sets the category's default expectations for years. Cursor is most aggressively writing that playbook. |
|
07
GitHub's Copilot brand is being subsumed under broader GitHub messaging
GitHub's in-market creative does not mention Copilot by name in any 30-day-window ad — all positioning is around "agentic AI + DevOps" with GitHub-the-platform as the brand. Microsoft's broader AI consolidation has put Copilot inside the platform narrative, not above it. Microsoft Build 2025 — Agentic Web Platform Strategy ↗ |
The Copilot brand equity is being diluted, intentionally or not. For Cursor and Claude Code, this is uncontested category-naming whitespace — every paid impression that names "AI coding tool" goes to the brand that puts itself in front of the words. |
03 · Pattern counts
Pattern counts
Messaging themes
- 1 of 3 advertisers brands itself as the AI coding tool. Cursor names Cursor in every creative; GitHub never names Copilot; Anthropic positions Claude Code as one feature among many. Implication: in a category where the buyer is shopping for an AI coding tool, Cursor is the only advertiser advertising one.
- 3 of 3 advertisers center "agentic" framing, but execute it completely differently. Cursor: Bugbot agents on every PR + Cursor 3 end-to-end agents. GitHub: "agentic AI playbook" + DevOps. Claude Code: /goal and /remote-control slash commands. Same theme, three different products being sold.
- 1 of 3 advertisers runs named ABM DMs in the set. Cursor's MESSAGE-format DMs (signed by 'Emily', naming NVIDIA, Stripe, Salesforce, Adobe, Hilton, Fox) are the only ABM personalization in the AI-coding category right now.
Cursor: "Hey [FIRSTNAME] — The teams that are shipping faster with AI aren't bolting it onto their existing workflow. They're redesigning their workflows around it and pulling ahead. NVIDIA, Stripe, Salesforce, Adobe, Hilton, and Fox all made that shift with Cursor. — Emily"
GitHub: "AI that collaborates, not just assists. See how agentic AI helps your team keep code moving safely and efficiently."
Claude Code (Anthropic): "With Opus 4.8, you can hand off long-running coding work to Claude Code and walk away. Ship features with /goal and step away from your computer with /remote-control."
Three openers, three positions. Cursor names six customers in one DM. GitHub names no products, no customers. Anthropic names the model first, Claude Code second. The first reads like enterprise B2B SaaS; the second reads like a generic enterprise IT pitch; the third reads like a product-launch announcement.
Industry targeting
- Cursor runs named-enterprise-customer rotation (NVIDIA, Stripe, Salesforce, Adobe, Hilton, Fox, Rippling) but no vertical-specific tracks. GitHub runs no named vertical or customer — pitches to "the enterprise" generically. Anthropic (Claude Code) runs the only explicit vertical-specific campaigns in the AI-coding category: a full Legal track ("87% of general counsel," contract review, M&A diligence, litigation prep — 5 variants) plus an AI-native Startup playbook track (3 variants).
- Implication: Vertical targeting in AI coding is uncontested by Cursor and Copilot. Anthropic's Legal + Startup vertical tracks are a structural moat the other two are leaving open. Cursor's strongest move is to extend named-customer rotation into vertical-named cohorts (Healthcare engineering teams, FinTech engineering teams).
Content / offer types
- 1 of 3 advertisers runs owned original research. Cursor's Developer Habits Report ("the most comprehensive dataset on AI coding in the world") is the only owned-research engine in the category — but it runs at single-digit frequency, dwarfed by the $500 Credits creative (82x).
- 1 of 3 advertisers runs a direct-incentive offer. Cursor's $500 Teams credit is the only conversion-asset incentive — and it absolutely dominates Cursor's rotation, running ~82% of the library.
- 1 of 3 advertisers runs explicit vertical lead-magnet content. Anthropic publishes a Legal deployment playbook ("deployment plan," "deployment guide for teams ready to lead") and an AI-native Startup playbook — neither Cursor nor GitHub has vertical-specific content in active rotation.
Format preferences
- Cursor runs ~95% single image (the $500 Credits creative dominates), ~5% MESSAGE-format ABM DMs, <1% video and document combined. The format spread is wider in unique-creative-count terms but heavily concentrated by impression.
- GitHub: 100% single image. Zero video, zero document, zero message. Most format-concentrated advertiser in the study.
- Anthropic (Claude Code): ~50% document, ~30% video, ~20% single image. Document share is by far the highest in the AI-coding category. Video share carries the named-employee testimonial series (Lisa Crofoot, Chuma Kabaghe).
- Implication: Anthropic owns the document format in this category uncontested. Cursor's ABM DMs are the only MESSAGE-format play. GitHub's monoformat rotation is the most structurally exposed.
Funnel stage focus
- Cursor: barbell — bottom-funnel DMs targeting named engineering leaders + mid-funnel Teams-credit conversion offer + top-funnel Gartner badge and Developer Habits Report. Covers the full funnel.
- GitHub: exclusively top-funnel awareness — every creative is "Learn more" pointed at content, no demo CTA, no trial CTA, no direct conversion.
- Claude Code: mid-funnel — "Learn more" CTAs on every creative pointing to product announcement pages and feature docs.
- Implication: Cursor is the only advertiser playing the full funnel. Both rivals leave the bottom-funnel slot uncontested.
Social proof strategy
- Named enterprise customers in active creative: Cursor only — NVIDIA, Stripe, Salesforce, Adobe, Hilton, Fox, Rippling appear by name (in Emily DMs at low frequency).
- Named-employee testimonial proof: Anthropic only — Lisa Crofoot (product manager at Anthropic) and Chuma Kabaghe (product engineer at Anthropic) appear in video testimonials at the highest frequency in Anthropic's library. Parallel in spirit to Cursor's Emily DMs, but the named person is an internal employee rather than a sales lead.
- Vertical thought-leadership proof: Anthropic only — "the legal teams at the world's leading companies choose Claude," "87% of general counsel are using AI," Legal deployment guide.
- Owned original research: Cursor only — the Developer Habits Report.
- Third-party analyst signal: Cursor only — Gartner Magic Quadrant Leader recognition.
- Quantified customer outcome: Cursor only — "Rippling moved from 150 to 500+ engineers on Cursor."
- Implication: Cursor and Anthropic each have a substantial proof stack — Cursor on enterprise-customer-and-analyst proof; Anthropic on vertical-thought-leadership and named-employee proof. GitHub leaves every proof texture uncontested. Cursor's enterprise-customer proof is more defensible against an issuer (Microsoft); Anthropic's vertical content + employee testimonials are more defensible against an IDE-extension incumbent.
Tone / voice
- Three distinct voices. Cursor runs enterprise-B2B-mature voice — "engineering leaders at Stripe, NVIDIA, and Salesforce rebuilt how their teams work," "the gap between your best engineer's setup and everyone else's is a real cost." Sales-team-signed DMs read as direct outreach. GitHub runs IT-magazine voice — "AI that collaborates, not just assists," "scaling it requires the engineering rigor behind modern DevOps." Uniform, slightly corporate. Claude Code runs product-launch voice — "We're upgrading Claude Opus to a new version: Claude Opus 4.8. Available today." Feature-led and confident.
- Implication: Voice differentiation is real but isn't the white space. Structural moves (named customers, owned research, ABM DMs, full-funnel coverage) are where the gaps live.
04 · Whitespace
Strategic whitespace
01
Brand the Developer Habits Report as the category benchmark
GapCursor publishes the only owned research in the AI-coding category — the Developer Habits Report — but currently treats it as a single content drop rather than a recurring benchmark franchise.
Why it mattersForrester 2026: B2B buyers reward proof over promise. Mercury's New Economics franchise model proves that a recurring, named research series compounds — each annual edition cites the prior, deepening category authority. The AI-coding category has no equivalent today.
Recommended playProductize the Developer Habits Report as an annual benchmark. Publish a quarterly mini-update with one new finding. Open a public Cursor Insights microsite at cursor.com/insights with structured pages engineered for AI-assistant retrieval. The first advertiser to own "the State of AI Coding" research franchise locks in category citation authority.
02
Take the Copilot brand-dilution window
GapGitHub no longer names Copilot in its in-market creative — all positioning has shifted to GitHub-the-platform. The Copilot brand is being subsumed under broader Microsoft AI consolidation.
Why it mattersBrand dilution at an incumbent opens whitespace at the category-language level. The buyer searching "the best AI coding tool" doesn't have a single Copilot ad on the SERP today. Cursor is positioned to fill that void faster than Claude Code is.
Recommended playLead a "What is the best AI coding tool" GEO-optimized landing page with structured comparison data, schema markup and verified capability claims. Run a Cursor-vs-Copilot comparison page (factual, public) that AI assistants will cite when buyers ask. Extend ABM DMs to include the brand-dilution context: "Microsoft is folding Copilot into agentic platforms; here's why developer teams are switching to Cursor."
03
Build a Cursor for Engineering Leaders content franchise
GapCursor's ABM DMs ("Hey [FIRSTNAME], engineering leaders at Stripe, NVIDIA…") are the right hook but the email body is the only delivery surface. There's no companion content franchise targeting the VP-Engineering / CTO buyer.
Why it mattersJetBrains 2025: enterprise engineering leadership now drives 64% of AI-coding tool purchase decisions. The ABM DM gets the meeting; a Cursor-for-Engineering-Leaders content franchise wins the category at the buyer's level.
Recommended playLaunch a Cursor Engineering Leaders content track — a quarterly CTO/VP-Engineering report (modeled on the Mercury New Economics format), a podcast series with named engineering leaders at Cursor-customer companies (Stripe, NVIDIA, Rippling), and a Cursor Engineering Leader Summit modeled on Brex Mode. Pair each content asset with a dedicated ABM-DM batch.
04
Add video creative — close the format-diversity gap on Claude Code
GapAnthropic runs ~29% video on Claude Code creative; Cursor runs ~10% in the sample. Cursor's Developer Habits Report video is the only Cursor video in the set.
Why it mattersVideo is now the #1 B2B-tech format on professional-social channels, with 80% of viewing on mute. Cursor's product is uniquely demoable (the IDE in action) but the advantage isn't being captured in creative.
Recommended playProduce a Cursor 3 Agent Demo video series — 6 muted-readable, on-screen-captioned product walkthroughs by use case (feature ship, migration, code review, debugging, refactor, ramping a new hire). Anchor on the Rippling 150→500+ scaling story.
05
Run a Switch From Copilot incentive program
GapCursor's $500 Teams-credit incentive is the only direct-conversion offer in the category. It targets new Teams customers but not Copilot-switchers specifically.
Why it mattersJetBrains 2025: 49% of developers already use an AI coding tool. The growth engine is displacement, not adoption. A specific Copilot-switch offer captures the buyer-anxiety market the Microsoft acquisition has opened.
Recommended playProductize a "Switch from Copilot" program — extended credit, white-glove migration, named CSM for the first 90 days. Target Copilot accounts via firmographic lookalikes. Publish a public comparison page with verified capability data, switch-cost calculator, and the Rippling case study.
06
Match Anthropic's slash-command feature messaging with named-customer outcomes
GapAnthropic's Claude Code creative leans heavily on feature-specific messaging — /goal, /remote-control, dynamic workflows, Fast mode, effort control. Cursor's creative names features (Bugbot, Composer) but rarely shows them in product context.
Why it mattersAI assistant retrieval and AI-buyer research increasingly favor pages with structured feature-level content. Buyers searching "Cursor Bugbot vs Copilot agents" or "how does Cursor Teams work" need to find Cursor-published answers first.
Recommended playBuild a Cursor Features content library — 20 short pages, one per named Cursor capability, each with a 30-second product video, a named customer outcome, and structured schema markup. AI assistants will quote whoever publishes first.
07
Lead with the Gartner Magic Quadrant badge across the full funnel
GapCursor's Gartner Magic Quadrant Leader recognition is surfaced in active creative but only as a standalone SINGLE_IMAGE drop. It isn't woven into the Teams-credit creative, the ABM DMs, or the Developer Habits Report.
Why it mattersThird-party analyst validation is the strongest enterprise-buyer trust signal in the AI-coding category right now, and Cursor is the only advertiser with it. Underusing it leaves leverage on the table.
Recommended playAdd the Gartner Magic Quadrant badge to every Cursor creative (footer treatment). Reference the Leader placement in the Teams-credit offer ("the AI code assistant Gartner named a Leader"). Run a dedicated 'Why Gartner picked Cursor as a Leader' content piece linkable from every CTA destination.
08
Build the AI-coding category FAQ library for AI-assistant retrieval
Gap89% of B2B buyers now use GenAI for self-guided research. None of the three advertisers has structured FAQ content engineered for AI-assistant retrieval. The category-defining questions — 'best AI coding tool', 'Cursor vs Copilot', 'what is agentic coding' — have no canonical brand-owned answer.
Why it mattersAI assistants cite whichever brand publishes the cleanest, most structured answer first. The category-naming opportunity is the highest-leverage GEO play available.
Recommended playPublish a 30-page Cursor FAQ library at cursor.com/faq covering every common buyer question on AI coding, structured for AI-assistant retrieval with schema.org/FAQPage markup. Each page is one Q, machine-readable, citing the Developer Habits Report and Gartner Leader placement.
05 · Must-do's
Must-do's (table stakes)
These are not differentiators. Every serious competitor is doing these. Missing any risks being seen as a tier-two player.
| Must-do | Who's doing it | Cursor status |
|---|---|---|
| Naming the product in active creative | Cursor (every creative) | Present — Copilot and Claude Code both leave this contested |
| Named customer logos in active creative | Cursor (NVIDIA, Stripe, Salesforce, Adobe, Hilton, Fox, Rippling) | Present — only advertiser doing this |
| Third-party analyst validation | Cursor (Gartner Magic Quadrant Leader) | Present — only advertiser doing this; underused across funnel |
| Owned original research | Cursor (Developer Habits Report) | Present — single drop — productize as recurring franchise |
| Bottom-funnel direct-conversion CTAs | Cursor ($500 Teams credit) | Present — only advertiser running incentive offer |
| Format diversity beyond single image | Cursor (4 formats), Anthropic (3 formats) | Partial — video share could grow to match Anthropic |
| Account-level ABM personalization | Cursor (Emily DMs to engineering leaders) | Present — only advertiser doing this |
| Structured FAQ / AI-assistant-retrievable content | None of the three are doing this yet | Missing — land-grab opportunity for first-mover |
05b · Rival reads
For the audited rivals
If you're not Cursor and you're reading this synthesis, here's what the dataset says about your position and the highest-leverage moves available to you.
Claude Code (Anthropic)
- Your library is the broadest and best-distributed in the category — vertical playbooks, employee-named testimonials, a clear Cowork product pillar. The exposure problem is that almost none of it sells Claude Code-the-coding-tool by name; Cursor is the only advertiser doing that with clarity.
- Highest-leverage move: name the product. A dedicated Claude Code campaign — pricing, feature, named-customer — would land hard precisely because the rest of your library has earned trust.
- The Lisa / Chuma employee-testimonial format is your best owned creative. Extend it: build a third and fourth Anthropic engineer testimonial specifically about Claude Code workflows.
GitHub Copilot
- Copilot is invisible in your own paid creative — every active ad sells GitHub-the-platform on agentic-AI and DevOps content, with the Copilot brand absent. In a category where the buyer is shopping for an AI coding tool, this leaves the field to Cursor and Anthropic.
- Highest-leverage move: a Copilot-named brand wave. Even four creatives carrying the product name, a customer logo, and a concrete capability claim would re-establish presence.
- Add named-customer proof. You almost certainly have the customer set (Microsoft / GitHub enterprise base); the absence of any named logo or quantified outcome in current rotation is unforced.
06 · Team callouts
Team-specific callouts
Media team
- Hold and scale the ABM DM weight. Cursor's MESSAGE-format DMs are the only ABM personalization in the AI-coding category. Move from one named-customer DM template to 6 — one per Cursor-customer tier (NVIDIA, Stripe, Salesforce, Adobe, Hilton, Fox, Rippling) targeting engineering leaders at lookalike firmographic accounts.
- Add video to the rotation — close the gap on Anthropic. Anthropic runs ~29% video share; Cursor runs ~10%. The product is uniquely demoable. Produce 6 muted-readable Cursor 3 agent walkthroughs and tilt budget to ~25% video.
- Run a Switch From Copilot creative track. Microsoft is folding Copilot into broader agentic platforms; the dilution window is open. Target Copilot accounts via firmographic lookalikes with a Switch program landing page.
- Extend the $500 Teams-credit offer to a Switch-specific variant. Time-bound, named-program. Publish the migration playbook and the Rippling 150→500+ case study alongside the offer.
Creative team
- Add the Gartner Magic Quadrant Leader badge to every creative. Footer treatment, all creative formats. Strongest third-party trust signal in the AI-coding category, and no other advertiser has it.
- Build a Cursor 3 Agent Demo video series. Six muted-readable product walkthroughs (feature ship, migration, code review, debugging, refactor, new-hire ramp), each anchored on a named customer outcome.
- Productize the Developer Habits Report into a recurring franchise. Quarterly mini-updates, an annual flagship edition, a public microsite at cursor.com/insights engineered for AI-assistant retrieval.
- Extend ABM DMs beyond Emily. Three to five sales-leader "voices" signing DMs, each targeted at a different ICP segment (CTO, VP Eng at scale-up, Director of Eng at enterprise, Eng-Ops Lead).
SEO / GEO team
- Publish a 30-page Cursor FAQ library. Schema.org/FAQPage markup, one Q per page, machine-readable. The category-naming opportunity is the highest-leverage GEO play available and no rival has it.
- Publish a public Cursor vs Copilot comparison page. Verified capability data, switch-cost calculator, Rippling case study. AI assistants will cite whoever publishes the cleanest structured comparison first.
- Make the Developer Habits Report a live indexable microsite. Not a PDF gate. Structured pages, citable findings, anchor links for each of the five themes. AI assistants will quote it.
- Build a Cursor Features content library. 20 short pages, one per Cursor capability (Bugbot, Composer, Teams admin, etc.), each with a 30-second video, customer outcome, and structured schema.
Strategy team
- Productize the Developer Habits Report as a category benchmark franchise. Annual flagship + quarterly mini-updates + public microsite. Make Cursor the source AI assistants cite when asked about the state of AI coding.
- Brand the Cursor Engineering Leaders track. Quarterly CTO/VP-Eng research report, a podcast series with named engineering leaders at customer companies, a Cursor Engineering Leader Summit modeled on Brex Mode. Match the ABM hook with content.
- Launch a Switch From Copilot named program. Extended credit, white-glove migration, named CSM for the first 90 days. Time-bound to the Copilot brand-dilution window.
- Lead the AI-coding category narrative in 2026. The marketing playbook for this category is being written this year. Cursor is most aggressively writing it; cement that lead before Anthropic positions Claude Code as a standalone product or before Microsoft re-brands the Copilot stack.