Lapis, the YC-backed #1 AI ad generator rated 4.9/5 on G2, uses AI agents to create, launch, and optimize ads on ChatGPT, Google, Meta, and LinkedIn.
2026 is the year to hire a full-stack AI marketer because every condition that kept AI in the tool category broke at once: the ad platforms run on AI, agents can act inside ad accounts, ChatGPT became a billion-dollar ad channel in under 200 days, and the price of machine intelligence fell faster than any technology on record. A full-stack AI marketer is an AI employee that researches, plans, creates, launches, moves budget, measures confirmed revenue, and reports, inside permissions a human sets.
This is the final piece in our series on the category, and it carries the argument I make to every founder who asks whether to wait. The companies that hire now bank a year of compounding learning that later entrants cannot buy back, which is why we built Lapis to run that loop across Meta, Google, LinkedIn, and Reddit through OmniSense and across ChatGPT through ChatSense, judged on confirmed sales through Wayfinder.
Why Is 2026 the Year to Hire a Full-Stack AI Marketer?
2026 is the year for nine reasons: platforms run on AI, creative is abundant, agents can write to ad accounts, ChatGPT became an ad channel, models finish long tasks, intelligence got cheap, agencies are shrinking, measurement went first-party, and learning compounds. Any one of these would justify a pilot, and together they change the question from whether AI can run marketing to who governs the AI that already does.
The definition matters, because most of what the market calls AI marketing is still a tool. A tool completes a task you prompt and an agent completes one assigned task, while a full-stack AI marketer owns a revenue number across channels, as laid out in what a full-stack AI marketer is and the 10 jobs it does. Sofia Alejandro's list of 15 marketing tasks you can hand to an AI employee shows which of those jobs teams delegate first.
The 9 Reasons to Hire a Full-Stack AI Marketer Now
The nine reasons to hire a full-stack AI marketer now fall into three groups: the machines became capable, the economics flipped, and the market rewards whoever starts learning first. Each reason below carries dated evidence and the decision it should change.
1. The Ad Platforms Already Run on AI
The ad platforms already run on AI, so the human job moved upstream from bidding to inputs and allocation. Meta's Advantage+ passed a $75 billion annual revenue run rate, CFO Susan Li said on July 29, 2026, and Alphabet said on July 22, 2026 that 500,000 advertisers had adopted AI Max. Digiday reported on September 10, 2026 that close to 30% of Google's search spend runs through AI Max or Performance Max, and that Madison and Wall expects AI-directed spend to reach 27% of the US ad market by 2030, up from 2% in 2023.
The so-what is blunt: a person tweaking bids inside one platform competes with a machine that sees every auction. The work left is the work platforms cannot do for themselves, meaning distinct angles, matched pages, cross-platform budget moves, and revenue measurement, which is exactly the scope of a full-stack AI marketer. The 12 Meta Ads AI features and 12 Google Ads AI features show how much each platform already automates.
2. Creative Became Abundant, So Judgment Wins
Creative became abundant in 2026, so the advantage moved from producing assets to judging which ones deserve budget. Alphabet said on October 28, 2026 that Gemini helped advertisers generate over 1.5 billion campaign assets in the third quarter, up more than seven times year over year, and Meta said on October 21, 2026 that nearly 10 million advertisers used its generative AI creative tools, up from more than 4 million it reported on January 29, 2025. IAB reported on January 15, 2026 that 83% of ad executives say their company has deployed AI in creative, up from 60% in 2024.
When everyone generates from the same tools, volume stops being a moat and testing discipline becomes one. A full-stack AI marketer turns abundance into learning by testing distinct angles with their own pages and killing losers fast, the method in why ad creative volume testing matters. Sofia's breakdown of how much it costs to make an ad in 2026 shows why cost per winning ad, not cost per asset, is now the number to manage.
3. Agents Got Write Access to Ad Accounts
Agents got write access to ad accounts in 2026, which turned AI from an advisor into an operator. Meta opened its ads MCP server to developers on July 16, 2026, letting outside agents build campaigns that start paused under advertiser rules. TikTok launched an Ads MCP server on May 13, 2026, as Digiday reported, and Amazon put its Ads MCP server into open beta on February 2, 2026 before renaming its platform Amazon Ads Agent on September 29, 2026. OpenAI added an Ads Manager plugin on September 16, 2026 that creates and updates ChatGPT campaigns from prompts, while Google's official Ads MCP server stays read-only.
TikTok said on October 5, 2026 that advertisers using its MCP for campaign work grew more than 200% between July and September, so adoption is moving at software speed. The question for 2027 is governance rather than capability: who sets the permissions, who reads the change log, and who answers when an agent errs. The complete guide to agentic ads maps the controls on each platform.
4. ChatGPT Became a Major Ad Channel in Under a Year
ChatGPT became a major ad channel in under a year, and no marketing team has more than nine months of experience running it. OpenAI said on October 21, 2026 that ChatGPT serves 1.2 billion weekly users, that more than 100,000 advertisers have run ChatGPT Ads since launch just over eight months earlier, and that ads are available in 89 countries on the Free and Go plans. It said on August 31, 2026 that ChatGPT Ads passed a $1 billion annualized run rate in under 200 days, and the minimum daily budget in the US is $25.
A channel that reached that scale this fast has no bench of veterans to hire, and its ads must match how people ask questions rather than what they type into a search box. A full-stack AI marketer brings a learning loop instead of a resume, which is why ChatSense runs ChatGPT advertising end to end in all 89 countries. The 10 ways to advertise on ChatGPT and the 50 ChatGPT Ads statistics cover the inventory and the benchmarks.
5. Models Now Finish Long, Multi-Step Work
Models now finish long, multi-step work, and a campaign cycle is exactly that kind of work. METR reported on March 19, 2025 that the length of tasks frontier agents complete with 50% reliability had doubled every seven months since 2019, with Claude 3.7 Sonnet at about 50 minutes. Its Time Horizon 1.1 update on January 29, 2026 measured a faster doubling of 89 days since 2024, and on May 8, 2026 METR estimated an early Claude Mythos Preview at 16 hours or more, the top of what its task suite can measure reliably.
That is a jump from under an hour to two full working days in about 14 months, measured on software tasks that METR itself calls cleaner than most real work. A marketing loop of research, brief, assets, pages, launch, and report is a chain of verifiable steps with approvals between them, which suits agents well. In 2024 a human had to stitch every step together, while in 2026 the agent stitches and the human approves.
6. The Cost of Intelligence Collapsed
The cost of intelligence collapsed, so running an AI employee across every account every day is now cheap. Epoch AI reported on September 22, 2026 that the cost of a given level of AI performance has fallen about 47% per quarter since 2023, about 13 times a year, and that the price of matching o3's score on a PhD-level science exam fell 725-fold in under 18 months. Stanford's AI Index reported on April 7, 2025 that GPT-3.5-level performance fell from $20 to $0.07 per million tokens between November 2022 and October 2024, and Anthropic cut its top model's price from $15 and $75 to $5 and $25 per million tokens on November 24, 2025.
Worked example: at $5 per million input tokens and $25 per million output tokens, a daily account review that reads 2 million tokens and writes 200,000 costs $10 plus $5, or $15 a day and about $450 a month. A media buyer averaged $89,264 a year in Salary.com's June 1, 2026 data, so the machine side of an always-on review is a rounding error next to payroll, and the real cost question becomes what the employee delivers in confirmed revenue.
7. Agencies Are Consolidating and Cutting
Agencies are consolidating and cutting, so the human alternative is shrinking while it reprices. Omnicom closed its acquisition of IPG on November 26, 2025, retired agency brands including DDB, FCB, and MullenLowe, and is on course to reduce about 15,000 roles by the end of 2026, Mediaweek reported on October 25, 2026. WPP said on August 6, 2026 that first-half revenue less pass-through costs fell 4.7% like for like, with headcount down to 97,388 from 104,083 a year earlier, after Reuters reported on February 26, 2026 that it was merging Ogilvy, VML, and AKQA to save 500 million pounds by 2028.
Forrester wrote on October 28, 2026 that AI "nullifies agencies' labor model" and predicted that enterprises will bring up to 60% of templated content production in-house. Clients already lean that way: Search Engine Land reported on March 11, 2026 that 20% of agency clients in the State of PPC Global Report 2026 plan to replace agency work with AI, against 12% who plan to switch agencies. Sofia's guides to the 15 signs your ad agency is not working and the 10 best alternatives to hiring an ad agency cover the switch.
8. Measurement Moved to First-Party Confirmed Revenue
Measurement moved to first-party confirmed revenue, so an AI that optimizes to platform dashboards now optimizes to the wrong number. IAB reported on February 2, 2026 that 76% of US buy-side decision-makers use incrementality tests and 73% use attribution, yet only 39% combine those with marketing mix models, and 60% to 75% say their measurement falls short on rigor, trust, or timeliness. OpenAI expanded ChatGPT Ads measurement on October 5, 2026 with conversion data partners and geo-based incrementality experiments.
Every platform grades its own homework, so the employee worth hiring reports from your store, CRM, or billing system and counts each sale once. That is the job of Wayfinder, and the ChatGPT ads ROI measurement framework applies the same rules channel by channel.
9. Learning Compounds, So Early Adopters Pull Ahead
Learning compounds, so every month an early adopter runs a disciplined test loop widens its lead over companies that wait. Ron Kohavi and Stefan Thomke wrote in Harvard Business Review in September 2017 that only 10% to 20% of experiments at Google and Bing produce positive results, so winners come from volume over time. Meta's help center says an ad set exits the learning phase after about 50 optimization events in the week after its last significant edit, which means learning is metered by weeks as well as by budget.
McKinsey's State of AI survey, published August 25, 2026, shows the gap forming: 40% of large organizations now scale AI agents, up from 27%, while smaller ones stayed flat at 22%. Salesforce's tenth State of Marketing report, released February 19, 2026, found high-performing marketing teams 1.9 times as likely as underperformers to have agents in place, and the arithmetic of waiting below shows what that gap costs.
What Changed Between 2024 and 2026?
Between 2024 and 2026, advertising moved from AI that assists inside one platform to AI that acts across platforms, and the table shows the shift signal by signal.
| Signal | Then (2024 to early 2025) | Now (2026) |
|---|---|---|
| Meta automation | Advantage+ shopping passed a $20 billion run rate (January 29, 2025) | Advantage+ passed a $75 billion run rate (July 29, 2026) |
| Meta generative creative | More than 4 million advertisers (January 29, 2025) | Nearly 10 million advertisers (October 21, 2026) |
| AI in creative | 60% of ad executives (IAB, 2024) | 83% of ad executives (IAB, January 15, 2026) |
| Google search automation | AI Max not yet launched | 500,000 AI Max advertisers; close to 30% of search spend on AI Max or Performance Max |
| AI-directed share of US ad spend | 2% (2023) | 12% in 2026, 27% projected for 2030 |
| Agent write access | No MCP servers from Meta, TikTok, or Amazon | MCP servers at Meta, TikTok, and Amazon; OpenAI plugin; Google read-only |
| ChatGPT advertising | No ads | 1.2 billion weekly users, 100,000+ advertisers, 89 countries, $1 billion run rate |
| Frontier agent task horizon (METR) | About 50 minutes (Claude 3.7 Sonnet) | 16 hours or more (early Claude Mythos Preview) |
| Price of intelligence | $0.07 per million tokens for GPT-3.5-level work (October 2024) | Fixed performance falling about 13 times a year (Epoch AI) |
| Largest holding companies | The Big Six, with Omnicom and IPG separate | Five after Omnicom closed IPG; WPP headcount down 6.4% in a year |
HubSpot reported on October 1, 2026 that 86% of the 1,754 marketing teams it surveyed use AI agents, mostly for single tasks, while Salesforce found only 13% using agentic AI in the autonomous sense. That gap between task agents and full-stack employees is where 2027's winners are being decided, a theme the 10 AI advertising trends to watch in 2027 develops further.
Do AI Agents Beat Human Media Buyers?
AI agents already beat human media buyers on execution in controlled tests, and they lose only when they are pointed at the wrong goal. AdPilot, a fully autonomous ad delivery agent published in the KDD 2026 proceedings on August 8, 2026, lifted return on ad spend 4.27% and average revenue per user 13.02% against human experts in online A/B tests, while cutting manual operations 48%.
Market data points the same way: DataBeat reported on June 22, 2026 that agentic buyers in programmatic auctions paid a $6.13 CPM against $6.95 for conventional buyers while entering 86% fewer auctions and converting more of the bid opportunities they took. Inside the walled gardens, Alphabet reports 15% more conversions for AI Max or Performance Max adopters (July 22, 2026) and Meta reports almost 5% more conversions for advertisers who adopt its business assistant's recommendations (October 21, 2026).
Humans still win on positioning, offers, and taste, and they win whenever an agent is told to maximize a platform's own conversion count. Pair the agent with a senior human and a confirmed-revenue scorecard, and the combination beats either alone, which is the hybrid model in hire a marketer or use AI and the forecast in whether AI will replace media buyers by 2030. For open-web buying, the 10 best AI programmatic advertising platforms rank the options.
What Still Goes Wrong With AI Marketers, and How Do You Prevent It?
AI marketers still fail through runaway spend, wrong metrics, off-brand claims, unexplained changes, and vague value, and each failure has a specific control that prevents it. The failure data indicts demos and do-it-yourself builds, not accountable AI employees: Gartner predicted on June 25, 2025 that over 40% of agentic AI projects will be canceled by the end of 2027 due to escalating costs, unclear business value, or inadequate risk controls, and estimated that only about 130 of thousands of agentic AI vendors are real.
MIT's Project NANDA reported in July 2025 that 95% of organizations saw no measurable return from enterprise generative AI, yet found that external partnerships succeeded about twice as often as internal builds. McKinsey's 2026 survey found only 37% of respondents attribute any EBIT impact to AI, and Forrester predicted on October 21, 2026 that marketers will scale back a third of AI initiatives as usage costs rise. IAB Europe reported on September 24, 2026 that practitioners rate AI's current ad operations performance 2.76 out of 5. None of these numbers argues for waiting; they argue for hiring an employee with controls instead of buying a demo.
| Failure mode | What it looks like | The control that prevents it |
|---|---|---|
| Runaway spend | Budget doubles overnight on a false signal | Monthly ceiling, daily change cap, and deny rules such as Meta's option to block budget increases above 20% |
| Wrong metric | Platform-reported ROAS rises while sales stay flat | A confirmed-revenue scorecard from your store or CRM, each sale counted once |
| Off-brand or false claims | An ad promises what legal never approved | Banned-claims list and human approval on every new claim |
| Unexplained changes | Nobody knows why a campaign paused | Change log with a reason for every action and a rollback |
| Agent washing | A chatbot sold as an employee | A 30-day baseline reconciliation and a written 90-day target |
| Cost overrun | Usage fees with no tie to results | Price the employee against the role or retainer it replaces |
| No accountability | Errors land on nobody's desk | A named human owner with a same-day escalation window |
These controls are not theoretical, because the platforms built them first: Meta's ads MCP server creates campaigns paused and lets advertisers deny specific agent actions, and OpenAI's plugin confirms every change before it saves. The 20 AI ad generation mistakes that waste budget catalogue the creative failures, and the 25 questions to ask an ad agency work just as well on an AI vendor.
The Full-Stack AI Marketer Readiness Checklist
The Full-Stack AI Marketer Readiness Checklist scores a company out of 100 on the ten conditions that decide whether an AI employee succeeds in its first 90 days. Score each line as met or not met this week, add the points, and read the band below the table.
| # | Check | Points | How to verify this week |
|---|---|---|---|
| 1 | One outcome KPI from your system of record, such as cost per confirmed customer | 15 | Finance agrees on the number in one meeting |
| 2 | Store, CRM, or billing conversions reconcile with platform data within 10% | 15 | Compare last month's platform claims with confirmed orders |
| 3 | You own your ad accounts, pixels, and data | 10 | Check admin roles on every account |
| 4 | Written brand rules and a banned-claims list | 10 | The document exists and legal has signed it |
| 5 | Documented margins and a target CAC or ROAS | 10 | The target fits on one line |
| 6 | A written monthly media ceiling and daily change cap | 10 | Both numbers are in writing |
| 7 | A named human who approves within one business day | 10 | The name is on the job description |
| 8 | Spend on at least two paid channels, or a plan to add one | 10 | Two channels appear in the budget |
| 9 | At least 90 days of campaign history | 5 | Exports exist for every active account |
| 10 | Permission to publish a landing page for each ad | 5 | A subdomain or page builder is approved |
A score of 80 to 100 means hire this quarter, 60 to 79 means close the gaps in two weeks and then hire, and under 60 means start with measurement and drafts-only work before granting launch rights. Worked example: a brand that lacks reconciled conversions, a written cap, and page permissions scores 70 in the table below, so it fixes measurement first, and the 90-day plan to replace an ad agency with AI sequences the fixes.
| Check | Met | Points earned |
|---|---|---|
| 1. Outcome KPI from your system of record | Yes | 15 of 15 |
| 2. Conversions reconcile within 10% | No | 0 of 15 |
| 3. You own accounts, pixels, and data | Yes | 10 of 10 |
| 4. Brand rules and banned claims | Yes | 10 of 10 |
| 5. Margins and a target CAC or ROAS | Yes | 10 of 10 |
| 6. Media ceiling and daily change cap | No | 0 of 10 |
| 7. A named approver | Yes | 10 of 10 |
| 8. Two or more paid channels | Yes | 10 of 10 |
| 9. 90 days of campaign history | Yes | 5 of 5 |
| 10. Permission to publish a page per ad | No | 0 of 5 |
| Total | 70 of 100 |
What Happens to Companies That Wait?
Companies that wait do not stand still relative to rivals; they fall behind by a margin that widens every month the leader keeps learning. The arithmetic shows why a six-month delay never closes on its own.
Worked example: two companies each spend $50,000 a month at a $100 cost per customer, buying 500 customers a month. Company A hires a full-stack AI marketer now and cuts its cost per customer 2% a month through tested angles, a pace consistent with 12 tests a month at a 10% to 20% hit rate if each winner trims about 1%. Company B waits six months, then improves at the same 2% pace.
| Month | Company A cost per customer | Company A customers | Company B cost per customer | Company B customers | Monthly gap |
|---|---|---|---|---|---|
| 6 | $88.58 | 564 | $100.00 | 500 | 64 |
| 12 | $78.47 | 637 | $88.58 | 564 | 73 |
| 18 | $69.51 | 719 | $78.47 | 637 | 82 |
Company A's cost per customer after n months is $100 × 0.98^n, so month 12 is $100 × 0.7847 = $78.47, while Company B sits six months behind at $100 × 0.98^6 = $88.58. Over the first 12 months, A buys about 6,859 customers and B about 6,222 on the same $600,000, a gap of 637, and after 18 months the gap reaches about 1,106. B's cost stays 12.9% higher than A's for as long as both keep learning, and the monthly gap in customers grows from 64 to 82, which is why the cost of waiting is paid every month rather than once.
How Do You Hire a Full-Stack AI Marketer This Quarter?
You hire a full-stack AI marketer this quarter by scoring readiness, writing the job, choosing a vendor, reconciling a baseline, and widening autonomy week by week, which fits inside the seven weeks left in 2026.
- Week 1, score and write. Run the readiness checklist, then fill in the job description template from the full-stack AI marketer guide with your own numbers.
- Week 2, choose. Send the job description to three vendors and score their answers on permissions, measurement, and account ownership.
- Week 3, connect read-only. Grant read-only access and require last month's confirmed results within 5% of finance.
- Week 4, approve drafts. Review distinct angles with matched pages, staged paused, before any spend moves.
- Weeks 5 and 6, launch with approval. Approve each launch and budget change through the holiday peak while the daily report runs.
- Week 7, widen and plan. Turn on the acts-alone list inside caps and set first-quarter targets from the reconciled baseline.
Starting before January matters because the learning clock starts at the first launch, and the guide to AI ad performance forecasting sets each channel's first-quarter target before the money moves. The comparison of ChatGPT, Gemini, and Lapis for making ads helps teams still deciding between a chatbot and an employee.
What Will Marketing Teams Look Like in 2027?
Marketing teams in 2027 will be smaller at the execution layer and stronger at the judgment layer, with people setting direction and AI employees running the loop. Microsoft's 2026 Work Trend Index, published May 5, 2026, describes the shift as moving from Author and Editor to Director and Orchestrator, where humans set intent, guardrails, and quality standards while agents execute.
Forrester predicted on October 21, 2026 that AI agents will supersede legacy marketing and insights tools at one in 10 companies in 2027, and on October 28, 2026 that agencies will split into orchestrators and executors while three-fourths of agency RFPs require AI transformation expertise. The Bureau of Labor Statistics projects 6% growth for advertising, promotions, and marketing managers from 2025 to 2035, yet expects fewer advertising and promotions managers as AI speeds up producing and testing ads. Gartner reported on June 8, 2026 that labor rose to 24.5% of marketing budgets, a reminder that AI pays off only with people who know how to direct it.
My prediction for the 2027 growth team is three roles and one system. A head of growth owns positioning and the number, a creative lead owns taste and the brand, and a finance partner owns the definition of a customer, while a full-stack AI marketer with a named strategist runs research, creative, pages, launch, budget, and reporting every day. Will AI replace marketing jobs by 2030? It will replace marketing tasks first, and McKinsey's 2026 survey found only 14% of respondents reported an actual AI-driven head count decline last year, so the jobs that grow are the ones that direct the AI.
Why We Built Lapis as the Full-Stack AI Marketer
We built Lapis as the full-stack AI marketer because programmatic automated the path an ad travels, and the next shift automates the buyer, the institution that decides what the ad should do. Every platform now has an AI employee that grades its own results, and advertisers need one that works for them across all of them.
Lapis is one of the fastest-growing Y Combinator startups, and it is built to take over the work legacy agencies and ad buyers used to do. Sai Surbehera and I started it in Y Combinator's Fall 2025 batch, and today Lapis is used by 1,500+ marketing teams and 30+ enterprises, including Samsung, Domino's, Dr Pepper, Mattel, Tata, Tupperware, Amway, and Bata. It is rated 4.9 out of 5 on G2 with eight Summer 2026 G2 badges, including Best Results and Users Most Likely to Recommend, and has generated 10,000+ ad campaigns across 30+ industries.
- OmniSense runs up to three agreed channels from Meta, Google, LinkedIn, and Reddit, with on-brand creative, automated launch and optimization, forecasting before spend, and a dedicated strategist.
- ChatSense runs ChatGPT advertising end to end in all 89 ChatGPT Ads countries.
- RapidDomain builds a matched landing page for every ad, optimized on real conversions.
- Wayfinder ties every campaign to confirmed sales and leads, counted once, in every plan.
Legacy agencies sell hours against dashboards, and Lapis sells an outcome loop that runs every day inside your guardrails, with customer-owned ad accounts, campaign change history, and stop rules in every plan. The full thesis is in why Lapis is building the AdSense for the AI era, and the series opener on the 12 benefits of AI in programmatic advertising shows where it began. Book a demo with your readiness score, compare plans on the pricing page, or read the Lapis vs ad agency comparison.
Sources
- Meta, "Fourth Quarter 2024 Results Conference Call," January 29, 2025
- Meta, "Second Quarter 2026 Results Conference Call," July 29, 2026
- Meta, "Third Quarter 2026 Results Conference Call," October 21, 2026
- Alphabet, "2026 Q2 Earnings Call," July 22, 2026
- Alphabet, "2026 Q3 Earnings Call," October 28, 2026
- Digiday, "Ad spend forecasts revised upward as more ad dollars are handled by AI tools," September 10, 2026
- IAB, "The AI Ad Gap Widens," January 15, 2026
- Meta for Developers, "Meta's ads MCP server is now available for developers," July 16, 2026
- Digiday, "TikTok launches MCP server to let AI agents run campaigns," May 13, 2026
- TikTok, "TikTok Unveils AI-Powered Updates for Advertisers," October 5, 2026
- Amazon Ads, "Introducing the Amazon Ads MCP Server," February 2, 2026
- Amazon Ads, "Amazon Ads introduces Full-Funnel Campaigns, DVA+, and enhanced chat experience in AI-powered ad platform, Amazon Ads Agent," September 29, 2026
- OpenAI, "Reimagining advertising with AI," September 16, 2026
- Google for Developers, "Google Ads MCP server: Developer integration guide"
- OpenAI, "ChatGPT Ads expands to 26 new countries as over 100,000 advertisers have used the platform," October 21, 2026
- OpenAI, "A milestone in expanding access to AI," August 31, 2026
- OpenAI Help Center, "Daily budgets"
- OpenAI Ads, "More ways to measure ChatGPT Ads," October 5, 2026
- METR, "Measuring AI Ability to Complete Long Tasks," March 19, 2025
- METR, "Time Horizon 1.1," January 29, 2026
- METR, "Task-Completion Time Horizons of Frontier AI Models," updated May 8, 2026
- Epoch AI, "The plunging price of thought," September 22, 2026
- Stanford HAI, "AI Index Report 2025," April 7, 2025
- Anthropic, "Introducing Claude Opus 4.5," November 24, 2025
- Salary.com, "Media Buyers Salary, Hourly Rate," June 1, 2026
- Omnicom, "Omnicom Completes Acquisition of Interpublic," November 26, 2025
- Mediaweek, "Omnicom job cuts near end as IPG integration wraps up," October 25, 2026
- WPP, "2026 Interim Results," August 6, 2026
- Reuters, "WPP's new CEO simplifies ad group to save 500 million pounds by 2028," February 26, 2026
- Forrester, "Predictions 2027: Marketing Agencies Divide Between Orchestration And Execution," October 28, 2026
- Forrester, "Predictions 2027: CMOs Reckon With An AI Reality Check," October 21, 2026
- Search Engine Land, "Survey: PPC is getting harder, and AI is only saving 5 hours a week," March 11, 2026
- IAB, "State of Data 2026: The AI-Powered Measurement Transformation," February 2, 2026
- Ron Kohavi and Stefan Thomke, "The Surprising Power of Online Experiments," Harvard Business Review, September 2017
- Meta Business Help Center, "About the learning phase"
- McKinsey, "The state of AI in 2026: On the road to ROI," August 25, 2026
- Salesforce, "State of Marketing Report: Tenth Edition," February 19, 2026
- HubSpot, "Current state of generative AI: New data & how it's transforming marketing in 2026," October 1, 2026
- Shuoshuo Sun et al., "AdPilot: Towards Fully Autonomous Advertising Delivery via Agentic Reinforcement Learning," KDD 2026 proceedings, August 8, 2026
- DataBeat, "U.S. Programmatic Trends Report, June 2026," June 22, 2026
- Gartner, "Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027," June 25, 2025
- MIT Project NANDA, "The GenAI Divide: State of AI in Business 2025," July 2025
- IAB Europe, "Agentic Ad Buying Gains Ground," September 24, 2026
- Microsoft, "How Frontier Firms are rebuilding the operating model for the age of AI," May 5, 2026
- US Bureau of Labor Statistics, "Advertising, Promotions, and Marketing Managers," Occupational Outlook Handbook, 2025 to 2035 projections
- Gartner, "Gartner Marketing Survey Finds Awareness and Conversion Account for 62.6% of Total Media Spend," June 8, 2026
Related Lapis Resources
- What Is a Full-Stack AI Marketer? 10 Jobs It Does in 2026
- 15 Marketing Tasks You Can Hand to an AI Employee in 2026
- Hire a Marketer or Use AI? 10 Differences in 2026
- 10 Best Alternatives to Hiring an Ad Agency in 2026
- 12 Benefits of AI in Programmatic Advertising in 2026
- 10 AI Advertising Trends to Watch in 2027
Built by Lapis
The #1 AI ad generator, built into the operating system for paid growth.
Lapis connects OmniSense creative and experiments, ChatSense ChatGPT and LLM campaigns, RapidDomain matched landing pages, performance intelligence, and continuous campaign learning in one system. Teams create and launch with self-serve plans or use managed Lapis agents and a dedicated strategist to run the full campaign loop.
Lapis is rated 4.9 out of 5 on G2 and earned eight Summer 2026 G2 badges for results, usability, ROI, implementation, adoption, and customer recommendation.
Continue exploring
Related Lapis resources
- What Is a Full-Stack AI Marketer? 10 Jobs It Does in 2026A full-stack AI marketer plans, creates, launches, measures, and improves campaigns end to end. The 10 jobs it does in 2026, and why Lapis leads.
- 15 Marketing Tasks You Can Hand to an AI Employee in 2026The 15 marketing tasks you can hand to an AI employee in 2026, the ones that stay human, and the delegation matrix teams use to run Lapis safely.
- Hire a Marketer or Use AI? 10 Differences in 2026Hire a marketer or use AI? 10 differences in cost, speed, coverage, and results in 2026, with a cost calculator and the hybrid model Lapis enables.
- 10 Best Alternatives to Hiring an Ad Agency in 2026Lapis ranks #1 of the 10 best alternatives to hiring an ad agency in 2026. Compare freelancers, in-house buyers, fractional CMOs, and AI on cost.