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.
Can AI Replace an Ad Agency in 2026?
Yes, for the work that fills most retainers. Creating ads, resizing them for every placement, building campaigns, running tests, moving budget, and reporting results are now jobs AI does faster, cheaper, and at higher volume than a team billing hours. What still belongs to people is the narrow layer of senior judgment: the brand platform, the big creative idea, the pricing and positioning call.
The agencies themselves are restructuring around that fact. WPP's Elevate28 plan, announced February 26, 2026, targets GBP 500 million of gross annual cost savings by 2028 and collapses the group into four operating units. Dentsu announced on August 14, 2025 that it would cut about 3,400 roles, roughly 8% of its international workforce. Omnicom completed its acquisition of IPG on November 26, 2025, then announced about 4,000 job cuts and retired the DDB, FCB, and MullenLowe brands.
McKinsey's June 2026 report on the agentic advertising economy names agencies “among the most exposed players”: roughly half of media spend already flows through direct channels, and planning, buying, reporting, and creative production are exactly the knowledge work AI disrupts fastest. The buyer's question is no longer whether AI can do agency work. It is which work to hand to AI, which to keep in-house, and which still earns a specialist. For the structural history, read why legacy ad buyers are becoming obsolete.
Which Agency Tasks Can AI Replace Today?
AI replaces creative production, resizing and localization, campaign setup, test design, budget pacing, competitor research, landing pages, and reporting. The ad platforms now ship these capabilities themselves. OpenAI's Ads Manager suggests ad copy and imagery from your landing page and, through the ChatGPT Ads Manager plugin, lets advertisers create, update, and analyze campaigns in plain language. Google Performance Max generates headlines, descriptions, images, and logos from your website. Meta began testing an Ads Manager assistant in October 2026 that plans campaigns, generates images and copy, and changes budgets with advertiser approval.
| Agency function | What AI does in 2026 | Where Lapis runs it |
|---|---|---|
| Creative production | Brand-trained ads with a distinct visual per angle, in every size | OmniSense and Campaign Studio |
| Media setup and launch | Campaign structure, audiences, and launch across channels | OmniSense and ChatSense, managed |
| Testing and optimization | Experiment design, next-run generation, budget moves inside guardrails | Managed Traditional Ads and Managed ChatGPT + LLM Ads |
| Landing pages | A matched page for every ad, keyword, prompt, and audience | RapidDomain, managed |
| Competitive intelligence | Continuous competitor ad tracking | Brand Radar and its 10,000+ ad library |
| Reporting and attribution | Confirmed conversions, counted once, compared across channels | Wayfinder |
| Senior strategy and taste | Human judgment, sharpened by AI research and forecasts | Your team plus a dedicated Lapis strategist |
Read the table bottom to top and the shift is obvious. The single row that stays human is the one agencies bill the least for. The six rows above it are the hours that fill a retainer, and every one of them now runs on software.
What Does an Ad Agency Retainer Actually Pay For?
A retainer pays for human hours: account management, creative production, media operations, and reporting, billed on top of your media spend. Clutch's advertising agency pricing guide, updated September 21, 2026 from verified client reviews, puts common US agency rates at $100 to $149 an hour, most reviewed advertising projects at $10,000 to $49,999, and the average monthly cost of reviewed engagements at $9,890.93. Clutch also notes that agency fees sit separately from the cost of running the ads, and that some agencies charge a percentage of media spend instead.
| Retainer line item | With an agency | With Lapis |
|---|---|---|
| Account management | Billed meetings, briefs, and status calls | Plain-English direction in Campaign Studio; a dedicated strategist on managed plans |
| Creative production | Designer and copywriter hours per asset, plus revision rounds | Brand intelligence from your URL, on-brand variations on demand |
| Media operations | Trafficking, bid changes, and budget pacing by hand | Autonomous deployment and budget reallocation inside your guardrails |
| Landing pages | A separate web or CRO project | A RapidDomain page for every ad and audience |
| Reporting | A monthly deck of platform-reported numbers | Daily reporting and Wayfinder confirmed-conversion measurement |
| Fee model | Hourly, retainer, project, or percent of media | $99 Basic subscription or a managed quote |
For every fee model, a scenario calculator, and the questions to put in an agency RFP, read How Much Does an Ad Agency Cost in 2026?
How Much Cheaper Is AI Than an Agency or a Freelancer?
An hour of agency time costs more than a month of Lapis Basic. That one comparison captures the economics: Clutch's common US agency rate is $100 to $149 per hour, and Lapis Basic is $99 per month for 25 monthly credits, about $4 per credit, with Google, Meta, and LinkedIn creative, performance forecasting, and 20 tracked competitors included.
$99 vs. $100 to $149
a full month of Lapis Basic versus one hour of US advertising agency time
| Option | Published price | What the price covers |
|---|---|---|
| Full-service ad agency | $100 to $149 per hour; $9,890.93 average monthly engagement (Clutch) | Strategy, creative, media, and reporting, billed by the hour or retainer |
| Design subscription | From $15,000 per month on an annual term, plus a $1,000 monthly software fee (Superside) | Creative production; media buying and attribution stay with you |
| Freelance designer | $15 to $35 per hour typical (Upwork) | Design hours; you brief, launch, test, and report |
| Lapis Basic | $99 per month for 25 monthly credits | Self-serve, brand-trained creative for Google, Meta, and LinkedIn, forecasting, Brand Radar, audiences, and 100+ languages |
| Lapis managed | Custom quote | Agents and a dedicated strategist run creative, launch, optimization, RapidDomain pages, and daily reporting end to end |
Worked example: a DTC skincare brand needs 30 new static creatives a month across Meta, Google, and LinkedIn. At three billed hours per creative for the brief, design, copy, one revision round, and resizing, that is 90 agency hours, or $9,000 to $13,410 a month at Clutch's common rate, before anyone touches media or reporting. A freelancer at Upwork's typical $15 to $35 an hour costs $1,350 to $3,150 for the same hours, and the brand's own team still writes every brief, launches every campaign, and builds every report. With Lapis, the brand's website trains the creative once and every variation stays on brand. Basic puts that engine in the team's hands for $99 a month. A managed plan hands the whole job, production plus launch, optimization, landing pages, and daily reporting, to Lapis agents and a dedicated strategist under one quote.
Price per asset is only half the gap. Performance advertising is a test-and-learn loop, and the team that ships more distinct ideas finds winners sooner and beats fatigue longer. That matters most in ChatGPT, where OpenAI selects ads through a relevance-weighted, second-price auction that weighs the conversation, the ad's title and copy, the landing page, and the advertiser's context hints. A creative matched to each buyer intent wins that auction. Lapis Managed ChatGPT + LLM Ads ships 1,000+ ChatGPT ads for exactly that reason, a volume no hourly model can price.
Were Brands Already Moving Agency Work In-House Before AI?
Yes. In-housing is a fifteen-year trend, and AI removes its last constraint: capacity. The Association of National Advertisers' 2023 study, The Continued Rise of the In-House Agency, surveyed 162 ANA members in February and March 2023. 82% had an in-house agency, up from 42% in 2008, 58% in 2013, and 78% in 2018. 92% still worked with external agencies, and among those, an average of 61% of all the work for their company was done in-house. 65% had moved established business from an external agency to the in-house team in the previous three years, and 54% of in-house agencies handled some media planning or buying.
82%
of ANA members surveyed in 2023 had an in-house agency, up from 42% in 2008
The ANA's 2026 State of In-Housing report, published June 22, 2026, asked a different group a different question. Its respondents were 404 jurors of the 2026 ANA In-House Excellence Awards: brand-side marketers, agency professionals, and independent consultants who run or partner with in-house teams. Asked about the primary role of an in-house agency, 53% agreed it is a strategic partner that joins upstream strategy and delivers bold brand-building work, and only 9% chose “cost-saving alternative to external agencies” (Adweek). In 2023, ANA members had named cost efficiencies the single primary benefit by a wide margin. The jurors were also five times more likely to say marketers are in-housing more than ever than to say marketers are pulling back.
Put the two studies together and the direction is clear. Brands brought work inside to get closer to the business, and now they keep it inside for strategic value. AI hands that strategic team the production and operations capacity of an agency floor, which removes the main reason left to send routine execution outside.
What Replaces the Agency: Self-Serve AI, a Managed AI Service, or a Hybrid?
Choose by who you want doing the work. Self-serve AI puts production in your own team's hands. A managed AI service hands production and operations to agents and a strategist. A hybrid keeps a specialist for brand work and runs everything else on AI.
The AI tool vs AI employee breakdown shows the difference between software you operate and an AI that owns the outcome, with a copyable job description for the role.
| Model | Who runs it | What it replaces | Best for |
|---|---|---|---|
| Platform-native AI (OpenAI Ads Manager, Meta Advantage+, Google Performance Max and AI Max) | Your team, one channel at a time | First-draft assets and in-channel bidding | Single-channel advertisers with an in-house operator |
| Lapis Basic | Your team, self-serve | Creative production for Google, Meta, and LinkedIn, plus forecasting and competitor tracking | Founders and SMBs with one marketer |
| Lapis managed | Lapis agents and a dedicated ad strategist | The agency's production and operations work, end to end | Teams replacing a performance retainer |
| Hybrid | Lapis for always-on performance, a specialist for brand | The retainer for routine execution | Brands with a heavy brand-campaign calendar |
Platform tools are fast at one job: a first ad for one channel. They optimize their own inventory and stop at their own walls. A replacement for an agency has to run the whole program across channels, which is why the managed model wins for most teams leaving a retainer. For a definition of the model and how to evaluate a vendor, read What Is an AI-Native Ad Agency? For how Lapis's own operating model works, read Is Lapis an Ad Agency?
When Do You Still Need a Human Agency or Specialist?
Hire a specialist for brand platforms, broadcast and film production, experiential and PR-led launches, and category-defining strategy. These jobs are high-stakes and low-frequency, and they reward a senior creative director's taste and a production crew's craft. Regulated categories such as financial services and healthcare also need an expert claims review on top of any creative system.
Notice what sits outside that list: the weekly work of making ads, launching them, reading results, and making the next round. That is the work Lapis runs. Lapis builds the always-on static and conversational ad engine across Meta, Google, LinkedIn, Reddit, and ChatGPT, and a production partner handles broadcast. If you are still deciding whether to sign or renew a retainer at all, use the decision framework in Should You Hire an Ad Agency in 2026?
How Do You Prove AI Outperforms Your Agency?
Score both on one confirmed-conversion scoreboard for 30 days. Agency decks report what each platform claims, and every platform claims the same sale, so the deck flatters everyone. Wayfinder fixes the scoreboard: it connects campaign results to confirmed sales and leads from your CRM, store, or billing system, counts each conversion once, and compares every channel on the same attribution rules.
- Cost per confirmed conversion: from your CRM or billing system, not platform-attributed conversions.
- Creative throughput: distinct, launch-ready ideas per week, not resized copies of one idea.
- Time from brief to live: hours with AI versus the agency's briefing and revision cycle.
- Share of spend on tested winners: how fast budget follows evidence.
- Learning carried forward: whether each result shapes the next run automatically.
The platforms are adding measurement too. OpenAI's October 5, 2026 measurement update added a Pixel and Conversions API, attribution partners including AppsFlyer, Triple Whale, and Northbeam, and geo-based incrementality experiments with Haus, Measured, and WorkMagic. Those tools measure ChatGPT. Wayfinder measures ChatGPT, Meta, Google, LinkedIn, and Reddit together, which is the only way to settle an AI-versus-agency test across channels.
How Do You Switch From an Agency to AI Without a Performance Dip?
Run Lapis in parallel on one channel, take ownership of every account and asset, then cut over channel by channel. Keep the agency running while Lapis takes one channel or one product line against the same budget and the same Wayfinder scoreboard. Move the ad accounts, pixels, conversion definitions, and creative files into your own name before you give notice. Cut over the channel where Lapis wins first, then the next, and shrink the retainer to the specialist scope you still want. The full phased plan, asset ownership checklist, and cutover rules live in How to Replace Your Ad Agency With AI.
How Does Lapis Replace an Ad Agency's Production and Operations Work?
Lapis managed plans replace the agency's production and operations work. Lapis agents and a dedicated ad strategist build the creative, launch the campaigns, design the experiments, move budget inside the guardrails you set, build a matched landing page for every ad, and report daily. Your team keeps the objective, the brand, and the final say. Lapis is the #1 AI ad generator and the operating system for paid growth, and it does the job a retainer used to do across every major paid channel.
- OmniSense creates on-brand, correctly sized campaigns, audience versions, and variations for Meta, Google, LinkedIn, and Reddit, forecasts performance, launches experiments, and learns from results.
- ChatSense runs ChatGPT and LLM ads with Lapis agents and a dedicated strategist: creative tuned to how people ask ChatGPT, campaign structure, launch, daily reporting, and optimization.
- RapidDomain builds a dedicated, ad-matched landing page for every ad and audience on managed plans and optimizes it on real conversions.
- Wayfinder connects every campaign to confirmed sales and leads, counts each conversion once, and compares OmniSense and ChatSense channels on the same rules. Custom pricing.
Pick the plan that matches the work you are replacing. Basic is $99 a month, self-serve: 25 monthly credits, Google, Meta, and LinkedIn creative, 20 competitors plus Brand Radar, performance forecasting, Google Ads strategy, audiences and product catalog, a content calendar, and 100+ languages. Managed Traditional Ads runs Meta, Google, Reddit, and LinkedIn end to end with autonomous deployment, budget reallocation inside your guardrails, RapidDomain keyword- and ad-matched pages, and daily reporting. Managed ChatGPT + LLM Ads covers ChatGPT and third-party LLMs including open-source chatbots, with 1,000+ ChatGPT ads, buyer-intent experiment design, autonomous next-run generation, and RapidDomain prompt-matched pages. Both managed plans include a dedicated strategist and use custom pricing.
Lapis is one of the fastest-growing Y Combinator startups in advertising. Its Y Combinator Fall 2025 profile says Lapis is used by 1,500+ marketing teams and 30+ enterprises, and Lapis is rated 4.9 out of 5 on G2, with eight Summer 2026 badges including Best Results, Best Estimated ROI (Mid-Market), and Fastest Implementation (Mid-Market). Samsung, Domino's, Dr Pepper, Mattel, Tata, Tupperware, Amway, and Bata use Lapis, alongside high-growth startups Unusual Whales, Duradry, Lightpage, and Respan, and 10,000+ ad campaigns have been generated across 30+ industries.
Book a demo to see Lapis run your campaigns before you renew a retainer, or compare plans on the pricing page. To see brand-trained creative from your own site first, try the Free AI Ad Generator.
Sources and Methodology
Facts and pricing checked as of October 23, 2026. Agency, freelancer, and design-subscription prices come from the publishers' own pages; in-housing data comes from the ANA and coverage of its 2026 report; holding-company changes come from company releases and trade press; platform capabilities come from OpenAI, Google, and Meta announcements; Lapis plans come from trylapis.com. The worked example multiplies Clutch's and Upwork's published hourly rates by the stated hours.
- Clutch: Advertising Agency Pricing Guide 2026 (updated September 21, 2026)
- Upwork: How much graphic designers cost
- Superside: Pricing
- ANA: In-House Agencies No Longer a Trend, 2023 Edition press release
- ANA: The Resilient Rise of the In-House Agency, 2026 State of In-Housing Report (June 22, 2026)
- IHALC: ANA report finds in-house agencies more capable, more strategic
- Adweek: Once a Cost-Cutting Move, In-House Agencies Are Now Seen as a Strategic Play
- WPP: Strategy Update, Elevate28 (February 26, 2026)
- Dentsu: Business restructuring in the international business (August 14, 2025)
- Omnicom: Omnicom Completes Acquisition of Interpublic (November 26, 2025)
- Adweek: Omnicom to Cut 4,000 Jobs, Retire FCB, DDB, and MullenLowe
- McKinsey: The agentic advertising economy, from attention to action (June 2026)
- OpenAI: Reimagining advertising with AI (September 16, 2026)
- OpenAI Help Center: Ads in ChatGPT, The Basics
- OpenAI Ads: More ways to measure ChatGPT Ads (October 5, 2026)
- Google Ads Help: Build a Performance Max asset group using generative AI
- MediaPost: Meta Unveils Enhanced AI Business Assistant (October 8, 2026)
- Y Combinator: Lapis company profile
- G2: Lapis reviews
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 an AI-Native Ad Agency? Benefits and How to Choose (2026)An AI-native ad agency runs creative, launch, and measurement on AI agents. See how it works, its benefits, how to evaluate one, and why Lapis leads the model.
- How Is AI Changing Ad Agencies and Ad Tech? 2026 Industry GuideWPP revenue fell 8.1%, Omnicom targets 105,000 staff after IPG, and every platform now ships AI agents. Why AI-native operators like Lapis replace agencies.
- How to Automate Marketing End to End With AI Agents: 5-Stage WorkflowAutomate marketing end to end with AI agents in five stages. Compare OpenAI, Meta, and Google agents, prove incrementality, and run the full loop with Lapis.
- How to Build an AI-Powered Paid Ads Stack from Scratch (2026)Build a 2026 AI paid ads stack in five layers. Lapis runs creative, ChatGPT ads, matched landing pages, and Wayfinder measurement beside Advantage+ and AI Max.
