Home › Services › In-house AI marketing

The AI marketing agency that builds it in-house

Build your marketing in-house with AI.

Tired of doing your own marketing, or of agency after agency? Take it into your own hands. PELORA builds your whole AI marketing system inside your business with Claude Code, trains your team to run it and sets up your marketing finances. After handover, if you want, we stay on until your team doesn't need us.

From $50,000, one time, by complexity, up to $100,000. Optional support after handover from $2,000 a month.

For healthcare practices, health and wellness businesses and health product brands nationwide, from Newport Beach, California.

Diagram: the marketing system sits inside your business. A rulebook in the middle feeds content, ads, search, follow-up and reports, your team approves the work, and it is all saved in your own GitHub. YOUR BUSINESS Content Ads Search and AI Follow-up Your team approves Rulebook facts, voice, do-nots Saved in your own GitHub
The short answer

An in-house AI marketing build puts your whole marketing system inside your business and trains your team to run it. PELORA builds it with Claude Code in your own accounts, sets up your marketing finances and coaches your team through handover. From $50,000, one time, by complexity, up to $100,000. Optional support after handover from $2,000 a month.

Last updated . What we checked: our price book line for this build, and the script and review notes of the film on this page (November 12, 2026); five live pages from firms that sell AI marketing builds and training, for the questions buyers ask (November 12, 2026); the studies, surveys and company reports linked on this page, each opened and read on November 12, 2026; keyword, question and results data for these searches in the SEO tool Ahrefs for the US, and our own Google Search Console for October 10 to November 7, 2026 (read November 9, 2026).

$5M+In ad spend managed
10 yearsOur team has worked together
12 yearsOur founder has owned and operated treatment centers

All three are from PELORA's verified company figures and our founder's record.

Watch: how we build your marketing in-house with AI

This video uses AI-generated voice, avatar and scenes. The screens shown are examples.

Read the film's transcript

Tired of marketing agencies promising you the moon? You're not the only one. And here's the twist: we're a marketing agency. Plenty of businesses still want us to run everything, and we do. But some owners want their marketing in-house, with AI. So we build it for you. A complete Claude Code system for your marketing, set up on your computer and saved in your own GitHub. It connects to the platforms you already use, through MCP connectors and APIs. Like Meta Ads. Google Ads. Google Analytics. Search Console. Ahrefs. Microsoft Clarity. Local Falcon. Apify. Canva. And your inbox. Inside, it's organized like a real team. Ads. SEO. Content. Influencers. Reporting. Each with its own instructions, skills and checks. Every morning it can pull your numbers, write the report, and flag what needs fixing. It drafts your ads, your posts and your creator briefs. It tracks your marketing finances: budgets, cost per lead, and ad spend. And a person on your team approves anything before it goes live. For health businesses, patient data stays out unless a signed BAA covers every tool. It runs in your accounts. You own it. We train your team, and hand it over. Built by an agency that runs on this same system every day. Ready to bring your marketing in-house? Book a consultation below.

Who it's for

Who is an in-house AI marketing build for?

It's for owners of healthcare practices, health and wellness businesses and health product brands who are tired of doing their own marketing, or of paying agency after agency, and want the system and the skills inside their own business.

We hear it from owners all the time: they have been through agency after agency, paid every month, and still own nothing that works without the agency. Others have been doing the marketing themselves, on top of running the business. Take it into your own hands. We build it out for you and coach someone on your team to run it.

Healthcare practices

Licensed medical, dental and specialty practices, treatment centers and medical groups, like the clinics we market today. The system follows calls, leads and bookings by channel, keeps patient data out of the AI, and checks ads against Google's and Meta's health rules before they run.

Health and wellness

Gyms, studios, recovery and wellness clubs, and multi-location and franchise groups. The system follows trials and memberships, location by location, with one report across them.

Health products

Supplements, health drinks, skincare, devices and other products that help people's health, as on our health brand page. The system follows orders on your own site with order follow-up and review requests, and your Amazon, TikTok Shop and retail sales go into the monthly report where you can export them. What each marketplace allows your product is checked against its own rules first.

Who it isn't for

  • Businesses outside health and wellness. We focus there.
  • Owners who want software with nobody on their team to run it. The build needs one named owner on your side.
  • Brands we decline anywhere, such as research-use-only peptides, kratom and kava.
Why own it

Is it better to do marketing in-house or hire an agency?

If you want control and the skills inside your business without hiring a full team, there's a third choice: have the system built in-house and own it.

You wouldn't be the first to bring it in-house. In the Association of National Advertisers' 2023 survey of its member companies, 82 percent said they had an in-house agency, up from 78 percent in 2018, and the ANA named the desire to own, control and protect first-party data as a driver of bringing work in-house (ANA, May 2, 2023, 162 respondents).

Three paths: hire a team, hire an agency, or have the system built in-house, with the third path highlighted. Hire a teamHire an agencyBuilt in-house
QuestionHire a marketing teamHire an agencyBuilt in-house with PELORA
Who does the daily work?Your hiresThe agency's staffYour staff, with AI doing the drafting and checking
Who owns the accounts?YouDepends on the contract, so askYou, from day one
What are you paying for?Salaries, tools and trainingA monthly fee for their timeA one-time build, then optional support
What if someone leaves?You rehire and retrainYou start again with a new agencyThe rulebook and written guides stay in your business
Where does AI fit?Whatever your hires bringInside the agency's own toolsBuilt into your accounts and routines
What we build

What do we build inside your business?

A complete AI marketing system, run with Claude Code, Anthropic's coding tool: AI agents draft, check and report against your own written rules, and a person on your team approves every post, page, email and ad before it goes out.

The system is set up on your computer, saved in your own GitHub, and connected to the platforms you already use through connectors (MCP, a standard way for AI tools to plug into other apps) and APIs. Inside, it's organized like a real team, with a part each for ads, SEO, content, creators and reporting, and each part has its own instructions, skills and checks.

You own the accounts from day one, and at handover every login and document is in your name. The content, the customer list in your CRM and the website are yours too, so if you ever stop working with us, all of it stays with you.

What Claude Code can do in marketing, in its maker's words: Anthropic says its own growth marketing team built a Claude Code workflow with two sub-agents that "generates hundreds of new ads in minutes instead of hours" (Anthropic, July 24, 2025). That is Anthropic's result on its own team, not a result we promise you.

The six parts stacked: the rulebook at the base, then content, ads, search and AI answers, follow-up, and reporting on top. 6. Reporting 5. Follow-up 4. Search and AI answers 3. Ads 2. Content engine 1. The rulebook
1

The rulebook

Your facts, prices, voice and a do-not list, written once. Every AI agent in the system reads it before it writes, and the person you name checks each draft against it before anything goes out.

2

Content engine

Posts, emails, pages, ad copy and creator briefs drafted in your voice, checked against the rulebook, and approved by the person you name before anything publishes.

3

Ads

The workflows your team uses to plan, write and check Google and Meta campaigns in ad accounts you own, against Google's health rules and Meta's health rules. Your team runs them; managed ad buying is priced separately.

4

Search and AI answers

Service and product pages, and your Business Profile where you have one, written so Google and AI assistants can find you. The method is the one on our AI SEO agency page.

5

Follow-up

Fast replies, booking links or cart and order follow-up, reminders and review requests in your CRM, built like our CRM automation. An AI phone agent is its own build, on our AI phone agents page.

6

Reporting

Every morning the system can pull your numbers, write the report and flag what needs fixing, and once a month it builds one report by source: calls, leads, bookings, orders or memberships next to what each channel cost.

How it's set up

How is the system set up, in detail?

The same way PELORA runs itself: one AI agent per department, each with a written role, all reading one rulebook, with a second agent checking the work and a person approving anything a customer sees.

This is how our own agency runs every day, and it's what we set up inside your business, sized to the departments your marketing needs. AI agents for the rest of your business, such as the front desk or back office, are a separate service on our AI agents and automation page.

Departments, each with its own agent
  • Ads
  • Content and creative
  • SEO and AI search
  • Creators
  • Reporting
  • Marketing finance
How the system is organized: a rulebook at the top feeds six department agents, ads, content and creative, SEO and AI search, creators, reporting and marketing finance; their work goes to a second agent for a check, then to a person on your team who approves it. Rulebook: facts, prices, voice, do-not list Ads Content and creative SEO and AI search Creators Reporting Marketing finance A second agent checks it A person on your team approves A daily clock wakes the agents. Tickets track every job. Saved in your GitHub.
How the system is organized. Your departments are set in your written plan.
1

One agent per department

Each part of your marketing gets its own AI agent, for example ads, content and creative, SEO and AI search, creators, reporting and marketing finance. PELORA runs its own strategy, operations, finance, sales and proposals, media buying, creative, SEO and AI search, reporting, PR and outreach this way.

2

A written role for each

Each agent has a written job: what it owns, what it never does, who it hands work to and who it reports to. When a rule changes, it changes in one file, and every agent reads the new version.

3

One rulebook, read first

Your facts, prices, offers, voice and do-not list in one place. Every agent reads it before it writes, and a figure that isn't in it doesn't go out.

4

Skills and playbooks

Step-by-step skills for each kind of work, such as Google Ads, Meta ads, creative, copywriting, SEO and AI search, and reporting. Before a new build, the agent reads the platform's own help page that day and the best live examples, and saves what it learned as a playbook it reuses for 30 days.

5

A second check

Before anything ships, a second agent reads the work against the rulebook, the platform's rules and the job it was given, and sends it back if something is off.

6

A person approves

Nothing a customer sees goes out until the person you name approves it: posts, pages, emails, ads and replies.

7

A daily clock

The agents wake at set times each day. Every morning the system can pull your numbers, write the report and flag what needs fixing, including a new issue an ad platform raises in your account.

8

Tickets to done

Every job handed from one agent to another gets a ticket and is tracked until it's done. The agent that did the work doesn't close its own ticket: another agent checks it first.

9

Saved in your GitHub

Every rule, role, skill and template is saved in your own GitHub, with who changed what and when, so you can see the history and go back.

The tools

Which tools does the system use, and what does each do?

Claude Code is the workbench, GitHub saves the system, and connectors link it to the platforms you already use. We pick the tools your build needs and name each one in your written plan.

Tools a build can connect
  • Claude Code
  • GitHub
  • Google Ads
  • Meta ads
  • Google Analytics
  • Search Console
  • Ahrefs
  • Microsoft Clarity
  • Local Falcon
  • Apify
  • Canva
  • Buffer
  • ElevenLabs
  • OpenArt
  • Higgsfield
  • Retell
  • Your inbox
ToolWhat it does in your system
Claude CodeAnthropic's coding tool. It runs the agents, their roles, rules and skills, on your computer.
GitHubSaves the whole system in a private repository in your business's name, with a record of every change.
Google Ads and Meta adsYour ad accounts on Google, Facebook and Instagram. The system reads results and drafts campaigns and ads for your team to approve.
Google Analytics and Search ConsoleVisits, key actions and how you show up in Google search, read for your reports.
AhrefsKeyword, question and competitor research for SEO and AI search.
Microsoft ClarityHow visitors move through your pages, so you can see where they drop off.
Local FalconWhere you show on the Google map, location by location.
ApifyCollects public web data, such as the ads and listings competitors run in public.
CanvaDesign templates for posts, ads and graphics in your brand.
BufferSchedules approved posts to your social accounts.
ElevenLabsAI voice and music for videos.
OpenArt and HiggsfieldAI images and video clips.
RetellAI phone answering, when you add a phone agent, which is its own build.
Your inboxDrafts replies and reports for a person on your team to approve.

Every tool here belongs to its maker, and naming it doesn't mean its maker endorses PELORA. Claude and Claude Code are Anthropic's products, and PELORA isn't affiliated with Anthropic. You hold each account and pay each vendor directly, never marked up by us, and we set up the plans that fit your size.

What you need on your side

  • A computer

    The one your marketing owner works on every day. We set the system up on it with them.

  • A paid Claude plan

    One that includes Claude Code, in your business's name. You pay Anthropic directly; we never mark it up, and we check the current plans with you before we quote.

  • A GitHub account

    In your business's name. The system's roles, rules, prompts and templates are saved there in a private repository, with a record of every change, so they belong to you, not to a laptop or to us.

  • Your ad and analytics accounts

    Your Google and Meta ad accounts, Google Analytics, Search Console, your Business Profile and your CRM, all in your business's name. If an agency holds any of them today, moving them into your name comes first.

  • One owner on your team

    Someone who already handles your marketing, front desk or operations and is comfortable with software. They hold the rulebook and the logins and approve the work.

The finances

How do you set up marketing finances for every department?

We set up your marketing budget department by department, track what each one spends against what it brings in, and review it with you every month until handover, then on the optional support. It's marketing finances only, never your accounting or taxes.

  • Ads: spend by platform and campaign, with cost per lead and cost per new patient or customer.
  • Content and creative: shoots, editing, creators and design tools.
  • SEO and AI search: research tools, pages and articles, and what search brings in.
  • Tools and subscriptions: your Claude plan and every other AI and software subscription, in one list with what each one costs you.
  • Agencies and freelancers: what you pay today, and what stops once your team runs the system.
  • Reporting: one monthly report with spend next to calls, leads, bookings, orders or memberships, department by department.

Spend tracking compares each department's budget with its actual spend every month. The monthly review is run with the person who owns the system: with us until handover, then on the optional support, or from the written guide if you run it alone.

Marketing finances only: never bookkeeping, tax or accounting. Your accountant keeps your books.

Example: a marketing budget by department, ads, content and creative, SEO and AI search, tools and subscriptions, and agencies and freelancers, each with a budget and an actual column filled with your own numbers, flowing into a monthly review of spend next to results. Marketing budget, by department BudgetActual Ads Content and creative SEO and AI search Tools and subscriptions Agencies and freelancers yoursyoursyoursyours yoursyoursyoursyours yoursyours Monthly review: spend next to results
ExampleThe rows we fill in with you. No figures are shown because they are yours.
Training your team

How do you train our team to run it?

We train your staff live, by role, on your own work, with written guides they keep, and we coach the person who will run the system until handover.

  1. Map who does what. Each person's part of the system, and who approves what before it goes out.
  2. Write the guides. Step by step for each task, in your words, saved in your own accounts.
  3. Practice on real work. Your team runs the tasks with us beside them, and every draft goes through the approval routine.
  4. Coach one owner. One person on your side holds the rulebook and the logins. We coach them through handover, then on optional support for as long as they need us.

Does your team need AI or technical experience? No. You need one person who already handles your marketing, front desk or operations and is comfortable with software. We train the rest of your staff by role, so each person learns their own part of the system first.

Why training is part of every build: in Microsoft and LinkedIn's 2024 Work Trend Index, only 39 percent of people who use AI at work had gotten AI training from their company (Microsoft and LinkedIn, May 8, 2024, a survey of 31,000 knowledge workers in 31 markets).

Example: a page from your written guides, titled Reply to a Google review, with four steps, three checked off. guides / front-desk / review-replies Reply to a Google review Run the review-reply prompt Check it: no patient details Send it to the approver Post once it is approved Example guide
ExampleA drawn example of a written guide, not a real account.
Marketing skills

What marketing do you teach our team?

We teach the marketing behind each workflow your team runs, not only the buttons: advertising on Google and Meta, creative, copy, branding, landing pages, SEO and AI search, tracking and reading results, so your team knows what good looks like and what to look for.

The lessons come from the same written skills our own agents follow, taught on your own accounts and your own work.

Google Ads

  • Pick one number first: cost per booked appointment or qualified call, not clicks.
  • Set up tracking before you spend.
  • Read the search terms every week and add negative keywords.
  • Make the search, the ad and the page say the same thing.
  • Know Google's health ad rules before you write the ad.

Meta ads

  • Fewer, larger ad sets, so Meta has enough results to learn from.
  • Let the creative do the targeting: test different ideas, not small tweaks.
  • Launch the whole ad set, then leave it alone while it learns.
  • Judge leads by quality, not just count.
  • Never write copy that implies something about the viewer's health.

Creative

  • The hook in the first two seconds, then the reason to buy, then the offer.
  • Real people and real footage first in health ads.
  • Read hook rate and hold rate, and retire what doesn't move them.
  • Learn from winning patterns, and never copy another brand's ad.

Copywriting

  • Start from what the reader already knows about their problem.
  • Use your customers' own words in headlines.
  • Benefits over features, short and plain.
  • One clear action per page or ad, with claims your business can back up.

Branding

  • Your voice, colors, type and logo rules written down once.
  • Every post, page and ad checked against them.
  • A look made for your business, not a template.

Landing pages

  • One offer and one action per page.
  • The page says what the ad promised.
  • Fast on a phone, with a short form and real proof.
  • Test one change at a time.

SEO and AI search

  • Build pages for questions people actually search.
  • Answer first, then the detail.
  • Every number with its source and date.
  • Keep your Business Profile complete, and measure in Search Console.

Tracking

  • Conversion tracking working before any spend.
  • Key events set up in Google Analytics.
  • No health details sent to ad platforms.

Reading results

  • Compute every number; never eyeball it.
  • Compare the same date ranges.
  • Never add Google's and Meta's counts together.
  • Platforms over-claim, so ask new customers how they found you.

What your team will make, step by step

Drawn examples of how the work runs. Your own steps are written into your guides during training.

Example

Make an AI video or animation

  1. Write the script in your voice and check it against your rulebook.
  2. Make the voice, music and scenes with AI tools.
  3. Put it together with captions and an AI label.
  4. Your approver watches it before it posts.

The film on this page was made this way in-house: the script approved by our founder, the voice and music made in ElevenLabs, the avatar and scenes in OpenArt, and the picture put together in code.

Example

Build a website page

  1. Write the brief: the question the page answers and who it's for.
  2. Draft it in your voice from your rulebook's facts and prices.
  3. Check it: spelling, rules, links and speed on a phone.
  4. Your approver reviews a preview, then it publishes.
Example

Make a landing page for an ad

  1. Match the page to the ad's promise and who it's for.
  2. Give it one offer and one form or booking button.
  3. Check the claims against the platform's rules and your do-not list.
  4. Approve it, publish it, and watch its cost per lead in your report.
Example

Save a template

  1. Pick a piece that worked: an email, a post or an ad.
  2. Save its layout with blanks for what changes each time.
  3. Name it and store it in your GitHub.
  4. Next time, fill in the blanks; the same checks still run.
Example

Save your prompts

  1. When a prompt works, save it as a file named for the task.
  2. Write in what it needs: the inputs, the rulebook and the checks.
  3. Turn it into a skill in the system, so anyone runs it the same way.
  4. Improve it in one place, and everyone gets the update.
Example

Set up GitHub

  1. Create the account in your business's name.
  2. Make one private repository for your marketing system.
  3. Add the people who run it, each with their own login.
  4. Every change is saved with who made it and when, so you can go back.
Compliance and HIPAA

How do you handle compliance and HIPAA?

No patient data goes into the AI unless a business associate agreement covers your whole tool chain, and a person on your team approves everything before it goes live.

We build the marketing system to run without patient health information, and we don't promise HIPAA compliance. Before PELORA has access to any system that holds patient data, the practice signs a business associate agreement with us.

  • A person approves. Every post, page, email and ad waits for the person you name. Automatic replies run only from templates your team approved once.
  • Ads checked first. Ads are checked against Google's health rules and Meta's health rules before they run.
  • Tools read first. Each tool is named in your written plan, and we read its data terms before we connect it.
  • Your do-not list. The claims your business never makes are written into the rulebook, and every draft is checked against them.

Why every person gets their own login: in IBM's 2025 Cost of a Data Breach report, 13 percent of organizations reported breaches of AI models or applications, and of those, 97 percent reported not having AI access controls in place (IBM, July 30, 2025, 600 organizations that had a data breach, March 2024 to February 2025).

Example screen: an approval queue where a staff member's drafts wait for the owner's approval, and one changes from waiting to approved. Example screen Approval queue Instagram post draftFrom: marketing coordinator Waiting Approved Review reply draftFrom: front desk lead Waiting Monthly reportReady for the owner Approved
ExampleA drawn example, not a real account or person.
The process

How does an in-house AI marketing build go, step by step?

It runs in four steps, from your finances to handover, and the time it takes depends on how complex your system is. You get the timeline in writing with your quote.

The four steps of the build in order: finances and audit, the build, live training, then handover with 30 days of fixes, followed by optional support. 1 2 3 4 FinancesBuildTrainingHandover Support (optional)
  1. 1. Finances and audit

    Your marketing spend, your tools and your accounts on one page, and the rulebook started: your facts, prices, voice and do-not list.

    You get
    Your marketing budget on one page and the first draft of your rulebook.
    We need from you
    Access to your accounts, your current costs, and the person who will run the system.
    Ends with
    Your sign-off.
  2. 2. The build

    The content engine, the ad workflows, search pages, follow-up and the monthly report, set up in your own accounts and saved in your GitHub.

    You get
    The six parts, working in your accounts.
    We need from you
    Your facts, prices and voice for the rulebook, and answers when we ask.
    Ends with
    Your sign-off.
  3. 3. Live training

    Your staff trained live, role by role, on your own work, with written guides, and we coach the person who will run it.

    You get
    Trained staff and written guides they keep.
    We need from you
    Time on the calendar for the people we train.
    Ends with
    Your sign-off.
  4. 4. Handover

    Documents, ownership and credentials in your name and a recorded walkthrough, then 30 days of fixes for defects in what we built.

    You get
    Every login and document in your name, and the recorded walkthrough.
    We need from you
    The person who runs the system, at the walkthrough.
    Ends with
    Your sign-off.
Build cost

What does an in-house AI marketing build cost?

From $50,000, one time, by complexity, up to $100,000. Optional support after handover from $2,000 a month.

What sets the level is how complex your system is: how many workflows, AI agents and tools we connect, any custom code, how many locations or brands, and how many staff we train. We scope it on a call and quote it in writing.

Every build includes

  • The system build, in your accounts.
  • Live staff training plus written guides.
  • Your marketing finance setup.
  • Handover: documents, ownership and credentials in your name, a recorded walkthrough and your sign-off.
  • 30 days of fixes for defects in what we built.

Not included

  • Ad spend, which you pay to the platforms directly.
  • Your Claude plan and other AI and software licenses, which you pay directly and we never mark up.
  • Managed ad buying, done-for-you creative and shoots, which are priced separately.
Example, not a promise

How a build can pay for itself

Example: a business paying an agency $5,000 a month that stops and keeps the $2,000 support plan saves $3,000 a month, so a $50,000 build pays back in about 17 months.

The assumptions behind it: the agency fee is $5,000 a month and stops; the business takes the optional support at $2,000 a month; the build is $50,000; ad spend, the AI and software licenses you pay directly, and the time your own staff spend running the system are left out and compared in your finances step. Your numbers will differ, and this is never a promised saving.

Try it with your own numbers. B is the build from your written quote, M is the monthly fee you'd stop paying, and R is the monthly support if you take it. Months to pay back = B divided by (M minus R).

Enter your own three numbers to see the months.

Left out of this math: ad spend, the AI and software licenses you pay directly, and the time your own staff spend running the system. We compare those in your finances step. An example, never a promised saving.

Book a consultation

We scope your build and quote it in writing.

After handover

Do you stay with us after the handover?

Yes, if you want us: optional support after handover is from $2,000 a month, and you keep it until your team doesn't need us.

  • A monthly system review and the improvements that come out of it.
  • New workflows as your business grows.
  • Model and tool updates when the AI and software you use change.
  • A monthly report and review call, including the marketing finance review.
  • Staff questions by email, so the person running it isn't on their own.

It doesn't include ad buying or done-for-you content. If you take it, there's a three-month minimum, then month to month with 30 days notice.

The optional monthly support loop: review the system, improve it, update models and tools, report, and review again. Review Improve Update Report Every month
Track record

Have you built this for a client before?

Not yet: this offer is new, and no client has had this build. PELORA runs its own marketing and operations on this kind of system every day, with a person approving what goes out, and you can check parts of it on this site.

Our own system works the way yours would: AI agents in set roles, a second agent checking their work, people approving it, and outside specialists where a job needs one. We'd rather show you that than a case study we don't have.

How our own work moves: an AI agent drafts, a second agent checks, a person approves, then it goes live. Draft Check Approve Live AI agentsecond agenta personon our word
How we work

One rulebook, read first

One file of company facts, one price book and a do-not list. Our agents read them before they write, and our rule is that a figure not in them doesn't go out.

How we work

A second check

Our rule: anything a client or a client's customer will see is checked by a second agent before it goes anywhere.

How we work

A person says yes

Nothing deploys to our site or a client's site until our founder says so.

Check it yourself

This page

Researched, written, built and checked by our agents, with the sources in the "What we checked" line above, and published only on our founder's word.

Check it yourself

The film on this page

Scripted, voiced and animated in-house with AI tools, with the script approved by our founder and the film checked by two review agents.

Check it yourself

Our prices

This build's price is on this page, and our monthly plans are on our pricing page. Ad spend is billed to clients directly, never marked up.

Book a consultation

Ask us to show you how a draft moves from agent to approval.

Questions

What owners ask about building marketing in-house with AI.

How much does an in-house AI marketing build cost?

From $50,000, one time, by complexity, up to $100,000. Optional support after handover from $2,000 a month. The level depends on how many workflows, AI agents and tools we connect, any custom code, how many locations or brands, and how many staff we train, and we quote it in writing after a call.

What is included in the build?

The system built in your own accounts, live staff training with written guides, your marketing finance setup, and a handover with documents, ownership and credentials in your name, a recorded walkthrough and your sign-off, plus 30 days of fixes for defects in what we built. Ad spend, your tool subscriptions, managed ad buying, done-for-you creative and shoots are not included.

Who owns the system and the code?

You do. The accounts are in your business's name from day one, the system and every rule, skill and template are saved in your own GitHub, and at handover every login and document is in your name. If you stop working with us, all of it stays with you.

What if the person we train leaves?

The rulebook, the roles and the written guides stay in your business, and every login is in your name, so the next person learns from the same guides. If you have the optional support, we answer their questions by email.

Is our data secure, and what about HIPAA?

We read each tool's data terms before we connect it, name every tool in your written plan, and give each person on your team their own login. No patient data goes into the AI unless a business associate agreement covers your whole tool chain, and we don't promise HIPAA compliance. Before PELORA has access to any system that holds patient data, the practice signs a business associate agreement with us.

Will it replace our staff?

No. The AI does the drafting, checking and reporting, and your people set the direction and approve the work. The point is that someone on your team runs it, not that nobody does.

What does it cost to run each month?

It depends on the tools your build uses, and you pay each vendor directly: your Claude plan, any other AI and software subscriptions, and your ad spend, never marked up by us. We set up the plans that fit your size, list every tool in your written plan, and put the running costs on your monthly report. Optional support after handover is from $2,000 a month if you want it.

How fast will we see results?

We don't promise a timeline or a result. The system works from handover, and what it brings in depends on your offer, budget, market and how steadily your team runs it. In the finance step we set the numbers you'll watch, such as cost per lead and cost per new patient or customer, and you read them in your report every month.

Why not just hire an agency?

You can, and we're one: plenty of businesses still have us run everything. The build is for owners who want the system and the skills inside their own business, in their own accounts, with their own team running it.

We already use ChatGPT. Why would we need a build?

Because one chat tool isn't a system. The build gives every agent the same rulebook to read first, a second check and an approval step before anything goes out, workflows set up in your own accounts, staff trained live with written guides, and one monthly report. You can keep the tools you like; we name each one in your written plan.

Your marketing system. In your business. Yours to keep.

Book a consultation. We'll look at what you pay for marketing today, what the system would look like inside your business, and whether a build makes sense for you.

Or pick a time on our booking page.