Traditional agencies sell attention. The retainer covers a fixed number of hours each month, you get a deliverable each week, and on the day the contract ends, you do not have anything that keeps running on its own. The work stops because the attention stops. I do not do that. I install marketing infrastructure that keeps producing when I am gone. The retainer is for upkeep, not for output.
This post lays out the model I run at Pfender Marketing Co., and the version of it any operator can run for themselves. It is not a pitch. It is a playbook. If you take it and execute it in-house, or hire someone else to, the post still did its job.
What is the rented contractor model?
The rented contractor model is the standard agency retainer: a 6 to 12 month commitment where the client pays for ongoing access to people who execute campaigns on the client's behalf. The fee buys hours, deliverables, and someone to call. It does not buy a transferable system.
The shape is familiar. A discovery call, a proposal with a monthly fee somewhere between $5,000 and $25,000, a kickoff, weekly status calls, monthly reports that summarize what the agency did, and an auto-renew clause at the end. The work output - the ad accounts, the analytics dashboards, the CRM workflows, the content calendar, the creative files - sits in a shared drive or, more often, inside accounts the agency owns. When the retainer ends, the access ends with it.
The model persists because it is good for the agency. According to the 2025 ANA/4As Client-Agency Relationship Tenure study, retainer clients now stay an average of 56 months versus 24 months for project clients, and retainer agencies post 18% annual churn versus 42% for project shops. Recurring revenue is structurally easier to run a business on. I am not pretending otherwise. But the same structure that makes the retainer comfortable for the seller makes it expensive for the buyer over time, because the buyer is paying for continuity of a system they do not own.
Why some clients are moving away from it
Some companies get to the end of a multi-year retainer and ask themselves a reasonable question: what do we own? The honest answer in many cases: a logo refresh, a website that needs a developer to touch, and a set of campaigns running inside accounts the agency controls. The strategy lives in the agency's head. The data lives in the agency's tools. None of that transfers in a tidy folder when the contract closes.
That is not a knock on agencies - it is a description of the standard deliverable set. The question is whether it matches what the client actually needs.
What is an installed marketing engine?
An installed marketing engine is a working marketing system, configured under the client's ownership, that continues to produce results after the person who installed it walks away. The client owns the accounts. The client owns the data. The client owns the documentation. The operator who built it can leave the room and the engine keeps running.
In concrete terms, an installed engine is the set of artifacts and accounts a business actually uses to acquire and retain customers, all of it under the client's login. For a mid-market service business or a venture-backed SaaS, that usually looks like this:
- A website built on a platform the client's team can edit (in my case, Webflow), with the CMS structured for ongoing content, and the codebase in a repo the client can hand to any developer.
- GA4 and Google Tag Manager configured under the client's Google account, with events named consistently, conversions defined, and a documented measurement plan.
- A CRM with contact records, lifecycle stages, email and SMS automation flows, scheduling, and a defined data model the client owns end-to-end. For my clients, this is usually Tree CRM, the SaaS I built for exactly this slot.
- Paid media accounts (Meta, Google, Apple Search Ads where the client has an app) created under the client's billing and ad account ID, with naming conventions, audience definitions, and tracking documented in a runbook.
- A messaging framework, brand identity files, and a content cadence the client's team can produce against without me in the loop.
- A runbook that documents the weekly, monthly, and quarterly operating rhythm of the engine, including who does what, where, with what access.
None of that is exotic. The work is normal marketing work. What is different is the default ownership of every artifact and every account, and the documentation that lets the client run it without me.
The retainer model rents you attention. The install model gives you a system. The first one stops the moment the contract closes. The second one is yours.
How does the 90-day install sprint run?
The install sprint is a fixed-scope, fixed-fee, 90-day engagement that builds and ships the engine. There is no monthly recurring fee during the sprint. The deliverable is the engine itself, documented and accessible. The sequence I run is the same every time, because the order is what makes the install hold together.
- Weeks 1-2, full-funnel audit and measurement plan. I audit the current site, GA4, ad accounts, CRM, and content. I document what exists, what is broken, and what is missing. The output is a written measurement plan and a prioritized backlog.
- Weeks 3-5, foundation build. GA4 and GTM are rebuilt under the client's accounts. The CRM is configured: lifecycle stages, properties, list segmentation, the data model. The website rebuild starts in parallel in Webflow.
- Weeks 6-8, acquisition layer. Paid media accounts are created or restructured under client billing. Automation flows are built and tested. ASO and on-page SEO are addressed for the priority pages. Messaging and brand identity get finalized and applied to the site and the funnel.
- Weeks 9-11, integration and content cadence. Everything is wired together. Events fire correctly. Forms route correctly. Automations trigger correctly. The first month of content gets produced inside the cadence the client's team will own.
- Week 12, handoff. A runbook ships covering account access, weekly operating rhythm, where things live, and how to extend the system. We do a working session walking through every artifact. The client decides at this point whether they want a retainer for operation, a retainer for advisory only, or no retainer at all.
The fixed scope is the point. The client knows what they are buying and when it ships. The fixed fee removes the incentive to drag the work out. If something out of scope shows up mid-sprint, it gets logged for after the install and we keep moving.
What does the retainer look like after the install?
After the install, the retainer is for operating the engine, not for building it. The build is done. The retainer covers running it: producing content against the cadence, managing paid media, maintaining the CRM, watching the analytics, and iterating on what is working. The fee is smaller than the install fee because the heavy lifting already happened.
The client has three real choices at the end of the sprint, and I show all three honestly.
Option 1: I operate the engine
The client keeps me on a smaller monthly retainer to run the system I installed. This is the right move when the client's team is busy with the product or service they actually sell and does not have anyone with the bandwidth or skill to run marketing operations. The retainer covers a defined set of weekly outputs against the runbook. It is not a blank check for hours.
Option 2: The client's team operates it
The runbook is written for this. A reasonably capable marketing coordinator can run an installed engine because the documentation tells them what to do, where, and when. I stay available as an advisor at a small monthly minimum or hourly, and the client team owns execution. This is the model fractional CMO firms have been productizing - typical fractional CMO engagements run $5,000 to $15,000 per month for 2 to 4 days a week of senior leadership without execution - except in this version, the execution layer already exists.
Option 3: No retainer
The engine is running. The team is trained. The client has everything they need and they do not need me. This is a real outcome, not a courtesy line. The model is built so the client can leave cleanly.
What is the pricing model contrast?
The rented contractor model is monthly recurring revenue from day one. The installed engine is a one-time project fee for the install, then a smaller optional retainer if the client wants operation. Over a three-year window, the installed engine model usually costs the client less, even before factoring in what they own at the end.
| Dimension | Rented contractor | Installed engine |
|---|---|---|
| Ownership of accounts and data | Agency-owned or shared, often inside agency's logins | Client-owned from day one, every account under client's billing and admin |
| Primary deliverable | Monthly hours and campaign outputs | A configured, documented system the client owns |
| What happens when the engagement ends | The work stops; accounts are reclaimed or revoked | The engine keeps running; client team or successor takes over |
| Pricing model | Ongoing monthly retainer, typically 6-12 month commitment | Fixed project fee for 90-day install, then optional smaller retainer |
| Purpose of the retainer (if any) | Continued execution; output stops when the retainer stops | Upkeep and iteration on a system already producing |
| Scaling path | Bigger retainer, more agency hours | Hire in-house, extend the engine, layer a fractional CMO on top |
| Documentation | Monthly status reports | Runbook covering every system, account, and operating cadence |
When does the rented contractor model still make sense?
It makes sense when the client genuinely does not want to own anything. There are categories where this is the right call: large enterprises with brand budgets that need pure creative production, companies running short campaign bursts with no intention of building a system, and any business where the agency's media buying scale or production volume materially outperforms what an installed engine could do at the client's scale.
Agencies win on execution capacity. A 30-person shop can ship more creative in a week than I can ship in a month. If the client's bottleneck is production volume and they already have a working system, an agency retainer is the right tool. If the client's bottleneck is not having a working system in the first place, paying an agency $20,000 a month to execute against the gap is a slow way to never close it.
How does this differ from a fractional CMO?
A fractional CMO is strategic leadership without execution capacity. I install execution capacity. The two models pair well: a fractional CMO can direct a team or an agency more effectively when the underlying system is already built. What I do is the layer underneath the strategy, the thing that converts strategic direction into working software and accounts the company owns.
Most fractional CMO engagements run $5,000 to $15,000 a month for senior leadership 2 to 4 days a week. That is the right spend if the company's gap is strategic clarity. If the gap is that the website is broken, GA4 was never set up properly, the CRM is a spreadsheet, and the paid accounts are running inside a former agency's login, no amount of strategic clarity fixes that. The install does.
Common questions
Does the client really own the ad accounts and analytics?
Yes. Every account is created under the client's Google, Meta, Apple, and Webflow logins, with billing on the client's card and admin role assigned to the client's email. I am added as a manager or editor. If I disappear tomorrow, the client revokes my access and nothing breaks.
What happens if I want to bring marketing fully in-house after the install?
That is the point. The runbook is written for it. A marketing coordinator or marketing manager hired by the client can take over the operating rhythm because the documentation specifies what runs when, where it lives, and how to extend it. I have had clients do exactly this within six months of the install closing.
Is 90 days actually long enough to build all of that?
For a mid-market service business or an early-stage SaaS, yes, because the scope is bounded and the sequence is rehearsed. The install is not "build a custom marketing stack from scratch." It is "configure a known stack of platforms under your ownership, instrumented correctly, with documentation." The sprint is tight because the work is defined.
Why does the model work for a solo consultant but not for a bigger agency?
It works for me because I am AI-augmented and the scope is fixed. The leverage that used to require a team of five now requires one operator with the right tools. Bigger agencies have payroll to cover and incentives to keep clients on long retainers, so the rented contractor model fits their economics better. The installed engine model fits mine, and it fits many clients better too.
This is the model I run at Pfender Marketing Co. If you are evaluating whether to hire an agency, hire a fractional CMO, build a team in-house, or work with someone like me, the right question is not which model is best in the abstract. It is which model leaves you owning the engine at the end of the year. Pick the one that does.
