01
Drift gets caught early
Weekly OARS scoring, monthly reports, the specific deltas highlighted in plain language. You hear about regressions from me before you hear about them from a customer — or from your board.
Consulting · Retainer
Ongoing Infrastructure Retainer
The agent ecosystem shifts every quarter — model behaviors change, standards drift, new AI sources appear and disappear. A retainer is the simple answer to "who's watching all that for us?" so you can spend your attention on running the business.
— the problem
What this is actually for
The infrastructure shipped, the audit said you'd cleared Level 3, and for a few months the dashboard looked great. Then Schema.org rolled out a new property, OpenAI changed how it handles allowed-bot directives, your MCP server's behavior shifted under a Claude update, and OARS published a new revision. Nobody on your team noticed — they had real work to do.
This is what rot looks like in agent infrastructure. Not catastrophic failure, just steady drift, until one quarter you notice that AI-referral traffic is down by a third and nobody can say when it happened. A retainer puts a human on watch and a hand on the tiller, so the score stays steady and the surprises stop landing in board meetings.
— outcomes
What's different with a retainer in place
Three things that change once we're running.
02
The work actually gets done
Spec changes, schema updates, performance fixes — they happen, by me, before they pile up. Not "flagged for your team to handle when they get a minute," because that minute rarely comes.
03
There's a person to call
When a new AI source appears, when a partner asks if you've integrated with X, when your engineering lead has a "wait, does this affect agents?" question — you have someone whose actual job it is to know.
— scope
What's included
Four things, every month.
The retainer is a flat-fee arrangement covering ongoing monitoring, maintenance, and direct access. No per-ticket billing, no nickel-and-diming on small changes. Larger projects that fall outside the retainer scope (a new MCP module, a major refactor) are quoted separately and credited against the retainer when sensible.
Always-on
Continuous OARS monitoring
Your site and infrastructure scored weekly against the live OARS spec — every level, starting at Level 0. Uptime, security header drift, new CAPTCHA-on-agent issues, and Core Web Vitals regressions flagged alongside the higher-level findings, so the foundation gets watched as carefully as the action layer above it.
As needed
Schema & spec updates
When Schema.org, MCP, or OARS itself changes, your implementation gets updated. Standards drift is my problem to track, not yours — and the work is done, not just flagged.
Proactive
Performance optimization
Agent capability index movement, schema quality drift, AI-referral trend shifts, MCP error rates — actioned, not just reported. Action is what you're actually paying for.
Same person
Direct access
Slack or email, same-day response on business days, a monthly strategic sync. Same person every time — no handoffs, no escalation paths, no account managers reading a script back to you.
— cadence
What a year actually looks like
The rhythm of an ongoing engagement.
Month 1
Onboarding
A short baseline audit of whatever you've got — built by me or anyone else. Monitoring goes live, alerting wired in, first report drops at month-end.
Months 2–11
Steady-state
Monthly report plus a sync call. The work happens between syncs; the syncs are for direction, trade-offs, and the strategic questions that don't fit in Slack.
Quarterly
Strategic review
A longer session, every three months: where the ecosystem moved, what it means for your business, what we should reprioritize. Recorded and summarized for stakeholders who couldn't attend.
Always
On-call within reason
When something material changes that warrants a real conversation — a new AI source, a spec break, a partner ask — you get the call. Not a ticket, a call.
— fit
Who this is for
Good fit: Live infrastructure already running, multi-property or multi-channel exposure, regulated industries, businesses where a 20% drop in AI-referral traffic would actually hurt the quarter. Most retainer clients started with implementation, but I'm happy to take over infrastructure someone else built — we'll just start with a one-time baseline audit so I can see what I'm signing up for.
Not a fit: "We just launched and want help building" — that's implementation, not retainer. Or "we want one-off help every few months" — that's an annual audit refresh, which I'm happy to do as a fixed-price engagement instead.
Pricing: a single-site retainer with light monitoring runs $1,500/mo. Multi-property businesses with active development needs land at $5,000–$10,000/mo. Six-month minimum, after which it converts to month-to-month.
— FAQ
Common questions
Questions people ask before starting a retainer.
Did you have to build it for you to maintain it?
No — happy to take over infrastructure built by another vendor or your in-house team. We'd start with a one-time baseline audit so I have a clear picture of what's running and where the risk lives. After that, the retainer runs the same way regardless of who built the original system.
What's the minimum commitment?
Six months. Three is too short to demonstrate value through the ecosystem's drift cycle, and twelve feels heavy for a first engagement. After six months it converts to month-to-month with 30 days' notice on either side.
What's in scope vs. out of scope?
In scope: monitoring, schema and spec updates, performance optimization, MCP server maintenance, attribution upkeep, direct access. Out of scope: net-new module builds (e.g., adding a brand-new booking layer that didn't exist), full site rebuilds, or major refactors. Those get quoted separately and the retainer is paused or credited as makes sense.
Can I pause if priorities shift?
Yes, after the six-month minimum. Pauses are month-by-month. The infrastructure keeps running, but monitoring and maintenance stop until we resume.
How is this different from implementation?
Implementation is the one-time build. Retainer is what comes after — monitoring it, keeping it current, fixing the small things before they become big ones, and being on-call for the strategic questions. Different work, different cadence, different pricing model.
Do you offer SLAs?
Soft SLAs at the standard rate (same-day response on business days, monthly reports on the first business day). Formal contractual SLAs (e.g., 4-hour response, 24/7 on-call) are available on a custom-priced retainer for regulated businesses or those with formal procurement requirements.
Let's talk about fit
Not every business needs one of these.
If your infrastructure is stable and your priorities are elsewhere, an annual audit refresh might serve you better than a monthly retainer. We can figure out which fits on a quick call — and I'll tell you honestly which one I'd pick if I were in your shoes.
I respond personally within 1 business day. No pitch — just a real conversation.