Every Agency Course Teaches You How to Get Clients. None Teach You How to Keep Them.
There are hundreds of resources on how to start a Google Ads agency. How to find clients. How to pitch. How to close. The acquisition side is well-documented.
The operations side - what happens after you sign the client - is a black hole. Most agency owners figure it out through trial, error, and expensive mistakes. We certainly did.
After managing 160+ client relationships across $184M+ in spend, here's what we learned about the operational systems that actually determine whether an agency grows or burns out.
None of this is glamorous. All of it matters more than your next pitch deck.
The Audit-as-Sales-Tool: Why Free Audits Don't Close Deals
Most agencies offer a "free audit" to win new clients. They look at the account for 30 minutes, write up some general observations, and present them in a sales call.
This approach has two problems.
First, generic observations don't demonstrate competence. "Your campaigns could be better structured" tells the prospect nothing they didn't already suspect. It doesn't show them you know something their current setup doesn't.
Second, "free audit" positioning signals desperation. You're giving away work to earn the right to pitch. The prospect knows it. You know it. The dynamic starts wrong.
What works instead: a scored, systematic assessment.
We run every prospect through a 100-point account health scoring rubric before writing a proposal. Five categories, 20 points each:
Conversion Tracking Integrity (0-20): Are all key actions tracked? Is Enhanced Conversions active? Are there duplicate conversion actions inflating numbers? Is the attribution model appropriate? Is GA4 linked and consistent?
Campaign Architecture (0-20): Segmented by intent? Clear budget allocation? Campaigns competing against each other? Logical structure for the business model? Ad types diversified?
Feed Quality - Ecommerce (0-20): Title character count, custom label usage, description completeness, GMC disapproval rate.
Keyword and Search Term Quality (0-20): Active negative keyword management, weekly search term reviews, match type appropriateness, evidence of a keyword research process.
Creative and Copy Quality (0-20): Complete ad assets, evidence of copy testing, landing page relevance, extension coverage, testing cadence.
The score tells the story. Below 40: critical issues. 40-59: significant waste. 60-79: material structural problems. 80-100: reasonable management with room for efficiency gains.
Why this works: prospects argue with opinions. They don't argue with a scored rubric they can see. "Your account scored 47/100" hits differently than "we think we can do better."
After 160+ account relationships, scored audits close at 2-3x the rate of narrative audits. The mechanism is simple - a score forces specificity, and specificity demonstrates competence in a way that general observations never can.
Client Qualification: The Scorecard That Saves Months
Not every client is worth taking. This lesson costs most agency owners at least one terrible quarter before they internalize it.
Bad clients aren't just unprofitable. They're contagious. They consume disproportionate time, create stress that bleeds into other accounts, and attract referrals who are similarly difficult. One wrong client can undermine an entire quarter of growth.
We score every prospect before investing time in an audit or proposal. Takes 15 minutes. Saves months.
Ad Spend Level (0-15 points): Below $3K/month scores 0 - not viable at any reasonable service level. $3-8K scores 5 - marginal, flat retainer only. $8-25K scores 10. $25K+ scores 15.
Product-Market Fit (0-15 points): Does the product actually sell? Check reviews, Trustpilot, their site. You can't run good ads for a bad product. Every agency owner has this scar.
Decision-Maker Access (0-15 points): Can you talk to the person who approves budget and strategy changes? If the answer is "our marketing coordinator will relay feedback," the engagement will be painful.
Realistic Expectations (0-15 points): "I want 10x ROAS on a $3K/month budget in a competitive category" scores 0. A prospect who understands that growth takes time and investment scores well.
Technical Readiness (0-15 points): Website loads fast, tracking is implementable, they have product data in some manageable format. Accounts requiring 60 hours of technical setup before any optimization starts need different pricing.
Communication Style (0-15 points): During the discovery call, do they listen? Do they ask good questions? Do they respond to emails within 48 hours? The discovery call is the client at their best. If communication is already strained, it only gets worse.
Total (0-100): Take clients scoring 70+. Seriously evaluate 60-69. Decline below 60.
We turn away most inquiries now. Not from a position of arrogance - from experience. The engagements that started below 60 on this scorecard cost us more in time, margin, and stress than the revenue was worth. Every single time.
Pricing That Protects Margins
Three pricing models exist for Google Ads management. Each has a context where it works and contexts where it quietly kills your margins.
Percentage of Spend (10-20%)
Works at $15K+/month spend. Your margins scale with the client's investment. Incentives are aligned - you both benefit from increasing budget when it performs.
Breaks at low spend. At 15% of $5K/month, you're collecting $750. That doesn't cover the time required to manage the account properly. You either underservice or lose money.
Flat Retainer
Works at $3-15K/month spend where percentage pricing would undercompensate you. Predictable revenue. Predictable costs. Clean math.
Breaks at high spend. A $100K/month account at a $5K flat retainer means you're managing 20x the complexity of a $5K account for the same fee. Your effective hourly rate collapses.
Hybrid (Base + Performance)
Works when you have high confidence in the outcome and an established relationship. Base covers operations. Performance component rewards results.
Breaks on new relationships where you don't know the account, accounts with attribution issues you can't control, and volatile industries with external factors affecting performance.
The decision tree we use:
Under $8K/month spend: flat retainer ($1,500-$3,000). $8-25K: flat or percentage, depending on complexity. $25-100K: percentage (12-15%) or hybrid. Over $100K: hybrid with carefully negotiated performance baseline.
The presentation order matters. Show the prospect the opportunity value from your audit before showing the price. If they've seen "$4,100/month in recoverable waste" before they see "$3,000/month retainer," the ROI math is obvious without you having to sell it.
Scope Definition: The Section That Saves Your Margins
Eighty percent of scope creep is preventable with one section in your proposal: the explicit "not included" list.
Most proposals describe what you will do. Almost none define what you won't do. That gap is where margin erosion lives.
Included in this engagement:
- Google Ads campaign management (Search, Shopping, PMax as applicable)
- Monthly performance reports (delivered by the 5th)
- Bi-weekly check-in calls (30 minutes)
- Negative keyword management (weekly)
- Ad copy testing (minimum 2 active tests per campaign)
- Feed optimization for Merchant Center (up to 2 passes per month)
Not included:
- Meta Ads, LinkedIn Ads, or any non-Google channel
- Creative production (video, photography, design)
- Landing page design or development
- Email marketing or CRM integration
- SEO or content strategy
- Attribution platform setup
When a client says "can you also just...," you have a professional, non-adversarial response: "That falls outside the current scope. Here's what adding it would look like."
Without this list, "can you also just" slowly expands your workload by 30-40% without expanding your revenue by a single dollar. We've seen agencies die from this. Not from losing clients - from winning too many without protecting their margins.
Weekly Optimization: Checklists Over Intuition
The most common "optimization" process at agencies is opening the dashboard, looking around, and adjusting whatever catches your eye.
This is how experienced marketers miss obvious problems. You default to checking what you checked last week. You gravitate toward the campaigns you find interesting. You skip the tedious stuff because nothing looked wrong last time.
Checklists fix this. Not because they're smarter than you. Because they're consistent when you're not.
Our weekly optimization checklists by tier:
Tier 1 (22 points): Budget pacing against monthly target. Bid strategy alignment check. Search term report review with negative keyword additions. Impression share trends on priority campaigns. Creative performance (CTR, conversion rate by ad). Feed quality score (for ecom). Competitor activity via auction insights. Landing page performance by campaign. Geographic performance review. Device performance split. Dayparting analysis. Audience signal review for PMax. And 10 more covering conversion tracking health, quality score trends, and ad extension performance.
Tier 2 (14 points): Budget pacing. Bid strategy validation. Negative keyword management. Creative rotation check. Feed basics (ecom). Conversion tracking health. Impression share on top campaigns. Search term review. Performance against targets. Plus 5 structural checks.
Tier 3 (8 points): Budget pacing. Conversion tracking health. Search term negatives. Performance against targets. Creative freshness. Feed approval status. Bid strategy check. Monthly trend comparison.
The discipline is doing it the same way every week. Week 47 gets the same rigor as week 1. The checklist ensures that.
Reporting: Match the Depth to the Client
A $5K/month client does not need a 30-page performance deck. They need 2 pages with clear actions and a 5-minute read.
A $100K/month client needs depth - competitive positioning, trend analysis, strategic recommendations, and forward-looking projections.
Our reporting templates by tier:
Tier 1: 8-10 page deck. Performance summary. Campaign-level breakdown. Competitive intelligence. Trend analysis. Strategic recommendations. Testing results. Next 30-day plan.
Tier 2: 4-5 page deck. Performance summary. Top campaign highlights. Key actions taken. Key actions planned. Questions for the client.
Tier 3: 2-page summary. Performance vs. targets. Actions taken. Actions planned. One recommendation.
Every template is pre-built. Data pulls are automated where possible. An account manager doesn't build a report - they review a generated draft, add commentary, and send it. Report generation time dropped from 2-3 hours per account to 20-30 minutes.
Over 50 accounts, that's the difference between 125 hours per month on reporting and 20 hours. That 105 hours goes back into strategy and optimization.
The Escalation Framework
When something goes wrong - and things go wrong regularly in Google Ads - the response speed matters more than the response quality. A good response 6 hours after a problem starts beats a perfect response 3 days later.
Escalation triggers we've documented:
Immediate (same-hour response): GMC account suspension. Conversion tracking failure. Budget overspend by 50%+. Client account access revoked.
Same-day response: ROAS drops below floor for 48+ hours. CPA exceeds target by 30%+ for 3+ days. Impression share drops 25%+ week-over-week on priority campaigns.
Next-review response: Gradual performance trends (5-10% shifts). New competitor entering auction. Seasonal pattern changes.
For each trigger, the SOP defines: who gets notified, what the first diagnostic step is, what the communication to the client looks like, and what the resolution timeline expectation is.
This removes the "should I tell the client about this?" ambiguity. If it hits the trigger, the communication sequence starts. No judgment calls about whether to mention a bad week. Transparency is the default.
What Nobody Tells You About Agency Operations
The best system is the one your team actually follows. We've built beautiful SOPs that nobody used because they were too complex. The winning SOPs are simple enough to follow on a bad day, detailed enough to catch the important stuff, and short enough that people don't skip to the end.
Client retention is an operations problem, not a sales problem. The agencies with the best retention aren't the ones with the best results. They're the ones with the most consistent communication, the fastest response to problems, and the clearest reporting. Clients leave agencies that go quiet, not agencies that have a bad month.
Your first 10 clients teach you how to serve the next 50. Every painful client relationship reveals a gap in your operations. Wrong expectations? Your qualification scorecard needs a new factor. Scope creep? Your proposal needs a clearer "not included" list. Missed a campaign issue? Your weekly checklist needs another point.
Documentation compounds. Every SOP you write today saves time forever. Not dramatically - maybe 15 minutes per week per SOP. But 50 SOPs saving 15 minutes each is 12.5 hours per week. Over a year, that's 650 hours. That's a full-time hire you didn't need to make.
The Honest Bottom Line
Agency operations isn't a topic that trends on Twitter. Nobody builds a following by posting about their client qualification scorecard or their escalation framework.
But the agencies that grow past 20 accounts, past 50 accounts, past the point where the founder's personal capacity is the ceiling - they all built this infrastructure. Some documented it. Most learned it through expensive mistakes.
We spent 3 years building the operational system behind Tegra. 10 modules. 52 production commands. 12 contract schemas. 152-point quality scoring.
It's the most boring competitive advantage we have. And it's the only reason a team of 3 can do what most agencies need 15 people to do.
Operations isn't the exciting part of running an agency. It's just the part that determines whether the agency survives.
Gate Scores: insight:11/15 | hook:8/11 | viral:8/10 | authority:5/5 | entertainment:7/10 | info_density:7/10 | composite:7.6
Ruslan co-founded Tegra in 2017. Runs the Google Ads practice - feed, PMax, search, attribution. Writes weekly about the parts of paid search operators are afraid to touch.