Agency Growth, Sale Prep & Why Agencies Fail

Profit over revenue, recurring revenue and retention, sell ROI not traffic, instrument what you cannot measure.

At the 2026 conference, the agency operators and acquirers converged on one idea: the work you sell is being commoditized, so the durable value is the company behind it (its profit, its retention, and whether anyone but you can run it). Read this page as the business layer that sits under the rest of the Playbook: it tells you what to instrument, what to keep recurring, and what to professionalize whether you plan to sell or to own for the long haul. Use it alongside your monthly cadence and the strategy page to turn delivery into an asset, not just a job.

SEO Spring Training 2026 was, on the surface, an SEO conference, but the business-of-the-agency thread ran through every block of both days. With AI commoditizing the deliverable (Roof, Buckner, and Eldar all argued the old playbook is decaying), the speakers who drew the most attention were the ones talking about the company behind the work: what an agency is actually worth, why profitable-looking agencies quietly bleed out, how to make clients stay, and what it takes to sell. Four sessions converged on a single uncomfortable idea: most agency owners are skilled technicians who never became operators, and the gap between those two roles is where agencies stall, fail, or sell for nothing. This brief synthesizes across Chris Martinez and Chase Buckner (day 1 part 2), Jacob Kettner (deck-only), Adam McChesney and Eldar (day 2), and Marty Marion (day 1 part 1).

The through-line

1. A CEO's job is to grow enterprise value, and most owners are technicians who never made that switch. Martinez framed the owner's one job as increasing the value of the company year over year, and said if you cannot answer the financial questions you are not a CEO yet, you are the SEO or designer or ads expert who happens to be a CEO. The day 2 moderator (Chris, last name not given) made the same point with The E-Myth Revisited (Michael Gerber): about half the room raised a hand when asked if they were a better marketer than business person, and he tied it back to Martinez's enterprise-value-and-EBITDA framing. Kettner is the operational version of the same diagnosis: good agencies fail not because they are bad at SEO but because they fly blind on their own numbers.

2. Profit, not revenue, is the only number that matters, and revenue hides the truth. Martinez: "revenue is vanity, profit is sanity, all that matters is what you keep," with EBITDA (not revenue) the metric a sophisticated buyer cares about. Kettner showed the mechanism by which top-line growth masks bottom-line bleed: his Web Dev department ran at 198% fulfilment cost for eight months while the healthy SEO department subsidized it through the top line, and nobody could name the moment it broke. Adam reports a real profit number alongside revenue (2.174M in 2025 at about 26% profit), treating margin as the headline rather than the run rate.

3. Retention and recurring revenue are what make an agency both survivable and sellable. Martinez wants roughly 80% monthly recurring revenue, low churn, and no client over 20% of revenue, because project work and concentration each cost "a turn" on the multiple. Buckner's whole Growth Flywheel argument is that selling only traffic makes you replaceable, while delivering more of the flywheel cements clients into higher-ticket, longer-term recurring revenue, which is also exactly what acquirers want to see. Adam built the same thesis into his org chart: client experience as the center of gravity, about 3% churn on roughly 100 clients, because clients rarely leave agencies that deliver ROI, are embedded in their systems, and treat them better than anyone else.

4. Sell ROI and outcomes, not traffic or vanity metrics, or you are a commodity. Adam refuses to sell on keyword rankings, impressions, or clicks; his "Results Matter" value states clients pay for ROI, not vanity metrics. Buckner's framing is that clients want "the house" (money, growth), not "the tool" (traffic), so agencies that sell only traffic are dangerously replaceable. Marion supplies the strategy underneath this: "better" is the most dangerous word in marketing, and you win by changing the criteria of evaluation (a category-class distinction) rather than claiming to be better at the same thing.

5. AI is collapsing the moat under the basic deliverable, so the defensible value moves up the stack. Eldar argues the 2024 and 2025 playbook is a losing strategy and visibility now lives across LLM engines (ChatGPT, Gemini, Grok, Perplexity, Claude). Buckner warns that website-building as a business is being commoditized by AI builders. The response across speakers is to move up the stack: Adam into client experience and systems integration, Buckner into the full flywheel, Martinez into a sellable enterprise, Kettner into operational instrumentation.

6. You cannot delegate, scale, or sell what you cannot measure. Kettner's leadership reframe is that he built his financial model not for better reports but so someone else could run the company, because "trust without measurement is just hope." Martinez's entire sale-prep checklist (clean financials, an operator who can run the business after you leave, documented CAC and churn and LTV) is the same idea aimed at a buyer instead of an internal hire: a business nobody can diagnose or run without you is a job, and "nobody wants to buy a job." Adam operationalizes it by giving every department measurable structure and promoting an operator (COO Alejandra) and a CX manager (Laura) who can run their layers.

Tactics & playbook

Sale prep (Martinez):

• Compute your real exit number, not the headline: write the price at which you become a seller, cut it in half (cash at close vs earn-out), reduce about 40% for worst-case taxes on the cash at close, then live on that through a non-compete of up to five years.
• Reframe the goal as a retirement number = annual need divided by 5.5% (e.g. 200K per year is about 3.64M), which the agency income should hit independent of any exit. The exit just gets you there faster.
• Know the EBITDA multiple bands: under 1M EBITDA = 2 to 4x (most likely 3x), 1M to 2M = 3 to 5x, 2M+ = 5 to 8x.
• Hold at least 3 (ideally 4) years of clean financials, and provide revenue and COGS per line of business, revenue per client, and revenue concentration.
• Keep no single client above 20% of revenue, target about 20% YoY growth, and about 80% monthly recurring revenue, or you lose "a turn" on the multiple.
• Install an operator (your number two) who can run the business after you leave; de-risk the team handoff with a "kiss goodbye" bonus and an earn-out tied to key people staying. Leadership must be US-based, full-time, W2.
• Track CAC, churn per line of business, and LTV (LTV = (1 / monthly churn as a decimal) months times average monthly revenue; 3% churn implies about 33 months).
• Run the full sell-side process in order (financials, valuation estimate, data room, attorney, CPA/tax plan, teaser, NDA, CIM, management meetings, IOI, exclusive LOI, then about 90 days of due diligence) and keep hitting your forecasts during diligence, because missing a 60-day forecast drops the price.

Operational and financial instrumentation (Kettner):

• Set an internal unit rate (per hour, per link, per citation, per 1,000 words) as the floor under every proposal, even if you sell on value. "Value-based pricing tells you how high; an internal rate tells you how low you can't go."
• Treat revenue as six buckets every price must pay rent on: fulfilment, overhead, sales, marketing, account management, profit. Set profit first and fit everything else around it.
• Price for the team you are growing into, not the team you have. Using a 50%-fulfilment formula (Price = cost / 0.50) at 30-person scale where overhead has grown silently costs about 15 points of profit; reprice to cost / 0.35.
• Diagnose the three profit killers separately, because they look identical on the P&L: pricing (structure wrong, fix by repricing, never by selling more), utilization (billable hours below target, often fixed by selling more), and over-delivery (delivered hours exceed budget, fix by scope and renewal repricing).
• Account for the "utilization tax": dividing cost by target assumes 100% billable, which is impossible, so a naive 110/hr rate is really 200/hr at realistic 55% utilization. Small teams sit in a structural red zone and must price for the midpoint utilization of the next team size up.
• Know your break-even revenue before every hire; expect each hire to pull you underwater for roughly six months until you sell into the new capacity.
• Build the instrumentation bottom-up: Layer 1 tracking in your PM tool, Layer 2 budgeting per client, Layer 3 a financial model reviewed every Monday morning. Start at Layer 2, not at the model.

Client experience and embedding (Adam McChesney):

• Make client experience the center of the org chart (a COO, a CX manager, and dedicated CX specialists each carrying roughly 20 accounts), not a side function.
• Run a fixed retention cadence: kickoff on day 1 (handoff from sales to CX), automated weekly check-ins, bi-weekly or monthly meetings, and a quarterly business review for deep-dive ROI and roadmap.
• Embed into the client's stack so leaving is hard: a CRM department that builds CRMs for clients who lack one, and a dedicated integrations specialist who wires call and lead tracking into existing tools (e.g. ServiceTitan via Zapier) so every job is attributed to its source.
• Expand revenue through upsells, cross-sells (SEO to Ads to Social to Web), and referrals, the "most underrated growth strategy"; turn happy clients into a flywheel of personal brand plus authority plus CX driving about 3 inbound leads per day.
• Solve the home-services attribution lag with a custom dashboard that credits a February lead closing in April across months, so ROI is visible over the last 6 months.

Growth flywheel and AI leverage (Buckner):

• Deliver more of the flywheel (Capture, Nurture, Close, Evangelize, Reactivate) rather than traffic alone, charging more and retaining longer the more you deliver (or have a commissioned partner deliver).
• Beat the five-minute lead-response shot clock (the cited average human response is 3 hours 26 minutes) with AI chatbots and voice AI that answers the calls humans miss.
• Add a trackable, textable number and a chat widget that redirects to SMS, automate reviews (you need 100+ to be competitive against an average of 39), and run reactivation campaigns over SMS (98% open rate) as "sales on demand."

Positioning and acquisition (Marion):

• Treat acquisition as displacement, not demand creation: 95%+ of a market already buys from someone, so de-position the incumbent (insert doubt or dissatisfaction) before positioning yourself.
• For agencies competing against incumbents, arm the prospect with an unavoidable question to ask their current provider (e.g. "show me the test logs and the revenue impact"), surfacing the incumbent's "trust us" accountability gap.
• Win by changing the evaluation criteria (a category-class distinction), since competing on "better" loses to familiar, easier, and safer defaults.

Tensions & disagreements

• Sell the business vs build the dream business you never sell. Martinez treats the agency as an asset to professionalize and exit, with the entire talk organized around preparing for a sale. Adam's arc runs the other way: after a partnership exit that went badly (a buyout offer that never came, then being fired), he rebuilt an agency he frames as the one he wants to keep and operate, "Profit, Freedom, Scale" rather than a sale. The two are not strictly contradictory (a sellable business is also a good one to keep), but the orientation differs: build-to-sell vs build-to-own.

• "Sell more" as the answer. Buckner's flywheel and reactivation tactics push agencies to sell more and capture more leads as the growth lever. Kettner explicitly warns that selling more is the wrong fix for two of his three profit killers (pricing and over-delivery), where more volume accelerates the bleed. They are addressing different situations (Buckner assumes the unit economics work; Kettner assumes they may not), but an owner taking both talks at face value would get opposite instructions about whether more sales fixes the problem.

• Value-based pricing. Marion's whole approach prices to outcome and ROI, and Adam sells on ROI rather than cost. Kettner agrees value sets the ceiling but insists the "I charge for value, not hours" story is often an excuse owners use to avoid knowing their internal numbers. The reconciliation he offers: value tells you how high, an internal rate tells you how low you cannot go, but the tension with a pure value-pricing camp is real.

• Where the mid-market lands. Martinez says agencies charging roughly 1,000 to 3,000 per month are getting squeezed, with very low-ticket and high-ticket doing better. Kettner's model is built precisely for a mid-market services agency at about 300K per month and argues it can be made structurally profitable through pricing and utilization discipline. One reads the squeeze as a market position to escape; the other reads it as an operations problem to solve.

Sources (conference sessions)

Conference session references, not pages on this site:

• Day 1 part 2 (Chris Martinez on agency M&A and sale prep; Chase Buckner on the Growth Flywheel and retention; Kyle Roof's AI-content portion is covered in the AI search-visibility brief)
• Jacob Kettner, Why Good Agencies Fail: operational and financial systems (deck-only)
• Day 2 (Adam McChesney on building the agency of your dreams via client experience; Eldar on AI visibility; Chris the moderator on technician vs business owner)
• Day 1 part 1 (Marty Marion on de-positioning and acquisition; Austin's mindset talk and Roof's SEO testing are covered elsewhere)

Connect it to your system

This business layer plugs into the rest of the Playbook through cadence, positioning, and the work you actually deliver.

• The strategy page is where the build-to-sell vs build-to-own decision lives; let it set the orientation before you pick which metrics to professionalize here.
• Run the retention and ROI-review cadence through your monthly cadence and weekly cadence pages, so QBRs, check-ins, and forecast tracking are scheduled rather than ad hoc. Map the Martinez sale-prep metrics (recurring revenue, churn, LTV, CAC) and Kettner's cost-center model onto whatever you already review there.
• Standardize the day-1 handoff and embedding that Adam's client-experience model depends on through your onboarding playbook.
• Marion's "sell ROI, not vanity metrics" and "change the criteria of evaluation" connect to the demand-and-conversion thinking on digital PR, conversion, and demand.
• The AI commoditization pressure that pushes value up the stack is covered in AI search visibility and AI agents and automation; the response-time and reactivation tactics in Buckner's flywheel live closest to the automation side.