Listicles call themselves "best."
Buyers know better.

Run four parallel checks. First, evaluate the operator stack — who actually does the work day-to-day, and how senior are they? Junior-led engagements at senior pricing are the most common bad outcome. Second, ask for two reference calls with current clients in similar revenue bands and competitive categories — not the agency's all-time best case studies. Third, request the audit deliverable that comes before the engagement; the quality of the audit is the single best predictor of the engagement quality. Fourth, look at the founder's published work — books, original-data studies, conference talks, contributor bylines on Forbes or Search Engine Land. Operators who publish methodology have a public reputation that disciplines their work; operators who don't publish have nothing forcing accountability beyond the contract.

Depends on what you're buying. Big agencies (100+ headcount, WebFX-scale) buy you process maturity, deeper bench coverage if a person leaves, and the ability to scale execution at large keyword footprints. The trade-off is that the senior strategist who pitched you usually doesn't run your account day-to-day. Boutique senior-led agencies (5-30 headcount) buy you direct access to the operator who decides strategy, faster decision velocity, and more customised work — at the cost of less bench depth and more concentration risk. The honest version: if your engagement is straightforward and your budget is mid-market, a big agency works fine. If your category is competitive, your business is between $1M and $10M revenue, and you want senior judgement on the line, a boutique is usually the better trade.

Methodology beats vertical specialisation in most cases — and the few exceptions are easy to spot. Vertical specialists make sense in heavily regulated industries (legal, medical, financial advice) where compliance language and trust signals matter more than SEO craft. Outside those, vertical specialists tend to ship the same playbook across every client in the vertical, which produces mediocre results because everyone in the vertical is running the same playbook. Methodology-led operators (who publish their approach openly and adapt it to your specific situation) compound better. The exception inside the exception: a boutique that's done five engagements in your specific vertical with public results may have learned something a generalist hasn't. Ask for the case studies and judge the depth.

When you can verify them, when the client size matches yours, when the timeline is recent (last 24 months), and when the metric is tied to revenue rather than vanity. A 2,414% organic traffic growth case study is meaningful if the client is roughly your size, the time period is bounded, and you can talk to the client. The same number is meaningless if it's from a $50K-revenue startup that grew to $200K — that's a base-rate story, not a methodology proof. Specifically ask: what was traffic before, what was traffic after, what was the timeline, what was the engagement scope, and can I talk to the client. Agencies that won't put you on a reference call have a reason for not doing so. Agencies that hand you three references inside 24 hours are usually telling you the truth about their work.

Three terms matter. First, no minimum contract length — month-to-month or with a 30-day exit. Long contracts (12+ months no-exit) protect the agency, not you, and usually correlate with mediocre delivery because the agency knows you can't leave. Second, IP ownership of all content, audits, and assets created during the engagement, including draft files and ranking data — you should own everything you paid for. Third, a clear scope document that lists what's included in the retainer (content pieces per month, hours of strategy time, tool access) so future scope changes are explicit rather than implicit. The fourth thing isn't a contract term but a posture: the agency should be willing to fire you as a client if the engagement isn't working, and you should be willing to fire them. Mutual at-will is the healthiest structure.

Two reasons, only one of them legitimate. The legitimate reason: SEO pricing genuinely varies 5-10× across engagements based on competitive density, content velocity, technical scope, and market complexity, so a single published number would mislead more than it would inform. The illegitimate reason: hidden pricing lets the agency price-discriminate based on what they think your budget is, which favours the agency over the buyer. The honest middle ground is to publish ranges with the factors that move price (which we do at /how-much-does-seo-cost), then tighten the quote on a discovery call. Agencies that won't publish ranges at all are usually optimising for opacity, not for accuracy.

Commit to enough time to see compounding returns, but in a structure that lets you exit if the work is bad. The minimum useful engagement length for SEO is roughly six months — that's how long it takes for substantive content to mature, for technical retrofits to be crawled and re-indexed, and for authority work to register in third-party trust signals. So plan for six months minimum. But structure the contract as month-to-month with a 30-day exit clause, so the commitment is a planning horizon rather than a contractual lock-in. If the agency insists on a 12-month no-exit lock, walk. Good operators don't need contractual locks because their work retains clients on its own merit.

Three options in order of escalation. First, raise the issue specifically and in writing — name the deliverables that are missing, the metrics that aren't moving, and the response time issues. Most agencies will course-correct when given a clear list. Second, ask for a 30-day improvement plan with named deliverables and named owners; if those don't ship on time, the trust is broken and the engagement should end. Third, exit. A clean exit means you take possession of all IP, content drafts, audit files, and ranking data created during the engagement. The wrong way to handle it is silent dissatisfaction — you keep paying, the work keeps drifting, and 18 months later you're worse off than when you started. Naming the problem early is the only move that preserves the relationship or ends it cleanly.