Agency evaluation checklist

You want a partner who delivers, not one that sells hope. Treat the evaluation like hiring a key exec.

• Ask for raw campaign data, not glossy slides. Demand the spreadsheets with traffic, costs, contact rates and pipeline attribution.
• Require a work sample. A 3-week paid brief where they diagnose and outline the first 90 days reveals more than promises.
• Check a current client reference for one thing only: what went wrong and how quickly they fixed it.

Strategic fit and specialization

Generalists are comfortable with everything and expert at nothing. Match specialization to the problem, not company size.

• If your product sales cycle is 6 months plus, favor agencies with experience in long sales cycles and content that nurtures.
• If you need rapid paid growth, choose vendors who can show cost-per-opportunity math and creative scale, not just click growth.
• Example: a product marketing problem often needs a strategist with category experience. Demand gen problems need operations-first teams.

Proof and performance validation

Proof is specific and reproducible. Stop accepting claims like "tripled pipeline" without context.

• Ask for baseline KPIs and the exact lift, with attribution model used. Insist on seeing the raw touchpoint timelines that produced the pipeline.
• Request a cohort analysis: how did leads sourced by the agency convert through the funnel over a defined period.
• If they refuse raw data, treat that as a red flag. Insist on contract language allowing audits of campaign data.

Team structure and capabilities

Titles alone mean nothing. Look at who will actually do the work.

• Insist on names and CVs for the core team. Not just an account director, but the content lead, media buyer, and analytics owner.
• Check for split roles that cause bottlenecks. One person owning creative and analytics is a recipe for slow learning.
• Ask about bench strength. Who covers when a senior leaves mid-engagement?

Tech stack and integrations

Compatibility matters. Integration gaps create manual work and reporting mismatches.

• Map their stack to yours: CRM, marketing automation, analytics, CDP. Ask for a simple integration plan and data schema.
• Avoid partners who want to force their own martech as the only option. That often means vendor lock and hidden fees.
• Confirm what they will own in your systems and what they expect you to own. This prevents scope creep.

Pricing model comparison

Price should reflect the objective, not an arbitrary hourly rate.

• Fixed monthly retainer works for steady-state work. Performance fees make sense for clear, attributable outcomes.
• Hybrid models often align incentives: a base retainer plus a variable tied to qualified pipeline.
• Insist on a clear change-order process for out-of-scope requests.

KPI and SLA expectations

Set these early. Vagueness kills.

• Define KPIs by funnel stage, not vanity metrics. Examples: MQL to SQL conversion rate, cost per qualified opportunity, opportunity velocity.
• Create SLAs for response times, campaign builds, and reporting. Example: 48-hour response for critical incidents, 5 business days for new creative.
• Tie a portion of variable fees to KPI achievement, but use reasonable windows. Results take time.

Core B2B service packages

Don’t buy services by label. Buy outcomes and a clear path to them.

Demand generation and pipeline

This is operations work with creative sprinkled on top.

• Demand gen must include campaign flows, lead scoring and lifecycle rules. If these aren’t in the proposal, it is incomplete.
• Require playbooks for top-of-funnel, mid-funnel nurturing and handoff to sales. Playbooks should include timing, content types and expected conversion rates.
• Example mini-scenario: for a 90-day push, expect ramped targeting, a mid-campaign creative refresh, and progressive profiling to reduce form friction.

Account-based marketing (ABM)

ABM is coordination, not magic.

• ABM needs tightly integrated sales and marketing workflows, plus named-account content. Demand to see account lists and intent signals they will use.
• Measurement must include account engagement scoring and influenced pipeline, not just leads.
• ABM pilots should run for 3 months minimum to test messaging and list hygiene.

Product and content marketing

Content should move deals, not just fill channels.

• Ask for evidence that content has shortened deals or improved win rates. The right piece of content reduces friction in the purchase process.
• Content plans must align to buyer stages and include enablement assets for sales.
• Include rep feedback loops. Sales should be able to flag content that closes or stalls deals.

Paid media, SEO and CRO

These three must be treated as a stack, not separate silos.

• Paid media without landing page conversion optimization is wasted budget.
• For SEO, prioritize technical fixes and content tied to buyer intent, not generic top-funnel traffic.
• Require A/B testing calendars for landing pages and creative with clear success criteria.

Marketing operations and analytics

This is where campaign promises become measurable outcomes.

• Ops should own runbooks, data hygiene, naming conventions and attribution models.
• Ask for a regular cadence of analysis that includes funnel leakage, creative performance and forecast accuracy.
• Demand a clear escalation path when data or integrations break.

Match agencies to company stage

Different stages need different muscle.

Startup and early growth fit

You need speed, experiments and a tight budget.

• Look for an agency that can move fast on paid tests and content that drives early lead flow.
• Prefer short pilots and clear stop criteria. If something doesn’t work in 6 weeks of proper testing, pivot.

Scale-up and expansion fit

This is about systemizing and doubling down on winners.

• Expect process creation: lead routing, predictable content production, sales enablement programs.
• The agency should be comfortable building repeatable funnels and handling budget scale without losing efficiency.

Enterprise and global fit

Complexity and governance become primary concerns.

• They should demonstrate multi-region operations, compliance experience and playbooks for stakeholder sign-off.
• Look for explicit plans to align with procurement, legal and data teams.

Industry needs: SaaS, tech, healthcare

Industry verticals change constraints.

• SaaS: focus on free-to-paid conversion mechanics and trial onboarding content.
• Tech: product marketing and developer evangelism matter. Look for technical content capabilities.
• Healthcare: compliance and privacy dictate every asset and tracking method. Demand HIPAA and regulatory experience.

Procurement, onboarding and governance

Make vendor management boring. That is good.

RFP checklist and scoring rubric

Keep it short and objective.

• Required deliverables, team CVs, tech integrations, 3 examples with raw data, price and timeline.
• Scoring rubric sample (100 points):

- Capability fit 30 - Proof and data 25 - Team and availability 15 - Tech fit 15 - Price and terms 15

Onboarding milestone checklist

Don’t assume onboarding will be quick.

• Week 0: legal, access provisioning, kickoff.
• Week 1: audit of current martech and past campaigns.
• Week 2-3: build first campaign assets, set tracking and baseline dashboards.
• Week 4: launch and immediate QA.

Reporting cadence and stakeholder map

Match cadence to decision-making rhythm.

• Weekly tactical sync for the delivery team.
• Biweekly performance review with marketing ops and revenue lead.
• Monthly strategy review with executives.

Sample 90-day plan with owners

• Day 1-7: kickoff, access setup, baseline audit — Owner: Agency PM and Marketing Ops
• Day 8-21: creative and campaign build, tracking, QA — Owner: Creative Lead and Analytics Engineer
• Day 22-30: soft launch, quick learnings, fix issues — Owner: Media Lead
• Day 31-60: full ramp, A/B tests, sales enablement deliverables — Owner: Demand Gen Lead and Sales Ops
• Day 61-90: cohort analysis, optimization roadmap, forecast update — Owner: Analytics Lead and VP Marketing

Measurement, ROI and optimization

Measure to decide, not to justify.

KPI framework by objective

Pick 3 KPIs per objective and be ruthless.

• Awareness: share of voice, engaged sessions, account reach
• Demand: cost per qualified opportunity, MQL to SQL conversion, pipeline influenced
• Growth: win rate lift, average deal size, sales cycle length

Attribution, forecasting and benchmarks

Be explicit about how you count things.

• Use a consistent attribution model and document when you switch models. If you mix models, report both last-touch and multi-touch influenced pipeline.
• Forecasting should be probabilistic. Don’t accept single-point forecasts without confidence bands.
• Benchmark against your historical performance, not industry vanity numbers.

Testing and iteration roadmap

Treat optimization like product development.

• Run structured experiments: hypothesis, variant, sample size, timeline, success metric.
• Limit concurrent tests on the same funnel stage to avoid interference.
• After a test, update the playbook. If a change fails, record why and what you tried next.

Dashboard metrics: weekly vs monthly

Weekly metrics show operational health. Monthly metrics show trend.

• Weekly: lead volume, cost per lead, creative CTR, delivery issues, top 3 blockers
• Monthly: qualified opportunities, pipeline influenced, conversion rates by stage, forecast vs actual

Negotiation and red flags (United States)

Protect your operations and IP.

Contract clauses to require

Be explicit and legal about the obvious.

• Scope of work with deliverables, timelines and acceptance criteria.
• Data ownership clause saying you own the data and final creative assets.
• Security requirements and audit rights.
• Termination for convenience with transition assistance.
• Liability limits and confidentiality.

Pricing negotiation tactics

Get creative. There is more than one lever.

• Start with a pilot project with a capped fee and clear success metrics. Scale on success.
• Propose a retroactive bonus for performance above specified thresholds, not a punitive clawback.
• Negotiate caps on incremental ad spend fees. Agencies sometimes charge percentages of media spend.

Deal-breakers and warning signs

Know when to walk.

• They refuse to provide raw performance data.
• No named team members or continuous staff turnover without mitigation plans.
• Vague answers on security or compliance.

Data security, compliance and IP

Don’t let "we handle security" be the final word.

• Require documentation of compliance certifications relevant to you, for example SOC 2 or HIPAA where applicable.
• Specify data deletion and export procedures on termination.
• Insist on encryption in transit and at rest for any hosted assets containing customer data.