How to Evaluate B2B SaaS Growth Marketing Agencies

Written by: Aaron Rovner, Founder, Saas Hero | Last updated: August 29, 2026

Key Takeaways

• Four structural shifts in platforms, measurement, staffing, and pricing created a gap between what B2B SaaS buyers need and what most agencies deliver.
• Agencies that optimize toward form fills instead of CRM-defined outcomes train bidding algorithms on the wrong audience, which inflates lead volume while pipeline stays flat.
• Post-click ownership determines whether an agency can improve the highest-leverage variable in the funnel, because landing pages drive conversion rate.
• Flat-fee retainers indexed to total ad spend remove incentive conflicts that percentage-of-spend models create when agencies recommend budget changes.
• SaaSHero is the only agency that meets every criterion in this framework; schedule a discovery call to see how your current account compares.

How to Evaluate Performance-Based Growth Marketing Agencies for B2B SaaS

Four independent conditions, not bad actors, created the current gap between buyer need and agency supply.

Platforms automated the lever-pulling and left data quality as the job. Manual bidding, keyword control, and placement selection now sit inside Smart Bidding, broad match, and Performance Max. What remains under human control is narrow: which conversion events the algorithm pursues, and how accurately those events proxy for revenue. An algorithm pointed at a form fill finds the people most likely to fill in forms, such as students, competitors, and job seekers, while reporting a falling cost per conversion. B2B SaaS advertisers who feed SQL and closed-won signals back to Google Ads via offline conversion tracking typically see SQL volume increase 30–50% at the same spend. The core skill in 2026 is choosing what the interface optimizes toward, not operating the interface.

The measurement layer broke before the ad platforms did. Third-party cookie restrictions, browser tracking prevention, consent requirements, and cross-device journeys each removed part of the path between a first impression and a signed contract. Few B2B SaaS companies have full pipeline attribution connecting ad spend to CRM revenue. Most still optimize on CPL, which reveals nothing about revenue outcomes. Boards now ask questions phrased in finance, such as CAC payback and pipeline coverage, and the reporting stack most companies have cannot answer them.

Mid-market teams are staffed for judgment and short on execution. A $10M–$50M software company typically runs two to four full-time marketers covering content, product marketing, events, lifecycle, and web. Nobody in the building has run a Google Ads account at scale, audited a search terms report, or configured offline conversion imports. The contractor layer fills the void: a freelance designer, a web developer for landing pages, and a campaign manager on the ad accounts. Each executes competently inside their own scope. Nobody owns the connections between them.

The standard agency scope stops at the click, and per-channel pricing holds it there. The conventional paid media retainer is scoped to the ad account. The landing page belongs to the client, the CRM to RevOps, and the conversion definitions to whoever configured the tag manager, often years earlier and often no longer at the company. Percentage-of-spend pricing creates an incentive misalignment because agencies earn more when clients increase ad spend, which conflicts with the client's goal of lowering cost per acquisition. Per-channel pricing holds the scope boundary in place, because testing a new channel raises the client's fees before it has returned anything.

The four shifts converge on one vacancy. Automation moved the work to data quality. Broken measurement moved the answer into the CRM. Mid-market teams hold the judgment but not the operators. The standard retainer stops short of the chain it is judged on. Nobody is accountable for the whole path from impression to CRM record.

Book a discovery call to find out whether your current agency owns the full chain, or whether you do.

15-Point Red-Flag Checklist Derived from Buyer Language

These structural shifts show up as specific, repeatable problems inside day-to-day agency relationships. The following conditions, drawn verbatim from the language buyers use when they arrive at a new agency search, each disqualify an incumbent relationship. Eight or more present simultaneously signal a structural problem, not a performance slump.

1. "We constantly have to tell them what to do and babysit them." The agency waits for direction rather than arriving with a standing test agenda.
2. "Why am I the one coming up with ideas for what our agency should test?" Strategy is being supplied by the buyer, not the vendor.
3. "Reporting doesn't answer pipeline questions." The monthly deck shows platform metrics while the board asks about pipeline coverage.
4. "Why do I have to keep asking what's actually working?" No proactive performance narrative exists between calls.
5. "How am I going to explain these results to the CEO?" Reporting requires manual reconstruction before it is presentable.
6. "Are we actually optimizing toward pipeline, or just leads?" The optimization target has never been confirmed against CRM data.
7. "Why are we still waiting on new creative?" Creative sits in a queue the buyer has to chase.
8. "When was the last time anyone tested our landing pages?" The post-click experience sits outside the agency's scope and inside a web team's backlog.
9. "Should we be spending more on Google? Less on LinkedIn? Testing Meta?" Channel-mix decisions are returned to the buyer by the party best placed to answer them.
10. "What are we doing this month that we weren't doing last month?" The account is maintained rather than advanced.
11. "Why am I finding problems in the account before our agency does?" The buyer is catching issues the agency is paid to catch first.
12. "We have no confidence the ad spend is returning everything it should." No CRM-connected attribution exists to validate channel economics.
13. Lead volume is up and pipeline is flat. The dashboard improves in exactly the metrics the board looks at while sales-accepted opportunities stay flat.
14. The scope is split across too many parties. Google sits with one vendor, LinkedIn with another, landing pages with a web contractor, and nobody is accountable for the result.
15. The incumbent did not own landing pages. A 2026 survey of 50+ B2B paid media operators found that 81% were not fully confident connecting paid media activity to pipeline or revenue, a gap that is structurally impossible to close when the post-click experience belongs to a different party.

Primary vs Secondary Conversion Architecture: How the Algorithm Learns

The distinction between an agency optimizing to form fills and one optimizing to CRM-defined outcomes is not a reporting preference. It determines which audience the ad platform finds tomorrow.

An account optimizing to form fills has told Smart Bidding that a form fill is the goal. The platform is not malfunctioning. It is succeeding at the goal it was given. It finds the cheapest people to convert, including students, job seekers, competitors, and existing customers. Cost per lead falls, lead volume rises, and the pipeline the sales team can actually work stays flat. Accounts that implement offline conversion tracking and value-based bidding generate 3x more pipeline at 31% lower cost per lead because Smart Bidding retrains toward clicks that produce SQLs.

The correction is to change what gets sent back to the platform. A primary and secondary conversion architecture separates the two. Secondary conversions, such as content downloads, webinar registrations, and low-commitment form completions, are tracked and visible in reporting but are never used for account-wide optimization. Primary conversions are CRM-defined outcomes, including qualified opportunities, lifecycle-stage events, and closed revenue. Importing Lead to MQL as offline conversions can reduce low-quality lead spend. MQL to SQL upload can improve SQL CPA. Opportunity upload can improve Opportunity CPA. Closed Won upload can improve ROAS.

A B2B SaaS client that previously tracked page views as conversions saw overall Google Ads conversions drop after switching to quality-lead signals, yet reported its third-highest pipeline month ever. The dashboard got worse. The business got better. That is the self-fulfilling-prophecy mechanism in reverse.

SaaSHero builds this architecture in every account during onboarding. Conversion tracking is rebuilt rather than inherited. Lifecycle-stage events are pushed back into the ad platforms from the client's CRM so the bidding algorithm learns from qualified outcomes, not page events. The mandatory discovery question that closes every SaaSHero intake is: "Are you optimizing campaigns around CRM data or just form submissions?" The answer determines whether CRM-level optimization is mechanically possible at all.

Three-Stage Demand Creation Framework for Paid Social

Most B2B paid social programs fail because they collapse a three-stage sequence into a single step. The targeting is correct. The ask is wrong. Nobody goes to LinkedIn to buy software. Asking a cold audience for a demo creates the demand-creation-versus-capture error that produces the most common complaint in agency reviews: "LinkedIn didn't work."

SaaSHero's Demand Creation Framework runs in three stages, each with defined audiences, messages, optimization goals, and explicit exclusion rules. The full arc is planned before launch.

Stage one: Awareness. The audience is cold ICP, people who fit the profile, have never encountered the company, and are not in a buying process. Messaging speaks to operational pain the person recognizes in their own week. The optimization goal is engagement, including clicks, reactions, video views, and landing page visits. The program deliberately excludes features, product walkthroughs, aggressive demo CTAs, and heavy social proof. Each assumes intent the audience does not have. Any engagement is sufficient to move a person into the consideration pool.

Stage two: Consideration. The audience is people who engaged in stage one, using retargeting pools only and never cold. Messaging introduces solutions, features, testimonials, case studies, and lead magnets. The optimization goal is traffic and content consumption, explicitly not conversions. The program deliberately excludes form-fill optimization and demo-request campaigns. Optimizing toward conversions here pulls the audience toward whoever converts fastest, a smaller and different group from the one being built. Demonstrated consumption, such as repeat engagement, content consumed, or meaningful time on site, moves a person forward.

Stage three: Conversion. The audience is warm only, fed entirely by the previous two stages. No new cold audiences are introduced. Messaging focuses on outcome and business impact, the state of the world after the problem is solved. The optimization goal is demo requests, sales-qualified lead generation, pipeline creation, and revenue outcomes. The program deliberately excludes cold audiences and awareness or consideration creative recycled into conversion campaigns. Pipeline becomes a fair measure only here, and only because the two prior stages did their work.

When one LinkedIn audience segment showed the lowest cost per lead and highest CTR, CRM data revealed a different segment converted to MQLs at nearly double the rate and produced significantly more opportunities, prompting immediate budget reallocation toward the pipeline-stronger segment. Platform metrics validate ad health. Pipeline metrics define strategy.

Flat-Fee vs Percentage-of-Spend Pricing: Incentive Map

The pricing model an agency uses shapes which recommendations the agency can make without a conflict of interest.

A percentage-of-spend arrangement, commonly 15–30% of ad spend and typically closer to 20%, ties agency revenue directly to client budget size. When the agency recommends increasing spend, its own revenue rises. When it recommends cutting a channel that is not returning, its own revenue falls. No bad faith is required for the consequence. Reallocation is the recommendation the pricing makes hardest to give. The percentage-of-spend model creates structural misalignment because an agency that improves performance enough to hit targets on lower spend cuts its own fee. Budget then calcifies where it was first placed, long after the opportunity has moved.

Per-channel pricing produces a second conflict of the same shape. If each additional channel carries its own fee, every test of a new placement raises the client's invoice. The agency has a financial interest in the channel mix staying exactly as it is. The client has a financial reason to refuse an experiment.

A flat retainer indexed to total monthly ad spend, not channel count, removes both conflicts. When SaaSHero recommends increasing a budget, it does so because the data supports scaling, not because the agency needs a raise. The same decoupling applies in reverse. When it recommends pausing a channel or reducing spend, the fee does not move, which removes any financial penalty for honesty. This structure turns channel-mix decisions into empirical questions. Shifting budget from LinkedIn to Google, opening a Meta test, or shutting a channel down entirely costs the client nothing in fees and earns SaaSHero nothing extra. For B2B SaaS companies with $5M–$50M in revenue and growth-stage budgets, flat-fee retainers are recommended over percentage-of-spend models because the workload does not scale directly with spend. The recommendation and the invoice are decoupled.

90-Day Validation Gate: Data Required by Day 90

A 90-day gate acts as the minimum window in which a properly built account can produce data clean enough to support a budget decision. B2B SaaS accounts switching to Target CPA bidding after reaching 30+ offline SQL conversions per month achieve measurable cost-per-SQL improvement within 60–90 days, given an average 84-day sales cycle. Shorter contracts do not give the work enough runway to produce the data the gate requires.

By day 90, the account must demonstrate three forms of validation: conversion infrastructure, optimization data, and strategic direction. On infrastructure, you need a documented primary and secondary conversion architecture with CRM-connected primary events, not inherited tracking from a previous configuration. This foundation enables the second requirement, which is at least one full sales cycle of offline conversion data flowing from the CRM back into the ad platforms, with lifecycle-stage events timestamped and attributable to specific campaigns. With that data flowing, the optimization layer can be validated through a search terms report reviewed weekly, a maintained negative keyword layer with documented exclusions, and at least one completed A/B test on landing page headline copy against the primary conversion event.

Strategic direction then becomes visible. A channel-level cost per SQL and cost per opportunity, calculated using CRM data rather than platform-reported conversions, shows whether the economics work. The median cost per SQL for B2B SaaS Google Ads is $800–$2,500 in 2026, varying by vertical, and must be evaluated relative to ACV to determine profitability. A Looker Studio or CRM dashboard should show platform spend alongside pipeline created, in the vocabulary the CFO uses, not as a PDF of platform metrics. A documented channel-mix recommendation with a stated rationale should rest on 90 days of clean data rather than the allocation inherited at onboarding.

For considered B2B motions, attribution windows should be set to 90–180 days rather than short ecommerce-style windows to align with longer buying cycles. A month-to-month contract evaluated at day 45 is being judged on its setup, not its results. The 90-day gate is where the work has compounded enough to evaluate outcomes rather than activity.

Book a discovery call to walk through what your current account would look like at the 90-day gate.

How SaaSHero Satisfies Every Criterion in This Framework

SaaSHero operates as the outsourced inbound growth team for B2B SaaS companies, with one team owning strategy and execution across paid media, creative, landing pages, attribution, and reporting, and aligning all of it to CRM revenue data rather than form-fill counts.

On measurement, SaaSHero rebuilds conversion tracking during onboarding, establishes a primary and secondary conversion architecture in every account, and pushes lifecycle-stage events back into the ad platforms from the client's CRM. The mandatory discovery question, "Are you optimizing campaigns around CRM data or just form submissions?", acts as the diagnostic that determines whether the engagement is mechanically possible. Reporting runs in Looker Studio and HubSpot dashboards connected to the client's CRM, showing pipeline, CAC, and payback period rather than impressions and clicks.

On post-click ownership, SaaSHero designs, builds, hosts, and A/B tests the landing pages its campaigns point to, using Figma for client approval and Unbounce for hosting and testing. Headline copy is treated as the largest lever on landing page conversion and is tested first. An agency that does not control the landing page cannot change the highest-leverage variable in the funnel.

On pricing, SaaSHero charges a flat retainer indexed to total monthly ad spend, not channel count. Adding, closing, or reweighting a channel leaves the fee unchanged, so the channel-mix recommendation and the invoice remain decoupled.

On scope, five capability areas, including paid media, creative, landing pages and CRO, attribution and reporting, and strategy, are delivered as one team under one accountability line. Creative is in-house. Concept, copy, and design are staffed capacity under the retainer, not production units purchased in blocks. Nothing is outsourced. All team members are full-time employees.

On proactivity, SaaSHero owns the strategic agenda. The client supplies goals, budget, and the approval decision. Everything between those inputs and the result, including the test queue, the creative, the channel-mix recommendation, and the competitive analysis, is staffed on SaaSHero's side. The bi-weekly strategy call is the standing forum where that agenda is delivered without being requested.

SaaSHero is a Google Premier Partner, a designation held by the top 3% of agencies, and has been a G2 High Performer in digital marketing for over two years, currently ranked #20 out of approximately 6,000 agencies. The firm has managed over $60M in lifetime ad spend across more than 100 B2B companies.

Decision Checklist That Disqualifies 80% of Agencies

Apply the following criteria in sequence. An agency that fails any one of the first five is structurally unable to deliver pipeline coverage regardless of execution quality.

1. CRM-connected optimization. Does the agency connect ad platforms to the client's CRM and optimize toward SQL and pipeline events, not form fills? If not, the bidding algorithm is trained on the wrong audience.
2. Post-click ownership. Does the agency design, build, host, and test the landing pages its campaigns point to? If not, the highest-leverage variable in the funnel belongs to a different party.
3. Primary and secondary conversion architecture. Does the agency maintain a documented separation between primary conversions used for bidding and secondary conversions tracked but excluded from optimization? If not, soft conversions are training the account toward the wrong audience.
4. Flat retainer indexed to total ad spend, not channel count. Does adding or removing a channel change the fee? If yes, the channel-mix recommendation carries an undisclosed financial interest.
5. Full-funnel attribution in the client's CRM. Does the agency produce reporting in the client's CRM, not a platform dashboard or a monthly PDF, showing pipeline created by channel, cost per SQL, and CAC payback? B2B marketers who invest in full-funnel attribution are 45% likely to significantly exceed their primary goals, compared with 24% for those who do not.
6. In-house creative. Are concept, copy, and design produced by full-time employees on the same team that runs the media? If not, the messaging cadence required by a three-stage demand creation framework cannot be maintained by rotating contractors.
7. Proactive strategic agenda. Does the agency arrive at calls with the next test, the next recommendation, and the next creative already prepared, or does the client supply the agenda? The answer determines whether the buyer has hired a vendor or acquired a direct report.
8. Client owns all assets throughout and at exit. Are ad accounts, landing page files, design files, dashboards, and conversion tracking configurations the client's property during and after the engagement? If not, switching costs function as a retention mechanism.

Frequently Asked Questions

How should B2B SaaS agencies structure performance-based pricing in 2026?

Performance-based pricing in 2026 should anchor to downstream revenue signals, including SQLs, pipeline created, and closed revenue, rather than top-of-funnel volume metrics like impressions or form fills. The two most common fee structures are percentage-of-spend and flat retainer. A percentage-of-spend arrangement ties agency revenue to client budget size, which creates a structural conflict because the agency earns more when the client spends more, regardless of efficiency. A flat retainer indexed to total monthly ad spend removes that conflict. When the retainer does not change with channel count, the agency can recommend consolidating, expanding, or pausing channels based on evidence alone. For B2B SaaS companies at the $15k+ monthly spend level, a flat retainer is the only structure in which a channel-mix recommendation is free of undisclosed financial interest. SaaSHero prices on total monthly ad spend, not channel count, so adding a new channel test or cutting an underperforming one leaves the fee unchanged.

What is the measurable impact of optimizing Google Ads toward SQLs instead of form fills?

The impact is material, not marginal. When Smart Bidding is pointed at a form fill, it finds the people most likely to fill in forms, a population that includes students, competitors, job seekers, and existing customers. When it is pointed at SQL and closed-won signals fed back via offline conversion tracking, it retrains toward the audience that actually buys. This retraining produces the 30–50% SQL volume improvement mentioned earlier, because the algorithm is no longer wasting budget on the cheapest-to-convert population. The dashboard often looks worse in the transition, because overall conversion volume drops when soft conversions leave the primary conversion set, while the pipeline the sales team can actually work improves. The self-fulfilling-prophecy mechanism runs in both directions. Feed the machine low-quality signals and it finds low-quality audiences. Feed it qualified pipeline events and it finds more of the people who produce them. SaaSHero establishes a primary and secondary conversion architecture in every account during onboarding, with lifecycle-stage events flowing from the client's CRM back into the ad platforms.

How does full-funnel ownership affect CAC payback for B2B SaaS companies?

Full-funnel ownership affects CAC payback through three compounding mechanisms. First, landing page conversion rate multiplies every other improvement in the account. A higher conversion rate changes the economics of every keyword and audience feeding it, while cutting wasted spend delivers a one-time gain. An agency that does not own the landing page cannot change the highest-leverage variable in the funnel. Second, CRM-connected attribution changes which channels receive budget. Last-click attribution assigns conversion credit to the branded search that happened after the buyer was already convinced, which defunds the channels that created demand and quietly starves the bottom of the funnel two quarters later. Multi-touch attribution, connected to the client's CRM, produces channel-level CAC payback figures that reflect how customers actually make purchase decisions. Third, in-house creative on the same team that runs the media means the messaging cadence required by a three-stage demand creation framework can be maintained and iterated continuously, rather than stalling in a contractor queue. B2B marketers who invest in full-funnel attribution are nearly twice as likely to significantly exceed their primary goals as those who do not. SaaSHero owns all five capability areas, including paid media, creative, landing pages and CRO, attribution and reporting, and strategy, under one retainer, so no seam between disciplines is owned by a different party.

Why does the 90-day gate matter, and what makes shorter contracts unworkable?

The 90-day gate matters because it aligns with both sales-cycle length and algorithm learning requirements. By day 90, a properly built account must have CRM-connected primary conversion events flowing back into the ad platforms, at least one completed A/B test on landing page headline copy, a maintained negative keyword layer with documented exclusions, and a channel-level cost per SQL calculated from CRM data rather than platform-reported conversions. The 90-day window is the minimum because the average B2B SaaS sales cycle is approximately 84 days, so a shorter evaluation window produces pipeline data from a period that does not yet reflect the full buying cycle. A month-to-month contract evaluated at day 45 is being judged on its setup. Google's Smart Bidding also requires roughly 30 conversions in a 30-day window to optimize with confidence. When real outcomes take longer, the algorithm defaults to optimizing form fills instead. Shorter contracts structurally prevent the data volume required for the bidding algorithm to retrain toward qualified outcomes, which means the account is evaluated before the optimization method has had time to work. SaaSHero's engagement model uses a validation period followed by a longer committed term, with the 90-day gate as the point at which the channel, the structure, and the messaging thesis can be evaluated on outcomes rather than activity.

How can a VP of Marketing tell whether an agency is proactive or reactive?

The clearest test is who writes the brief. In a reactive relationship, the marketing leader generates the test ideas, assigns the work, chases the status, and finds the problems in the account before the agency does. The agency executes a brief written elsewhere. In a proactive relationship, the agency arrives at every call with the next test already designed, the next creative already in review, and the next channel-mix recommendation already argued from the data. The buyer's role is to supply goals, budget, and the approval decision, not to direct the work.

Several practical signals make this visible. The agency's bi-weekly call agenda should arrive from the agency, not from the buyer. The monthly competitor analysis should land on its own schedule, not only when someone asks. The quarterly budget analysis should recommend reallocation, not reproduce last quarter's split. The agency should find problems in the account before the buyer does, not after. SaaSHero's operating model is built around a fixed cadence, including bi-weekly strategy calls, weekly performance updates, monthly competitor analysis, and quarterly budget analysis, all produced without being requested. The Senior Account Strategist owns the strategic agenda. The client does not supply it.

Conclusion: Decide Whether to Write the RFP or Schedule the Call

The four structural shifts, including platform automation moving the work to data quality, measurement degradation moving the answer into the CRM, mid-market teams holding judgment but not operators, and standard retainers stopping short of the chain they are judged on, converge on one requirement. A single agency must own the full path from impression to CRM record, price on total ad spend rather than channel count, and optimize toward SQLs and pipeline rather than form fills.

The 15-point checklist disqualifies most agencies on measurement and scope alone. The 90-day gate defines what clean data looks like and explains why shorter contracts prevent it. The pricing incentive map shows which fee structure allows a channel-mix recommendation to be made without a conflict of interest. The demand creation framework explains why many LinkedIn programs fail and what a three-stage sequence actually requires.

SaaSHero satisfies every criterion in this framework. One team covers five capability areas under one flat retainer indexed to total monthly ad spend. Every account runs on CRM-connected optimization. Post-click ownership is treated as a condition of accountability. The strategic agenda is proactive, and the client does not have to supply it.

If your current agency fails more than three items on the decision checklist, the relationship is structurally unable to deliver pipeline coverage, not because of execution quality, but because of scope and incentive design.

Book a discovery call and bring your current account. SaaSHero will tell you what it sees, including whether the relationship is worth saving.