Summary: Most agencies fail when asked one specific question: “What is your honest win rate?” The 11 other questions in this list are designed to expose the gap between what an agency claims and what it can deliver. Use them as a structured filter before you sign anything. The good answers will help you. The bad answers will save you a year and several crore.
Founders ask the wrong questions when evaluating growth agencies. They ask about packages, pricing, and “what would you do for us in the first 90 days.” Those questions get rehearsed answers. Every agency has the same elevator pitch about “data-driven strategies” and “compound growth.” Walking out of those conversations, you learn nothing about whether the agency can actually do the work.
The questions that matter are the ones agencies cannot rehearse. They expose internal process, hiring quality, attribution discipline, and willingness to disqualify themselves. After 13 years running engagements at upGrowth Digital, I can tell you which questions break good agencies away from bad ones inside 30 minutes. Twelve of them. Use this list before you sign anything.
Run them in this order. The early questions filter for capability. The middle ones filter for honesty. The last few filter for fit. An agency that handles all twelve well is rare. An agency that fumbles three or more should be off your shortlist.
Questions one through four: capability
Question 1. Show me three case studies in my vertical with the specific numbers and the specific timeframe.
Good answer: a senior team member walks you through three engagements in your vertical. They name the specific metric (organic clicks, CPL, ROAS, AI citation share), the specific number, the timeframe, and what they did differently from the previous agency. The case studies are within the last 24 months because patterns from 2022 do not transfer to 2026. They acknowledge what went sideways in each engagement, not just the wins.
Bad answer: case studies that are “anonymous” because of NDA, or that quote “300% growth” without naming the metric or the baseline, or that describe what the agency did without telling you what the client actually got. If they cannot show specific numbers in your vertical, they have not done the work in your vertical. The Lendingkart engagement at upGrowth ran 5.7x lead volume, 30% CPL reduction, and 4x spend scaling on Google Ads. That is the level of specificity to expect. If an agency cannot match that disclosure, they are either new to your vertical or hiding the numbers.
Question 2. Who exactly would work on my account, and what is their tenure with you?
Good answer: they name the strategist, the specialist (SEO, paid, content, GEO, depending on the engagement), and the account manager. Each named operator has been with the agency for at least 18 months. The strategist has worked with at least three companies in your stage range. The agency commits in writing that these people will not be swapped out without your approval.
Bad answer: vague references to “our team,” senior people on the pitch call who you never see again after signing, junior operators on the actual work, no commitment to staffing continuity. The pitch-and-switch pattern is the most common reason engagements fail. The work the senior team promised is executed by a junior team that does not understand the strategy. Get the names in writing before you sign.
Question 3. Walk me through your discovery and diagnostic process.
Good answer: they describe a structured diagnostic that runs across attribution, conversion, retention, paid efficiency, organic, AI visibility, and content. They explain how they will sequence the bottleneck fixes. They show you the framework they apply, and they can name it (the Organic Compounding System, the Paid-to-Organic Transition Model, the GEO Visibility Framework, etc). They commit to a written diagnostic deliverable in the first 30 days that you own regardless of whether the engagement continues.
Bad answer: “we’ll start with a discovery call and propose a strategy.” That is not a process. That is a placeholder. Agencies without a real diagnostic process default to their own playbook regardless of what your business actually needs. They fix what they know how to fix, not what is broken in your situation.
Question 4. What is your point of view on the SEO versus GEO question?
Good answer: they explain that the two disciplines run in parallel rather than as substitutes. They walk you through what changes in content engineering for AI extractors (FAQ schema, entity definitions, original data, structured sections). They name specific examples of brands that have successfully integrated both disciplines. They have an opinion about which signals matter most in 2026 and can defend it.
Bad answer: “GEO is just SEO with extra steps” or “we mostly do SEO and GEO is a future priority” or anything that suggests they have not thought carefully about how AI search has changed the discipline. ChatGPT has 883 million monthly active users in 2026. Google AI Overviews appear in 18% of all searches. Pew Research found pages featured in AI Overviews see a 46.7% drop in click-through rates. An agency that has not integrated GEO thinking by now is two years behind.
Question 5. What is your honest win rate? How many engagements work and how many do not?
Good answer: a specific number, ideally between 60% and 80% of engagements meeting their primary KPI. They explain what defines a win for them and what defines a loss. They tell you what kills engagements (client side: leadership change, budget cut, lack of internal owner; agency side: scope creep, wrong fit at signing). The number feels uncomfortable to share but they share it.
Bad answer: “almost all our engagements succeed” or “we have a 95% retention rate” or anything north of 90%. No agency wins 95% of engagements. The few that exist do so by quietly dropping clients before failure, redefining success, or operating at such low scope that the bar is meaningless. The most credible answer is a real number with real losses attached. Agencies that admit they fail occasionally are usually the ones that succeed more often.
Question 6. When have you walked away from a potential client and why?
Good answer: a specific story, maybe two. The client was the wrong stage, the wrong vertical, had unrealistic expectations, or had broken attribution that needed to be fixed before any agency work made sense. The agency declined the engagement and explained why. They lost the revenue. They were right.
Bad answer: “we work with anyone who is the right fit” or “we have never turned down a client.” Agencies that never disqualify themselves take on engagements they should not. Six months later both sides are unhappy. The willingness to lose a deal at signing is one of the strongest predictors of engagement quality. We use this signal at upGrowth and it has saved both us and the client multiple times.
Question 7. What does your reporting cadence look like and what metrics do you actually track?
Good answer: weekly reporting with specific metrics tied to the diagnosis they did at kickoff. The metrics include leading indicators (rankings, traffic, CTR, AI citation share, CPL, ROAS) and lagging indicators (qualified leads, pipeline contribution, revenue attribution). They can show you a sample report from another engagement (with names redacted) so you can see the format. They commit to a monthly strategy review with you, not just a metrics dump.
Bad answer: monthly reports with vanity metrics. “We grew your traffic 40%” without explaining whether that traffic converted. Reports built around the agency’s deliverables (X blogs published, Y backlinks built) rather than the client’s outcomes. Agencies that report on activity rather than outcome are agencies that cannot prove their work moved the needle.
Question 8. What is your attribution methodology and how do you handle the cross-channel question?
Good answer: they explain how they distinguish organic-assisted paid conversions, brand search lift from content marketing, AI citation referrals, and direct attribution. They acknowledge the limits of last-click models. They have a position on multi-touch attribution that is more sophisticated than “we use Google Analytics.” They are willing to integrate with your CRM to track downstream pipeline.
Bad answer: “we report on the GA4 default model.” That is not attribution thinking, that is dashboard interpretation. The 2026 reality is that 18% of searches return AI Overviews and 60.7% of AI search market share sits with ChatGPT. Last-click attribution misses most of the value chain. Agencies that have not built a sophisticated attribution view are reporting on the wrong things.
Question 9. How do you make decisions when our priorities conflict?
Good answer: a clear escalation path. The strategist owns the day-to-day. The senior leadership at the agency is available for cross-functional conflicts. They commit to a quarterly business review with you and your CEO/CMO. They have a process for surfacing disagreements rather than executing on conflicting instructions. They tell you they will push back when they think you are wrong, and they describe a recent example.
Bad answer: “we are very flexible and accommodate client preferences.” Translation: they will execute whatever you ask and not push back. That is not a partner. That is a vendor. Vendors are useful when you know exactly what you want. They are dangerous when your strategic clarity is incomplete because they will execute against your blind spots without challenge.
Question 10. What does the off-boarding process look like if we end the engagement?
Good answer: a structured 30-to-60-day handover. They commit to documenting the work, transferring asset ownership (content, dashboards, links, accounts), and providing a final strategic recommendation. The off-boarding includes the diagnostic, frameworks, and execution playbooks they built during the engagement. You leave with assets you can hand to the next agency or in-house team without losing six months of context.
Bad answer: anything that suggests off-boarding is an afterthought, or that key information lives in the agency’s tools without your access, or that asset transfer would require additional fees. Agencies that build hostage situations into their engagements are agencies that fear competition on results. The good agencies make the off-boarding clean because their reputation lives on what clients say after the engagement ends.
Question 11. What does success look like at the end of 12 months in our specific situation?
Good answer: a specific commitment tied to your bottleneck. If your bottleneck is organic compounding, they commit to a target organic traffic growth, featured snippet count, and topical authority expansion. If your bottleneck is AI visibility, they commit to citation share growth in target queries. If your bottleneck is paid efficiency, they commit to CPL reduction and ROAS targets. The numbers feel specific and conservative rather than aspirational.
Bad answer: vague language about “compound growth” or “marketing leadership transformation” without a specific metric or number. Or wildly aspirational targets that the agency will not be held accountable for. The right number is uncomfortable to commit to. If the agency commits without flinching, they are either confident or careless. The follow-up question to ask is “what conditions would make you fail to hit that number” and watch how they answer.
Question 12. Why should we not hire you?
This is the question that breaks most agencies. It is also the most useful question on the list.
Good answer: a specific honest answer. “If your bottleneck is product-market fit, we cannot fix that.” “If you do not have a senior internal owner for this engagement, we will fail.” “If you need execution at sub-six-month timelines and you are still pre-attribution, an SEO engagement is not the right fit; you need a fractional CMO first.” The agency is willing to articulate the boundary conditions where they would not succeed.
Bad answer: silence, awkward laughter, or “we will figure out a way to make it work for any client.” That is not confidence. That is a sales script. The agencies that can name when they are not the right fit are the ones that succeed when they are. The willingness to lose a deal at signing is the same muscle that makes the engagement work after signing.
I should run our own answers through this list because the questions are useless if the people who wrote them cannot pass them.
Capability questions: case studies are public at upgrowth.in/case-study/, with named numbers, named clients, named timeframes. Fi.Money grew 200,000+ monthly clicks in 9 months. Lendingkart hit 5.7x lead volume with 30% lower CPL. Scripbox crossed 198,000 traffic in 2 months. Vance hit 70% organic traffic from target geos in 3 months. Delicut Dubai went from 40,000 AED to over 2 million AED monthly sales. The named operators on each engagement are senior and have been with us for years. The diagnostic process runs through the seven-bottleneck framework before any execution work. The SEO versus GEO position is integrated rather than parallel because that is what the 2026 buyer requires.
Honesty questions: our engagement win rate sits in the 60-to-75% range depending on how you define success. We have walked away from prospects whose attribution was so broken that any agency engagement would have failed. Our reporting is weekly with metrics tied to the kickoff diagnostic. Our attribution methodology accounts for AI citation referrals, organic-assisted paid, and pipeline contribution rather than just last-click.
Fit questions: our escalation path runs from strategist to senior leadership to founder. We push back when we think clients are wrong, and we have ended engagements when the disagreement was structural. Our off-boarding includes documentation, asset transfer, and a strategic recommendation that survives the engagement. Our 12-month commitments are specific to the diagnosed bottleneck and conservative enough to be defensible. We tell prospects when we are not the right fit, and Grove (our diagnostic at upgrowth.in/grove) does this automatically when prospects are pre-revenue, sub-50K marketing budget, or looking for execution at freelancer pricing.
The honest answer to “why should we not hire upGrowth” is also worth stating. If you are pre-revenue or pre-product-market-fit, we are the wrong partner. Our work is built for funded, post-PMF companies that need to scale a working motion. If you have a strong existing in-house team and need pure execution, a specialist on retainer will cost less than our agency engagement and may produce comparable results. If your category does not yet have AI search demand, GEO work is premature and you can wait six months. We say all of this in the first conversation rather than after the contract is signed.
Six Common Questions About Vetting Growth Agencies
Q: What is the most important question to ask a growth agency before hiring them?
A: “Why should we not hire you?” Most agencies cannot answer this. The ones that can name the conditions where they would not succeed are the ones that actually succeed when conditions match. The willingness to disqualify themselves is the same muscle that makes the engagement work. We treat this as the single most important filter at upGrowth, and Grove (our diagnostic at upgrowth.in/grove) runs this filter on prospects automatically by walking them through stage, team setup, and bottleneck before recommending an engagement.
Q: How do I know if a growth agency case study is credible?
A: Three signals. The case study names the specific client (not “a Series B fintech”). The case study quotes specific metrics with specific timeframes (200,000 clicks in 9 months, 5.7x lead volume, 30% CPL reduction). The case study acknowledges what was hard, what almost did not work, or what the team did differently from a previous agency. Anonymous case studies, vague growth percentages, and stories that read like marketing copy are warning signs. Real case studies sound like operators talking, not marketers writing.
Q: What is a normal win rate for a growth agency?
A: 60% to 80% of engagements meeting their primary KPI is a credible range. Agencies claiming above 90% are usually quietly dropping failures, redefining success, or operating at low enough scope that any outcome counts as a win. Agencies below 50% have a process problem, a hiring problem, or a fit problem. The right agencies are honest about the failures and explain what kills engagements (leadership change, budget cut, lack of internal owner on the client side).
Q: Should I hire an agency that does both SEO and GEO, or specialists in each?
A: Integrated, in most cases. The two disciplines share the same content engineering surface and increasingly the same content engineers. Splitting them across two agencies creates coordination overhead and conflicting recommendations on what to publish, how to structure schema, and which queries to target. The integrated path produces better results faster. The exception is at very high spend levels (multiple crore per year on each discipline) where two specialist agencies with a fractional CMO coordinating between them can outperform a single integrated agency. For most companies, integrated is the right call.
Q: What red flags should I watch for in a growth agency pitch?
A: Five common ones. Senior people on the pitch who disappear after signing. Case studies that quote percentages without naming the metric or baseline. Reports built around agency deliverables rather than client outcomes. Inability to articulate when they would not be the right fit. Aspirational 12-month commitments without conservative anchor numbers. Any one of these is a yellow flag. Two together is a red flag. Three is a no.
Q: How long should the contract be with a growth agency?
A: 6 to 12 months for most engagements. Below 6 months, the agency cannot do compounding work; they default to short-term tactical execution. Above 12 months, you lose the option value of testing whether the engagement is working. The right structure is a 12-month engagement with quarterly business reviews and an explicit off-ramp at the 6-month mark if the agreed-upon KPIs are not tracking. The off-boarding terms should be in the contract from day one. Agencies that resist these terms are signaling they expect to perform poorly enough that you would want to leave early.
Your Next Move: Run the Twelve Questions Against Your Shortlist
Pull your current shortlist of growth agencies. Run all twelve questions on each one in the next pitch call. Score the answers as good (clean and specific), partial (acceptable but lacking specificity), or weak (rehearsed or evasive). The agencies with mostly good answers stay on the list. The agencies with three or more weak answers come off it. The exercise takes 30 minutes per agency and saves 6 to 12 months of misaligned engagement.
If you want to run the diagnostic conversationally before the agency pitches start, Grove at upgrowth.in/grove walks you through the framework in 5 to 7 minutes. The output tells you which agency structure your situation actually needs (full SEO, GEO-heavy, fractional plus specialist, or in-house). With that diagnosis in hand, the twelve questions become much sharper because you know what you are filtering for.
About the Author: I’m Amol Ghemud, Chief Growth Officer at upGrowth Digital. We help SaaS, fintech, and D2C companies shift from traditional SEO to Generative Engine Optimization. This shift has generated 5.7x lead volume increases for clients like Lendingkart and 287% revenue growth for Vance.
For Curious Minds
The "pitch-and-switch" is a deceptive practice where an agency uses its experienced senior team to win your business, only to assign junior, less capable operators to handle the actual day-to-day work post-contract. This is a primary cause of failed engagements because the strategy promised is not the strategy executed. You can counter this by asking for the names and tenure of the specific people who will manage your account. A stable team is a sign of a healthy agency. An agency that can commit to a strategist and specialists who have been with them for at least 18 months demonstrates internal stability and process maturity. Insist on getting this commitment in writing. This simple step shifts the power dynamic and ensures the team that impressed you in the pitch is the team dedicated to your growth. Discover how to codify this commitment in your contract by reading the full guide.
A structured diagnostic process reveals an agency's depth of thinking and proves they have a repeatable, methodical approach to growth, rather than relying on generic tactics. When an agency can name and explain their framework, it shows they have refined their strategy over many engagements. For instance, a process that evaluates attribution, conversion, retention, and paid efficiency in a specific sequence demonstrates a sophisticated understanding of business bottlenecks. This is the difference between hiring a strategist versus a simple executor. A good answer includes a commitment to a written diagnostic deliverable in the first 30 days that you own, showing confidence in their analysis. This approach uncovers their problem-solving ability, which is a far better indicator of success than any polished slide deck. See the other key process questions that separate top-tier agencies from the rest.
Specific, verifiable metrics are always more valuable than vague, high-level growth percentages. A claim like "300% growth" is meaningless without context: what was the baseline, what was the metric, and over what timeframe? In contrast, a detailed result, such as the one upGrowth Digital achieved for Lendingkart with a 30% CPL reduction alongside a 4x spend scaling, demonstrates a clear understanding of business impact. Credibility lies in the details. When weighing evidence, prioritize agencies that can provide:
The specific metric improved (e.g., CPL, ROAS, organic clicks).
The exact numerical change.
The timeframe of the engagement.
The specific actions that led to the result.
An agency that provides this level of detail is confident in its work and respects your need for concrete proof. Learn which other "bad answers" are immediate red flags during your evaluation process.
The Lendingkart example sets a high bar for what constitutes a compelling case study. It is effective because it connects agency actions directly to tangible business outcomes, not just marketing metrics. When you ask for a case study, look for the same level of granular detail and strategic depth. Your goal is to see evidence of their problem-solving skills, not just a highlight reel. Specifically, a strong case study should include:
The Starting Point: What was the initial challenge or the previous agency's failure?
The Core Metric: What was the primary business goal, like the 5.7x lead volume increase?
The Timeframe: How long did it take to achieve these results?
The "How": What specific, differentiated strategy did they apply?
The Setbacks: What went wrong and how did they adapt?
An agency that can walk you through these points for a recent client in your vertical is showing, not just telling. Explore the full list of questions designed to uncover this level of proof.
The digital marketing landscape changes dramatically every 18-24 months due to algorithm updates, new platforms, and evolving consumer behavior, especially with the rise of AI. An agency that relies on case studies from before 2022 is signaling that its strategies may be outdated. Past performance is only relevant if the playbook is current. An agency's ability to show recent success proves it can adapt and thrive in the current market, not the market of three or four years ago. A strong agency will proudly present recent work, demonstrating their command of modern challenges like building AI visibility. A refusal or inability to do so suggests they have not produced significant results lately, which is a major red flag about their current capabilities. Read on to find out how to test an agency's adaptability for future challenges.
While NDAs are a legitimate part of business, they should not be a complete barrier to demonstrating competence. An experienced agency knows how to showcase its wins while respecting client confidentiality. The key is to push for aggregated or anonymized data that still proves the point. A credible agency can navigate this by:
Sharing the specific metric improved (e.g., ROAS, CPL) and the percentage change without naming the client.
Providing a detailed walkthrough of the strategic process they used to achieve the result.
Offering to connect you with a current or past client for a reference call.
Showing specific numbers for clients who have given them permission, like how upGrowth Digital shares the Lendingkart data.
An agency that flatly refuses to provide any specific numbers, even in an aggregated form, is likely hiding a lack of results, not just protecting a client. Learn what other excuses signal a weak agency in the complete guide.
To maximize your evaluation efficiency, structure your initial calls around three core themes in a specific order: capability, honesty, and fit. This sequential filtering method quickly exposes weaknesses. Start with the questions that are hardest to fake and require hard evidence. A good structure for a 30-minute call would be:
Capability (15 mins): Begin by asking for three specific case studies in your vertical (Question 1) and details on their diagnostic process (Question 3). This immediately tests for relevant experience and strategic depth.
Honesty (10 mins): Ask who exactly will work on your account and their tenure (Question 2). This uncovers potential "pitch-and-switch" issues.
Fit (5 mins): Conclude with a question about their ideal client profile to see if there is mutual alignment.
An agency that struggles with the capability questions in the first 15 minutes can be disqualified, saving you from wasting more time. The full article provides the complete twelve-question sequence for a comprehensive evaluation.
You can prevent the "pitch-and-switch" by making personnel a formal part of your agreement. This moves the conversation from a verbal promise to a contractual obligation, which serious agencies will have no problem accepting. Your goal is to codify accountability directly into the statement of work. Work with your legal counsel to add a "Key Personnel" clause that includes the following points:
Names the specific individuals (e.g., Strategist, Account Manager) assigned to your account.
States that these individuals will not be reassigned without your prior written consent.
Specifies a transition period or specific qualification requirements for any potential replacements.
Outlines a remedy or penalty if the clause is violated.
An agency's reaction to this request is as telling as their answer. A confident partner will agree, while a hesitant one may be planning a switch. Find more ways to build accountability into your agency relationships.
Asking about AI visibility tests whether an agency is reactive or proactive in its strategic thinking. Traditional SEO is no longer sufficient as AI-powered search engines increasingly provide direct answers, bypassing website clicks. An agency that cannot articulate a clear strategy for this new reality is still operating on an old playbook. A forward-thinking agency is already developing methods to get their clients cited in AI responses. A strong answer will touch on:
Creating high-authority, data-rich content.
Structured data and schema markup.
Building a strong knowledge graph presence.
Measuring success through metrics like AI citation share.
An agency's fluency on this topic shows they are prepared to secure your market position not just for today, but for the next three to five years. The complete article explores other forward-looking questions you should be asking.
Partnering with an agency that lacks a forward-looking diagnostic process exposes your business to strategic stagnation and wasted investment. The primary risk is that they will apply outdated tactics to new market problems, leading to diminishing returns and eventual failure. Without a dynamic diagnostic framework, an agency is just guessing. The long-term risks include:
Wasted Ad Spend: Pouring money into channels with declining efficiency because they are not adapting to platform changes.
Loss of Market Share: Competitors who partner with more sophisticated agencies will capture new opportunities first.
Inability to Scale: Their one-size-fits-all approach will fail to identify and resolve your unique growth bottlenecks as you mature.
A clear diagnostic process is a sign of an agency built for the future. Learn how to spot these future-proof qualities in the full analysis.
The primary flaw in focusing on pricing and packages is that it treats agency services like a commodity, inviting rehearsed sales pitches rather than revealing true capability. Any agency can create an attractive-looking package, but very few can deliver on it. This approach evaluates the wrapper, not the contents. Shifting the conversation to process and people forces an agency to demonstrate how they think and operate. By asking about their diagnostic framework, team tenure, and specific case studies, you are testing for:
Strategic Depth: Do they have a repeatable system for solving problems?
Talent Quality: Do they retain experienced operators?
Proven Results: Can they back up claims with hard numbers, like the 5.7x lead volume achieved for Lendingkart?
This method allows you to evaluate their actual ability to deliver results, making the final price discussion much more meaningful. The full guide offers a complete script to steer these conversations effectively.
Using the diagnostic framework question is an excellent way to filter for an agency that prioritizes substance over superficial reporting. A top-tier agency's response will immediately shift the conversation from vanity metrics to business bottlenecks. They will focus on the 'why' behind the numbers, not just the numbers themselves. You can identify a superior partner if their diagnostic process aims to uncover core issues across the full funnel, including attribution, conversion, and retention. They should explain how they sequence fixes to address the most critical constraints first. This demonstrates a commitment to driving real business value, like the 30% CPL reduction seen in the Lendingkart example, rather than simply generating charts for a monthly report. Learn the other questions that expose an agency's true focus.
Growth Strategy and Planning
Inbound Growth
Growth Hacking
Fractional CMO
Search Engine Optimization
Paid and Performance Marketing
Social Media Marketing
Demand Generation
AI-Driven Growth Strategy
AI-Native Workflow Automation
Generative Engine Optimization
Grove — AI Growth Strategist
AI Automation
Framework
Case Study
Tools
Growth Tools
Decision Tree
Compare
Custom GPT's
Guides
Offer
Webstories
Quizzes
Blog
AI-powered Apps
Social Media Tools
Agency Fit Score
