Most SEO agency evaluations in 2026 score on the wrong axis. Traditional SEO metrics like backlink counts and keyword rankings now sit downstream of AI search behaviour, which means an agency that only tracks those signals will miss where your traffic is actually going. The seven due-diligence questions below separate agencies still optimising for 2019 Google from agencies built for AI Overviews, ChatGPT citations, and intent-driven search.
In This Article
Share On:
An SEO agency pitched us last month. Strong deck. Nice case studies. Rankings charts that climbed from page three to page one. The founder walked out of the pitch feeling good about it.
Then we asked the agency a simple question: show us the percentage of client traffic coming from AI Overviews, ChatGPT, and Perplexity. Silence. Then a slide about “we’re planning a GEO module next quarter.”
That’s the problem. A founder evaluating SEO agencies in 2026 is often evaluating them on criteria that stopped mattering in 2024. Backlink profiles still matter, but they matter less than whether your agency can get you cited in an AI Overview for a query that sends a buyer to your pricing page. Ranking for a keyword still matters, but it matters less than whether that keyword still has click-through traffic now that AI answers it directly in the SERP.
At upGrowth Digital, we’ve audited 40+ SEO agencies over the last 18 months, both as prospective partners and as an audit deliverable for clients questioning their current vendor. The pattern is consistent. Agencies that still frame the work as “rank for keywords, build links, publish blog posts” are measured on outputs. Agencies framing it as “win search queries wherever the buyer’s journey takes them” are measured on outcomes. The gap between the two approaches widens every quarter.
Here are the seven questions that expose whether an agency is built for how search actually works in 2026.
Question 1: What percentage of your clients’ organic traffic comes from AI surfaces?
Any competent SEO agency should be able to answer this in 30 seconds. If they fumble, ask for a follow-up report within 48 hours. If the report never comes, walk away.
The question matters because Google AI Overviews, ChatGPT, Perplexity, and Gemini now intercept a meaningful share of queries before the classic blue-link SERP ever renders. Depending on your vertical, that share is somewhere between 8 and 34 percent of search impressions. An agency that isn’t tracking it is flying blind.
Good answer pattern: “Across our B2B SaaS clients, AI surface traffic is 14 percent of organic sessions on average, up from 4 percent 12 months ago. We track it through UTM filtering on ChatGPT and Perplexity, referrer analysis for AI Overviews, and brand mention tracking across AI responses.”
Bad answer pattern: “We focus on rankings. AI is experimental and we’ll look at it when it matures.”
The second pattern means the agency has already missed the transition.
Question 2: How do you build topical authority, and what does a content cluster look like for a new client?
Topical authority is the mechanism by which Google and AI systems trust a site enough to cite it. Backlinks still matter, but they matter as one signal among many. What matters more in 2026 is whether you’ve built a coherent cluster of content that covers the entire buyer journey for a specific intent area.
The agency should walk you through their cluster-building process with specificity. Questions to probe:
How do they identify the cluster’s pillar topic? What does their keyword research process output? Do they build FAQ schema into every piece? How do they handle internal linking within a cluster versus across clusters? What’s the publishing cadence for a new cluster to reach authority?
If the agency describes a process that could apply to any content marketing agency in 2018, the answer is inadequate. A 2026 answer includes AI query mapping, entity associations, schema architecture choices, and a plan for mesh reinforcement across the content assets.
Question 3: Show me a case study where you replaced a competitor agency mid-contract. What broke, and what did you rebuild?
This question is a stress test. Most agencies can present a case study where they started with a greenfield client and drove results over 12 months. Fewer agencies can show you the grittier work of taking over a stalled account from a previous vendor and turning it around.
The quality of the answer reveals three things. One, whether they have enough client volume to have inherited accounts in trouble. Two, whether they can diagnose root cause in an existing SEO stack. Three, whether their turnaround plan follows a repeatable framework or is improvised.
What to listen for: specifics on audit methodology (technical crawl, content gap analysis, backlink toxicity review, internal linking map, schema validation), a clear first-30-day priority list, and honest reporting on what couldn’t be fixed (penalty recoveries, domain authority ceilings, redirect chain cleanup that required dev sprints).
If the agency only shows you six-month-ahead results without walking you through the diagnostic work, they may not have a real methodology.
Question 4: For YMYL verticals, how do you enforce E-E-A-T and factual accuracy?
If you’re in fintech, healthcare, legal, insurance, or any vertical where content accuracy impacts user wellbeing or financial decisions, this question is non-negotiable. Your Money Your Life (YMYL) content gets held to a higher editorial standard by Google, and AI systems refuse to cite sources that feel loose on credentials.
The agency should describe a process that includes: named author bylines with verifiable credentials, medical/legal/financial reviewer sign-offs where applicable, citation of primary sources (regulatory filings, peer-reviewed research, government data), date stamps on claims that age (compliance rules, interest rates, drug dosages), and an editorial calendar that triggers review of old content when regulations change.
Red flag response: “We use AI to generate drafts and have our writers edit them.” In YMYL, that’s a compliance risk waiting to surface. For healthcare clients we’ve flagged exactly this workflow in three agency reviews, and in two cases the client had to take down 40+ pieces of live content because it cited outdated dosage guidelines.
Question 5: Walk me through your monthly report. What KPIs are committed, and what KPIs are directional?
This question exposes whether the agency has thought through what they can guarantee versus what they can influence. Real SEO agencies separate the two cleanly. Sketchy SEO agencies promise everything and measure nothing.
Committed KPIs (what they’ll put in the contract): activity-level outputs. Number of pages published, technical fixes shipped, internal links built, schema implementations, audit deliverables.
Directional KPIs (what they’ll influence but not guarantee): ranking improvements, organic traffic growth, citation share in AI responses, conversion rate changes, revenue attribution.
An agency that guarantees ranking positions is lying or doesn’t understand how search works. An agency that only commits to “we’ll do our best” has no accountability. A good agency commits to the execution layer with precision, reports on the outcome layer with honesty, and has a clear protocol for when outcomes lag expectations.
Questions to probe inside their reporting: how often do they send reports, who on their side presents them, what’s their protocol when a KPI slips two months in a row, and can they show you an actual past report (not a template) with real client numbers redacted.
Question 6: How do you optimise content for citation in ChatGPT, Perplexity, and AI Overviews?
GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) are the next layer of SEO, not separate disciplines. An agency that treats them as a separate line item or a future roadmap item is already behind.
The right answer describes a specific process. Here’s what that sounds like in practice: query mapping to identify what prompts the buyer actually types into AI tools, structural optimisation of content (clear H2/H3 hierarchy, answer-ready formatting, extractable sentences under 30 words), schema markup that feeds structured data to AI indexers, entity association work to build semantic connections, and active citation monitoring through tools like Profound, Peec.ai, or similar.
If the agency doesn’t mention specific tools or describes GEO as “we’re thinking about it,” they’re not yet operational in this layer. We track citation share for clients as a primary KPI now. For one fintech client, 41 percent of their AI-referred traffic in Q4 2025 came from ChatGPT responses to queries where they’d built citation infrastructure six months earlier. That’s a 2.3x lift over pure organic search traffic from the same query set.
Question 7: Who will actually work on my account, and what’s the escalation path?
The pitch team is rarely the delivery team. This is the most predictable gap between what you’re sold and what you get. Ask for named team members, their roles, their availability per week on your account, and the escalation path when something breaks.
Good answer structure: account lead (senior, 8-12 hours/week on your account, runs strategy and monthly reviews), content lead (owns the editorial calendar and quality review), technical SEO (audits, schema, crawl work, 2-4 hours/week baseline plus sprint time), link acquisition (only if this is in scope), and analytics (dashboard ownership and reporting).
Red flags: generic titles without names, reluctance to share LinkedIn profiles of team members, account leads with 30+ other active clients, or an “offshore team” presented as an execution engine with no named manager.
One agency we audited last year had their senior strategist on four pitch decks simultaneously. When the client started, the senior strategist disappeared and a junior account manager with six-month tenure took over. The client was paying Rs 3.5L/month for junior execution dressed as senior strategy. This gap is common enough that insisting on named team members is non-negotiable.
How We Audit SEO Agencies for Clients
When a client brings us in to audit their existing SEO agency, we run a 14-day protocol. Day 1-3 is document review (contract, SOW, past 12 months of reports). Day 4-7 is technical audit of the client’s actual site and content assets to compare against what the agency reported shipping. Day 8-11 is interviews with the agency’s team (with client’s written approval) to probe process maturity. Day 12-14 is synthesis and recommendation.
The recommendation almost always falls into one of four buckets. Bucket one: keep the agency and restructure the scope. Bucket two: keep the agency but renegotiate the contract and reporting. Bucket three: run a 90-day bake-off between current agency and a challenger. Bucket four: replace the agency.
Across the last 18 months, the split has been roughly 35 percent bucket one, 22 percent bucket two, 15 percent bucket three, 28 percent bucket four. Meaning about 28 percent of SEO agencies we audit are so misaligned with the client’s actual business that replacement is the right call.
Seven Common Questions About Evaluating SEO Agencies in 2026
Q: How much should a quality SEO agency cost in India for a mid-market B2B SaaS?
A: Retainers range from Rs 1.5L to Rs 6L per month depending on content volume, technical scope, and vertical specialisation. Under Rs 1.5L, you’re likely getting junior execution with limited strategic input. Above Rs 6L, you’re paying for specialist verticals (healthcare, fintech, legal) or named senior leadership on your account.
Q: Can I ask an SEO agency to guarantee rankings?
A: No. Any agency that guarantees ranking positions for specific keywords either doesn’t understand how Google works or is planning to manipulate metrics in ways that will hurt you long-term. Guarantees should be on activity (pages published, audits delivered, technical fixes shipped) and targets should be on outcomes (traffic, citations, conversions) with clear protocols for when outcomes lag.
Q: How long before SEO work shows traffic results?
A: For a greenfield site in a competitive vertical, expect 6-9 months before meaningful organic growth. For an existing site with stalled SEO, a good agency should show improvement within 90 days because the quick wins (technical fixes, schema, internal linking, content refreshes on existing assets) are already sitting on the table.
Q: Should my SEO agency also handle my content writing?
A: It depends on your in-house capacity. If you have a competent content team, the agency should provide briefs, SEO guidance, and editorial review. If you don’t, they should provide end-to-end content production with named writers and editorial oversight. The trap to avoid is an agency that ghostwrites content with no editorial accountability.
Q: How do I know if my current SEO agency is actually working?
A: Look at three signals. First, month-over-month organic traffic trend for your core commercial pages (not branded search). Second, growth in keyword rankings for your target cluster, tracked in Ahrefs or SEMrush. Third, citation share in AI responses for queries your buyers actually type. If all three are flat or declining for six consecutive months, either the work isn’t happening or the strategy is wrong.
Q: What contract terms should I push back on?
A: Refuse auto-renewal without a 60-day opt-out window. Refuse exit fees that penalise termination. Refuse clauses that make the agency the owner of the content they produce for you (you paid for it, you own it). Refuse reporting language that doesn’t specify cadence, format, and response time for your questions.
Q: Is in-house SEO ever better than agency SEO?
A: For companies with annual organic revenue above Rs 30 crore and an existing content operation, yes. Below that, the specialist knowledge an agency brings (across 20+ client verticals, with live feedback loops on what’s working) is hard to match with a 1-2 person in-house team. The hybrid model (in-house manager plus agency execution) tends to beat both pure in-house and pure agency at mid-market scale.
Your Next Move: Audit Your SEO Agency in 14 Days
If you’re reading this because you’re unsure whether your current SEO agency is delivering, stop guessing. Book an audit. We run a 14-day SEO agency audit for Rs 1.5L that produces a written recommendation (keep, restructure, bake-off, or replace) with a cost-benefit breakdown of each path. The audit includes technical review of your site, content asset review, reporting cadence review, and agency interview (with your written approval).
For founders evaluating a new agency before signing, we offer a shorter 5-day pitch-review service at Rs 50K. You send us the deck, the SOW, and the references. We come back with the seven questions above mapped to the agency’s specific pitch, along with the red flags we’d flag before you sign.
Relying on traditional metrics means you are measuring an increasingly irrelevant portion of the search landscape. Focusing on classic keyword rankings ignores that AI Overviews and chatbots now intercept high-intent user queries, making old signals incomplete for assessing performance. A modern agency understands that success means getting your brand cited directly within these AI-generated answers where buyers are making decisions. According to industry analysis, AI surfaces can account for between 8 and 34 percent of search impressions, a share that is growing rapidly. For example, an audit by upGrowth Digital might reveal a significant portion of your potential traffic is being answered by AI before ever seeing your website link. Tracking AI surface traffic is not just a new metric, it is a direct measure of your visibility in the modern buyer's journey. Explore the full article to learn how to ask the right questions and ensure your agency is not stuck in the past.
Topical authority is a measure of how comprehensively your website covers a subject, making it a trusted source for both users and AI systems. It is more crucial than backlinks alone because AI models prioritize citing sources that demonstrate deep, interconnected knowledge, not just those with many inbound links. An agency building topical authority will create a strategic content cluster, which is a network of interlinked articles covering a core topic from every angle of the buyer's journey. This approach signals expertise, increasing the likelihood of being featured in AI responses that capture 14 percent or more of organic sessions in some sectors. While backlinks are still a signal, an agency like upGrowth Digital would argue that a well-structured content architecture is the foundation for earning AI citations. This strategy moves beyond ranking for a keyword to owning the entire conversation around a user's intent. Ask prospective agencies to detail their cluster-building process to gauge their modern capabilities.
An agency focused on outputs measures activity, while an outcome-focused agency measures business impact. The former will show you reports filled with link counts and ranking charts, which look good but may not correlate with revenue. The latter, however, will report on metrics like the percentage of traffic from AI surfaces, directly tying their work to how modern buyers discover solutions. Audits by upGrowth Digital consistently show a widening performance gap between these two models. An outcome-driven agency frames its success around your success, not the completion of a task list. When evaluating, weigh these factors:
Reporting: Do they report on AI surface traffic, which can be 14 percent for B2B SaaS, or just organic sessions?
Strategy: Is their plan about a content cadence or about building topical authority to dominate specific buyer intents?
Tools: Are they using modern tools to track brand mentions in AI or just standard rank trackers?
A focus on outcomes demonstrates a deeper understanding of how search generates value in 2026. The full analysis provides more detail on spotting the difference during the pitch process.
A modern content cluster is an engineered system designed for AI, not just a collection of related blog posts. While older strategies focused on a pillar page and spokes linked by keywords, a 2026 approach is far more sophisticated, incorporating AI query mapping and schema architecture to signal expertise to machine learning models. This is the difference between building a library and just stacking books. A forward-thinking agency, like the kind upGrowth Digital would recommend, will present a plan that includes entity associations and a structured plan for mesh internal linking to distribute authority effectively. Unlike older models, a modern cluster is designed to answer questions that will be pulled into AI Overviews, which now intercept between 8 and 34 percent of impressions. You should look for a proposal that explicitly outlines their strategy for FAQ schema, internal link architecture, and the publishing cadence required to establish authority. See the full guide for a checklist of what a modern cluster strategy must include.
To verify an agency's capabilities, you must move beyond slide decks and demand access to their actual reporting methodology. A competent agency should be able to immediately show you how they track traffic from AI surfaces like Google AI Overviews, ChatGPT, and Perplexity, breaking it down as a percentage of total organic traffic. The claim that B2B SaaS clients see 14 percent of traffic from these sources, up from 4 percent a year prior, is a benchmark you can use. Ask them to demonstrate their methodology, which should involve a combination of UTM filtering, referrer analysis, and brand mention tracking across AI responses. If they claim success for a client, ask for an anonymized report showing the specific queries generating AI citations and the resulting click-throughs. The analysis from upGrowth Digital shows that agencies unable to produce this data within 48 hours are likely not tracking it at all. The full article explains how to interpret these reports for red flags.
The audits reveal a clear dividing line between agencies prepared for 2026 and those stuck in the past. Underperforming agencies consistently frame their work around legacy outputs: keyword rankings, domain authority scores, and the number of links built. They often treat AI as an 'experimental' feature to be addressed later. In contrast, successful agencies measure themselves on business outcomes, such as winning high-intent queries on platforms like Perplexity and getting clients cited in AI Overviews, which can represent up to 34 percent of impressions. Thriving agencies operate as strategic partners who understand the entire buyer journey, not just the classic blue-link search results page. A key differentiator is their ability to answer, in seconds, what percentage of their clients' traffic comes from AI surfaces. A vague or dismissive response is a significant red flag indicating they are not equipped for modern search. Dig deeper into the seven critical due-diligence questions outlined in the full post.
To effectively vet an SEO agency, structure your evaluation to expose their readiness for modern search from the very first interaction. An organized approach prevents you from being swayed by a polished but outdated pitch. The key is to move from broad strategy to specific, data-driven proof. Here is a clear plan:
Initial Screening: Before the main pitch, ask: 'What percentage of your clients' organic traffic comes from AI surfaces?' An agency like upGrowth Digital would expect a swift, precise answer. A delay is your first filter.
The Pitch Meeting: When they present case studies, ask how they built topical authority and what their content cluster process looks like. Listen for specifics on schema, entity mapping, and internal linking, not just 'quality content.'
Data Verification: Request a follow-up report detailing their methodology for tracking AI traffic. The data should support claims like the 8 to 34 percent of impressions now happening on AI surfaces.
This structured process forces agencies to demonstrate modern capabilities rather than just talking about them. The complete article offers more probing questions for each stage.
You can guide your current agency by formally reframing your goals and key performance indicators. Schedule a strategic review and present them with the challenge of tracking and growing traffic from AI surfaces. Request a formal plan within 30 days detailing how they will build topical authority and measure citations in AI Overviews and chatbots. This plan should include new reporting that tracks this specific traffic source, which for some verticals already accounts for 14 percent or more of organic sessions. A capable partner will embrace this challenge. However, as noted in audits by firms like upGrowth Digital, there are critical red flags. The primary warning sign is pushback, such as claiming 'AI is just a fad' or 'the data is not reliable.' If they cannot provide a concrete plan or produce the requested report, it signals they lack the skills and are operating on an obsolete model. At that point, it is time to seek a partner built for the future of search.
Continuing with a traditional SEO strategy means you are effectively becoming invisible to a large and growing segment of your audience. The long-term implications are severe: a steady erosion of market share, declining lead quality, and a wasted marketing budget. Your competitors who adapt will capture the high-intent queries that AI answers directly, positioning their brands as the authoritative solution before a user even clicks a link. The up to 34 percent of impressions being intercepted is not just traffic; it is often the most valuable, bottom-of-funnel traffic. As models like Gemini become more integrated, this percentage will only grow. Audits by firms such as upGrowth Digital show that companies failing to adapt see their organic growth stall and then decline within 12-18 months. The choice is between evolving your strategy to be cited by AI or slowly being erased from the buyer's journey. Find out how to future-proof your strategy in the full article.
The integration of advanced AI is shifting the role of an SEO agency from technical optimizers to strategic visibility partners. The job is no longer just about optimizing a website for crawlers; it is about structuring information so that your brand's expertise can be easily understood, verified, and cited by AI models. This requires a new blend of skills. Core competencies now include data science, content architecture, and a deep understanding of natural language processing. An agency's value is measured by its ability to win share-of-intent across all search surfaces, not just ranking on a list of blue links. For instance, knowing that 14 percent of a B2B SaaS client's traffic comes from AI surfaces is the starting point; the real work is reverse-engineering why and scaling that success. As evidenced by upGrowth Digital’s findings, agencies that cannot evolve these skills are becoming obsolete. You should expect your partner to be as much a data analyst and content strategist as a technical SEO.
The core problem with these metrics is that they measure activity, not business outcomes. A top ranking for a keyword is worthless if an AI Overview now captures all the click-through traffic for that query. This common mistake, highlighted in audits by firms like upGrowth Digital, leads businesses to invest in strategies that no longer drive meaningful results. To solve this, you must reframe the conversation around business impact. Instead of asking 'What keywords will you rank us for?' ask 'How will you get our solution cited in AI responses for high-intent buyer queries?' This shifts the focus from outputs to outcomes.
Challenge them to connect every proposed activity to a revenue-generating goal.
Demand they report on metrics that reflect the modern search journey, like the share of traffic from AI surfaces.
Ask how they measure success for queries where AI intercepts up to 34 percent of impressions.
By controlling the evaluation criteria, you force them to prove their value in the context of 2026, not 2019. Read the full post for more ways to reframe the agency conversation.
This common internal misalignment stems from outdated success metrics and a lack of awareness about the seismic shift in search behavior. The solution requires a strategic re-education focused on new data and redefined goals. Start by presenting the evidence: AI now intercepts a significant share of search impressions, with reports showing that 14 percent of organic sessions for B2B clients already come from AI surfaces. Frame the pivot not as an abandonment of content but as an evolution toward creating more strategic, interconnected assets designed to establish topical authority. Introduce a new set of primary KPIs, such as 'brand citations in AI responses' and 'traffic from AI referrers,' making these the team's new north star. You can use case studies, such as the findings from upGrowth Digital’s audits, to illustrate the performance gap between old and new methods. This data-driven approach shifts the conversation from opinion to strategy, helping you build the internal consensus needed to adapt.
Amol has helped catalyse business growth with his strategic & data-driven methodologies. With a decade of experience in the field of marketing, he has donned multiple hats, from channel optimization, data analytics and creative brand positioning to growth engineering and sales.
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
Generative AI Content Creation
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
