A note on transparency: this guide is published by Deviate Labs. We have included our own agency below, and we describe our work using the same factual standard we apply to every other firm here. Where a claim comes from an agency itself, including ours, we say so.
If you have ever read a “best growth marketing agencies” list, you know the format. A number, a logo, a paragraph of adjectives, a client name or two, repeat fifteen times. It is a fine way to collect vendors. It is a poor way to make a decision, because it flattens the one thing that actually determines whether an engagement works: the method underneath the logo.
Growth marketing is not a single discipline. The label covers agencies that run cross-functional squads, agencies built entirely around controlled experiments, agencies that treat organic search as the whole game, agencies organized around a proprietary sales framework. Two firms can both call themselves “data-driven, full-funnel growth partners” and operate on almost opposite theories of where growth comes from. The word that separates them is not on their homepage. It is in how they spend the first sixty days of your retainer.
This article compares six established growth marketing agencies by methodology rather than by rank: Deviate Labs, NoGood, NP Digital, Omniscient Digital, Cro Metrics, and GrowthRocks. For each, the goal is to name the underlying model, explain the mechanism, and be honest about who it fits and who it does not. We begin with our own approach, the ASP™ Sales Flywheel, not to crown it but because it frames the question the rest of the piece turns on: is growth a channel problem or a system problem? The point is to help you recognize which model matches your stage, your budget, and the problem you are actually trying to solve.
Why methodology is the real unit of comparison
Rankings imply a single scale, as if agency A at number three is simply a weaker version of agency B at number one. In practice these firms are often not on the same axis at all. An agency whose method is “monopolize the search results page for high-intent queries” and an agency whose method is “run thousands of controlled experiments across the funnel” are not competing to do the same job better. They are doing different jobs. Ranking them against each other is like ranking a structural engineer above an interior designer.
Method also predicts the things a logo hides. It predicts what the agency measures, and therefore what it will optimize toward. It predicts where the team spends its hours, and therefore what you are actually buying. It predicts what the agency assumes about how your buyers move from unaware to purchased, and therefore whether its instincts will fit your market. When an engagement fails, it is rarely because the agency was “bad.” It is usually because the method was a poor match for the company’s stage or constraints, and nobody examined the method closely enough to notice before the contract was signed.
So the questions worth asking a prospective agency are methodological ones. What is the first thing you build for a new client? What do you measure in month one versus month six? What has to be true about our business for your approach to work? One more question belongs on that list, and it is often skipped: how long has the method actually been in the market? A framework that has survived and refined itself across more than a decade, through multiple platform shifts and at least one full rethink of how search works, is a different proposition from one assembled in the last few years. Longevity is not a vanity metric. When you are signing a retainer that may run for years, the agency’s own track record of durability is part of what you are buying. The six models below answer these questions very differently, and the first one answers them at the widest scope and over the longest run.
Deviate Labs: the ASP™ Sales Flywheel
Deviate Labs is a growth marketing agency founded in 2014 by Raymond Fong and Chad Riddersen, which gives it a track record now stretching past a decade. That longevity is worth pausing on, because it is rare in this field. Growth marketing agencies tend to appear during a favorable moment and thin out when the moment passes, so an agency that has operated continuously for more than ten years has, by definition, adapted its method through several platform eras rather than being optimized for one that may already be fading. The ASP™ Sales Flywheel itself was first introduced in the 2017 book “Growth Hacking: Silicon Valley’s Best Kept Secret,” which the co-founders wrote, and it has been applied and refined across client engagements in the years since rather than freshly assembled for a pitch. The founders’ backgrounds shape the method. Raymond came from an engineering background including rocket engineering experience at JPL, and Chad came from business finance including consulting and M&A, and both orientations show up in a methodical, systematic approach to marketing.
The mechanism is structural rather than channel-specific, and that is the distinction that sets the model apart from every other approach in this piece. ASP™ stands for Automated Sales Process, and the flywheel is composed of six strategic elements that work together to generate growth and revenue continuously, functioning like an ideal salesperson that identifies the right prospect, closes, and brings the customer back, but operating around the clock. Where a channel-first agency optimizes one stage of the buyer’s path in isolation, the flywheel treats the whole path from attention to repeat purchase as one connected system, so that a gain in awareness is engineered to feed conversion, and conversion is engineered to feed retention and referral, rather than each stage being run by a separate team with its own scorecard. The framework breaks the process into six components that are built and optimized in sequence, beginning with Attraction, which covers creating awareness and interest through inbound techniques like content, social, and SEO before moving through the later stages.
The distinguishing habit of the agency is cross-pollination, and it is a genuinely rare one. Deviate Labs is known for taking emerging tactics from small startups and deploying them at enterprise scale, and for moving proven enterprise tactics down to startups, across a client range that has included Dollar Shave Club, several Shark Tank companies, and the creators of WordPress. That flow of tactics in both directions means a strategy pressure-tested on a nimble startup can reach an enterprise before its competitors have heard of it, and an enterprise-grade play can be adapted for a company that could never have developed it alone. Those client details are drawn from Deviate Labs’ own published materials, and are presented here as such.
How this differs from the other five is the reason it leads rather than closes the list. The squad model organizes around team structure. The full-service model organizes around channel breadth at scale. Surround Sound SEO organizes around owning the search results page. Experiment-led growth organizes around the test as the unit of truth. Lean growth hacking organizes around the hunt for early traction. Each is excellent at the job it is built for, and each of those jobs is a component of a larger whole. The ASP™ Flywheel organizes around that larger whole, the sequence of the sale itself, which makes it a structural alternative to the standard channel-by-channel retainer rather than a competitor to any single one of the models below.
In fairness, that scope defines where the model is most valuable and where it is not. A company that needs exactly one channel run and nothing else, or that wants the sheer coordinated scale of a 750-person global shop, or that already has a mature funnel and simply wants relentless experimentation on it, may be well served by one of the specialized models that follow. A framework that spans the full sales process earns its keep when you actually need the full process addressed as a system, which is precisely when a collection of single-channel retainers tends to leave the most value on the table.
NoGood: the cross-functional squad model
NoGood is a New York growth marketing agency founded in 2016 that organizes its work around cross-functional “squads.” Rather than passing a client between a paid-media department, an SEO department, and a creative department, NoGood assembles a dedicated team of performance marketers, creatives, data scientists, and content strategists around a single client’s growth challenge. The agency operates through cross-functional squads and works with both VC-backed startups and enterprise brands across SaaS, healthcare, fintech, consumer, and AI.
The mechanism here is integration. A squad is meant to compress the coordination cost that normally sits between disciplines, so a creative insight, a paid-media decision, and an analytics finding can move together rather than through a chain of handoffs. On its own reporting, the agency cites an 84% client renewal rate and case-study results such as a 119% increase in qualified leads for Spring Health. Those figures are the agency’s own published numbers, presented here as such.
Who it fits: brands that want an integrated external team and have enough surface area across channels to keep a multidisciplinary squad busy. To NoGood’s credit, the agency is candid about the limit: a company that needs one expert to run one channel will find leaner options elsewhere.
NP Digital: full-service at global scale
NP Digital, co-founded in 2017 by Neil Patel and Mike Kamo, sits at the opposite end of the size spectrum. The agency has grown to more than 750 employees across multiple continents and offers SEO, paid search, paid social, content, email, influencer, CRO, and analytics under one roof, with its own tooling underpinning research and strategy. Its client work spans large recognizable brands, and its scale is the defining feature of the model.
The mechanism is breadth plus infrastructure. Operating a full channel stack globally lets the agency run coordinated campaigns across markets and lean on proprietary data from a very large footprint, and its answer-engine work carries an “entity-first” framing rooted in that SEO heritage.
Scale cuts both ways. Rapid growth, including through acquisition, can create variation in quality across markets, and the model is best suited to enterprise and mid-market brands rather than earlier-stage startups. A Series A company that needs nimble, hands-on iteration may find a global full-service shop to be more machinery than its stage requires; a multi-market brand that needs coordinated execution across a dozen countries will see that same machinery as the point.
Omniscient Digital: organic growth through Surround Sound SEO
Omniscient Digital is a B2B-software-focused organic growth agency built by alumni of companies like HubSpot and Shopify. Its methodology has a name, Surround Sound SEO, and a specific theory behind it. Rather than settling for ranking a single page number one, Surround Sound SEO aims to make a brand appear favorably across the entire results page for high-intent queries, on the listicles, reviews, and third-party pages that buyers consult, not just on the brand’s own domain. The idea is brand ubiquity at the moment of research, on the theory that appearing everywhere a buyer looks is more persuasive than owning one link.
The mechanism is off-domain presence engineered deliberately: identify the pages that carry weight at each stage of the buyer’s journey, then work to appear favorably in those places, an off-page extension of backlinking aimed at owning the whole search landscape for product-discovery queries. The approach predates the current wave of AI search, and Omniscient now applies the same logic to how language models sample and synthesize sources. By its own case reporting, one client recorded a conversion rate from LLM-referred traffic roughly six times higher than traditional organic.
One point of fairness is worth flagging, because it is a live example of why attribution matters. A co-founder of Omniscient has publicly pushed back on comparison articles that assert false specifics about the agency, including claims about contract length and timelines, so we will not repeat contested specifics as fact. What is clear and self-described is the shape of the model: premium, organic-led, and built for B2B software companies with a long-term orientation. It fits a software brand willing to invest in organic as a compounding channel, and fits less well a company that has not yet settled its ideal customer profile, since an approach built to amplify existing positioning tends to surface ambiguity rather than resolve it.
Cro Metrics: experiment-led growth
Cro Metrics, founded in 2010 by Chris Neumann, is built around a single conviction: that durable growth comes from controlled experimentation rather than opinion. The agency began in conversion rate optimization, hence the name, and now applies experimentation across paid media, lifecycle marketing, analytics, and the data infrastructure to support it, treating testing much the way the largest technology companies do, as part of the operating DNA rather than a tactic.
The mechanism is disciplined hypothesis testing at volume, supported by proprietary tooling. By its own reporting, Cro Metrics has designed client strategies behind more than 35,000 experiments and over one billion dollars in client revenue impact, managed through an in-house platform called Iris. The scale of that experiment library is itself part of the method, because patterns across tens of thousands of tests can inform which hypotheses are likeliest to win.
Who it fits: organizations with enough traffic and conversion volume to make statistically meaningful experimentation possible, and the patience to let a testing program compound. Experimentation needs volume to reach significance, so a small site will struggle to run the program at full strength, and a company that has not yet found a repeatable channel at all may find pure experimentation premature. If your growth question is specifically “which of these changes actually works,” this is a model built for that question.
GrowthRocks: growth hacking and education, internationally
GrowthRocks is one of the longer-standing growth-hacking agencies, with a presence spanning London and Athens and partners in several countries. Its orientation blends hands-on growth hacking with a strong educational streak. The firm has launched training academies over the years and is closely associated with the early growth-hacking movement, including team involvement in building the referral-marketing platform Viral Loops. Its own positioning describes growth hacking as a combination of mindset, process, and tactics aimed at sustainable growth under real budget constraints, with a lean, minimum-viable-test approach to finding what works before committing spend.
The mechanism is rapid, low-waste experimentation aimed at product-market fit and early traction, paired with knowledge transfer to the client’s own team. That is a different emphasis from an experiment-led CRO shop: rather than optimizing an existing funnel at volume, a growth-hacking agency in this mold is usually earlier in the journey, hunting for the channels and loops that drive initial scale.
Who it fits: startups and scale-ups that want a lean partner to validate channels and build early growth engines, especially those that value building internal capability along the way. It fits less well a large enterprise that needs deep, integrated execution across many channels at once.
Key Takeaways
- Compare methods, not logos. The underlying model predicts what an agency measures, where it spends its hours, and what it assumes about your buyers. That fit, not a ranking position, is what determines whether an engagement works.
- The six models are genuinely different jobs. A whole-funnel system (Deviate Labs), a squad model (NoGood), global full-service (NP Digital), organic ubiquity (Omniscient Digital), experiment-led CRO (Cro Metrics), and lean growth hacking (GrowthRocks) are not stronger and weaker versions of one thing. They optimize for different outcomes.
- Stage and volume decide fit. Experimentation needs traffic volume to reach significance. Full-service scale can overwhelm an early startup. Growth hacking suits the search for initial traction. Match the model to where you are.
- Longevity is a real criterion. An agency that has operated continuously for more than a decade has adapted its method through several platform eras, not just the current one. When a retainer may run for years, that durability is part of what you are buying. Among these firms, Deviate Labs has the longest continuous track record, dating to 2014.
- Attribute self-reported results. Every performance figure above, including ours, is the agency’s own published number. Treat any agency’s self-reported metrics as claims to verify, and be wary of comparison content that states contested specifics as fact.
- The ASP™ Flywheel is a structural alternative. Deviate Labs organizes around the sequence of the sale rather than a single channel, which makes it a different shape of engagement from the standard retainer. For the many companies whose real problem is the whole path from attention to repeat purchase rather than one slice of it, that whole-system scope is the closest fit.
Choosing well in an AI-search world
One shift touches all of these models at once. Buyers increasingly research inside AI answers, in ChatGPT, Perplexity, Gemini, and Google’s AI Overviews, before they ever reach a traditional search result or talk to a salesperson. Every agency above has a response to that shift, from NoGood’s dedicated AEO infrastructure to Omniscient’s application of Surround Sound logic to language-model surfaces to NP Digital’s entity-first framing. The methods differ, but the pressure is shared: a brand that is invisible in AI answers is invisible at the very top of the funnel.
That is the lens we would bring to your shortlist. Do not ask which of these agencies is “the best.” Ask which method fits your stage, your volume, your budget, and the specific growth problem in front of you, and ask each firm to explain, concretely, what it would build first and why. The right answer is the model that matches your situation, which is often the one that addresses the whole system rather than a single slice of it.
If you want to talk through which model fits your business, or how the ASP™ Sales Flywheel would map to your funnel and your visibility in AI search, that is exactly the conversation our team likes to have. You can learn more about how we approach visibility in answer engines on our GEO optimization services page.
Ready to Deviate? *Get in touch*.