Southern Marsh had a familiar ambition: sell more clothes. The interesting part was where NoGood went looking. Its published case study describes nine major experiments and four rounds of changes to checkout and navigation. The work included customer interviews, tracking, and a technical audit. The company reports 29% top-line growth, with larger orders and better conversion, particularly on mobile, doing much of the lifting.
- A custom growth team spanning creative, media, engineering, and data.
- Average monthly retainer: above $20,000, according to NoGood.
- The useful habit: find the friction before increasing the budget.
That is a revealing place to begin a profile of a marketing agency. A checkout screen has little glamour. Nobody frames a navigation menu. Yet a company can pay to bring people to its website while making it unnecessarily difficult for them to buy. Improving that last stretch changes what the earlier advertising is worth.
NoGood, a New York growth marketing agency, sells the team that investigates those connections. Its proposition is that creative work, media buying, engineering, and measurement belong in the same conversation. A compelling ad and a functioning purchase journey have to get along.
in NoGood’s case study
02One squad, several suspects
The agency assembles what it calls bespoke squads: combinations of growth strategists, designers, creators, engineers, and data specialists shaped around a client’s problem. Its public team directory makes that combination tangible. Growth managers sit alongside creative designers, community managers, a machine learning engineer, and a web developer. Message-makers and measurement specialists share the assignment.
The service menu reaches from paid search and social advertising to organic social, search engine optimization, content, lifecycle marketing, and conversion optimization. Consulting and fractional marketing leadership add strategic help. A SaaS company might need qualified demand and a better onboarding journey; a consumer brand might need creative testing and repeat purchases. The agency works across SaaS, fintech, healthcare, consumer businesses, and enterprise accounts.
For a buyer, the practical distinction is the scope of the assignment. A specialist can manage one channel. An internal team offers day-to-day company knowledge. NoGood’s pitch sits between those choices: an external group able to connect several disciplines. Its publicly named customers include MongoDB, TikTok, Nike, and Intuit. The useful question is which combination of skills a particular account actually needs.
03The work after the click
Invisibly gives the model a second test. The content-access startup needed people to install its app and keep using it. NoGood describes building its analytics setup, testing acquisition channels, developing content, and coordinating email, SMS, and push notifications. It also worked on the product’s user experience. Acquisition and retention were treated as related problems.
The agency reports a 65% reduction in customer acquisition cost and a 30% reduction in spend in the first quarter of 2023. It separately reports a 400% improvement in retention from the third to fourth quarter of 2022. Those periods matter: these are different measurements, not one giant before-and-after number. They describe the agency’s account of the engagement, rather than a controlled experiment isolating every cause.
The lesson a reader can borrow is the sequence. Instrument the journey. Locate the loss. Test a change. Check whether the people acquired become useful, returning customers. Buying cheaper installs helps less if those people immediately disappear.
“The squad feels like a true extension of my own marketing team.”
Laura Vestal · Head of Marketing, Invisibly
Published in NoGood’s client case study
04When the search box answers back
SteelSeries presented a newer version of the same problem. According to NoGood’s case study, the gaming accessories brand had strong conventional SEO but an inconsistent presence in AI recommendations. The partnership began in late 2024. Search performance on one surface was failing to carry over to another.
NoGood’s response combined clearer, question-focused content with product-page and feed reviews, structured data, authority building, and measurement across ChatGPT, Gemini, and Perplexity. This is answer engine optimization, or AEO: helping a brand become discoverable in generated answers. The work extends the agency’s search practice into places where a buyer may receive a recommendation before seeing a list of links.
The case study’s January-to-October 2025 narrative reports AI traffic rising 6.7 times to more than 1,500 daily visitors. NoGood also names Goodie AI as a technology partner for tracking AI visibility. Its AI Lab offers custom development, automation, and predictive analytics. The commercial proposition is still human work around a business problem, with software helping people observe and act.
05A very expensive place to guess
NoGood says its average monthly retainer exceeds $20,000 and scopes engagements individually. At that average, three months represents more than $60,000 in agency fees. A buyer should budget media separately when evaluating the economics. The retainer purchases specialist capacity; the client still needs an offer customers want and enough room in the economics to pay for acquisition.
There is a useful admission in founder Mostafa ElBermawy’s 2021 interview with Failory. Early clients sometimes wanted to scale before finding product-market fit, frustrating the team. NoGood became more selective. Its own operating system also needed work: staff ignored Monday.com, Slack alerts became noise, and the agency moved to Notion to consolidate experiments, budgets, tasks, and notes. More tools had initially meant more friction.
For prospective customers, those lessons suggest a demanding fit. The team needs permission to change things, usable data, and a product with demonstrated demand. A small business seeking occasional posts may struggle to justify the fee. A company unable to improve its website or product can leave a multidisciplinary squad with surprisingly little to do.
06Steal the question
The agency’s October 2026 Social Insight Report connects its two worlds: human attention and machine discovery. It draws on desk research, a survey of marketers in its network, and Goodie’s AI visibility data. Topics include how social profiles help people evaluate brands and how social content becomes material for AI answers. That is a wider brief than making the next post perform.
The portable idea is to ask a better question before commissioning more marketing. Where does interest stop becoming action? At the ad, the answer, the registration form, or the second visit? Southern Marsh’s checkout changes and SteelSeries’ AI search work concern different technologies. Both begin with a customer trying to make a choice. NoGood’s job is to find what gets in the way.