Most content about running an agency is a highlight reel. What is harder to find is an honest account of what a decade actually takes: the pivots, the top-heavy payroll, the eight-month plateau, the hire you should have made two years earlier. On an episode of MerchantSpring Marketplace Masters, Hamish Conway, founder and managing director of the Amazon agency Sell Global, marked ten years in business by talking through exactly that. Here are the lessons most worth stealing if you run, or are building, a marketplace agency.
From Accidental Agency to Enterprise Specialist
Sell Global started the way many agencies do: by accident. Conway was selling his own products on Amazon in the baby category, expanding into the UK and Europe and clearing a million dollars in his first full year, when other brands began asking him to help with their strategy. He already had the other half of the equation, having run several service businesses before, and the two skill sets met in an agency. The original vision was to help New Zealand companies export to the US, UK and Europe via Amazon.
Two pivots reshaped that. The first was geographic. Amazon Australia launched in December 2017 (ABC News), was quiet at first, then stepped up sharply through the 2020 COVID rush. Conway decided to double down with a clear vision: be the best Amazon specialist agency in Australia. The second pivot followed from it, a shift from lesser-known exporters to established brands, landing clients like Weber and Asahi's Australian beverages portfolio.
The logic was commercial. Australian vendors had mostly been onboarded fast and set up poorly, so an agency that knew what good looked like from the US could deliver significant results in about two months, versus the long, budget-blowing grind of taking an unknown brand from zero into America. A faster path to a client win is worth a great deal.
The Skill That Transferred Best Was Selling
Asked which of his pre-Amazon experiences mattered most, Conway didn't name a marketing or operations skill. It was selling, specifically the ability to step into a prospect's shoes, understand what they are dealing with, build rapport and trust, and convert. For an agency, whose product is intangible and whose results take time to show, the capacity to sell professional services credibly is foundational. A great service that can't be sold is not a business.
Hiring: The Top-heavy Trap, and Why You Need an Operating System
Early on, Amazon expertise was scarce in the region, so Conway recruited from the people he had helped start their own brands and from the FBA and private-label communities, operators who understood Amazon because they lived it. The first hires were support and ads help, but the business quickly needed senior, client-facing people who could hold strong conversations, and he leaned on people he already knew and trusted.
The trap came next. Enthusiastic about senior talent, he sometimes hired ahead of need and the agency became top-heavy, more senior people than support, which put real pressure on profitability. Growth also demanded far more coordination than he expected, and a misalignment opened up at the leadership level that took outside help and a tense restructure to resolve. The turning point was building a genuine leadership team, with partners owning the ad and delivery sides, so the business no longer depended on him for everything.
He is candid about the symptom that forced it: a point where he couldn't sell any more because operations would break and refuse new clients, the tail wagging the dog. The deeper lesson applies to every agency built on good people: talented, trustworthy hires will look after clients, but until you have an operating system rather than a set of heroes, it stays fragile. The team today is around 21, split roughly between the Philippines and Australia and New Zealand, with the senior-to-support ratio deliberately fixed.
The Mistakes He'd Skip
Conway is unusually frank about the detours. The biggest was staying too long in a coaching model built on one-off payments, when the agency's done-for-you model produced recurring revenue and much higher lifetime value; his average client tenure now runs about two and a half years, with some clients approaching a decade. A related error was a push into multiple marketplaces beyond Amazon, which mostly added complexity and made the business harder to scale, a case of trying to do too many things at once.
Then there is the perennial one: spending too much time on the wrong clients, brands that were always going to be hard to make work, which is a bad experience for everyone. His filter now is simply whether he can genuinely help, and the agency stopped chasing one-off setups in favour of recurring work.
The subtlest mistake was deciding too slowly, especially about people. Fearing the fallout, he held on to structures and roles past their moment, and in hindsight the calls always should have come earlier. His framing is gentle but firm: the people who get you to one stage are not always the ones who get you to the next, and nobody has to have done anything wrong for that to be true. Sitting with that ambiguity, rather than avoiding it, is part of the job.
The Vendor Bet That Paid Off
One decision Conway clearly got right was expanding from a seller (3P) focus into Amazon Vendor (1P), without abandoning 3P. The pull was that the enterprise brands he wanted lived on Vendor Central, so the agency asked its core question, what problems can we solve for them? Advertising was the obvious first win and delivered fast, because those brands were already set up and winning the buy box. From there, new revenue streams opened up in the harder problems: negotiating with Amazon, and managing shortages.
Much of the capability was built through conversations, learning, and using data to support annual vendor negotiations, pinpointing where a brand is strong and which terms are worth pushing on. Conway also runs vendor mastermind meetups, which give the agency a cross-brand view of what good looks like, and has since hired people who sold on the vendor side themselves. For agency owners weighing the same move, MerchantSpring's guide on the terms you can and should push back on in Amazon vendor negotiations is a useful primer on where the leverage sits.
Culture Across a Distributed Team
With a team spread across the Philippines, Australia and New Zealand, culture cannot be left to the office. Conway's approach starts with hiring well, then runs on rhythm. The all-hands moved to a monthly hour once weekly got unwieldy, covering how the month went, the key focus and the wins, with people sharing what is working; a separate mid-month moment is purely cultural, built around shout-outs for helping each other or landing client results. Underneath sits a deliberate cadence of one-on-ones led by strong team leaders, so no one is left out and the load doesn't all fall on the founder.
His deeper point is that culture is what people see you do. The agency talks constantly about caring for client results, because when clients win the agency wins, and it backs that with visibly hard calls: paying money back when warranted, saying no to work that isn't right. He reaches for Stephen Covey's idea of an emotional bank account, the small, constant deposits that build trust over time, and adds a blunt final ingredient: pay people properly. For more on how agencies operationalise this kind of best practice, MerchantSpring's resources for marketplace agencies and advisors are built around exactly this audience.
What the Next Decade of Agencies Will Demand
Looking forward, Conway believes the billable-hours model is finished; clients care about the outcome, not how you get there, so agencies have to get ruthlessly clear on the result they deliver. AI will run through everything, compressing research and speeding better decisions, but it raises rather than lowers the value of the human layer: people with strong commercial judgement who can ask the right questions and act as genuine advisors, drawing on the cross-client perspective only an agency has.
He is also investing in turning aggregated, normalised data into insights clients can't easily get elsewhere. And he is clear that the team matters as much as the tooling; people want growth, want to build AI skills, and want to feel part of something, so a founder's job includes helping them get what they want from their working life. The through-line across all ten years is the same: solve real problems, keep the human touch, and build the system before the heroics burn out.
Run Every Client Account From One Workspace
Conway's agency uses MerchantSpring to see client performance and to arm vendor negotiations with data, which is exactly what the platform is built for. MerchantSpring brings every seller and vendor account into one agency workspace, with scheduled client-ready reporting and the profitability signals that turn a negotiation or a QBR from opinion into evidence. See how agencies use MerchantSpring, or book a walkthrough of your own client accounts.
