Rokt started with a straightforward insight: the moment after a customer completes a purchase is the most receptive moment in all of digital commerce. The person has just committed money, their intent is confirmed, and they're still on the retailer's page. For anyone trying to place a relevant offer, that window is worth far more than a banner ad served to a stranger scrolling a news feed. From that premise, Rokt built a transaction marketing platform that now sits inside the checkout flows of some of the world's biggest companies.
The Sydney-founded company was started in 2012 by Bruce Buchanan and a small founding team. It took years of grinding through proof-of-concept deals with Australian retailers before the model proved itself clearly enough to pursue US expansion seriously. That expansion now defines the business. Rokt is headquartered in New York, but its engineering and product roots remain in Sydney, and the company still carries an Australian identity in the local tech conversation.
What Rokt actually does
Rokt's platform sits between a retailer's purchase confirmation page and a curated pool of advertisers. When a customer finishes a transaction on a site like Ticketmaster, Staples, or HelloFresh, Rokt's machine learning layer decides in real time which offer, if any, to show that customer. The offer might be a loyalty programme signup, a subscription trial, or a complementary product from a brand that has paid to reach people in exactly that buying state.
The key commercial claim is relevance at scale. Rokt charges advertisers on a cost-per-acquisition basis rather than cost-per-click, so the incentive structure is tied to actual customer action. Retailers receive a revenue share. Because the offer sits in the post-purchase flow rather than interrupting the purchase itself, the retailer's conversion rate is unaffected. That framing made the pitch easier to sell to e-commerce teams who are understandably protective of their checkout experience.
The machine learning layer is the core of what Rokt sells. It draws on transaction signals across its network of partners to predict offer relevance for each individual customer. A customer who just bought a concert ticket is a different prospect from someone who just bought office supplies, even if they're in the same demographic. Rokt's argument is that this transactional context makes its targeting more precise than almost any other digital channel.
The path from Sydney to New York
Buchanan spent several years building the business in Australia before concluding that the scale of the US market was necessary to make the model work at its full potential. The company opened its New York office and began signing US partners around 2016. Growth from that point was steady rather than explosive: Rokt isn't the kind of company that raised enormous funding rounds at spectacular valuations and then burned through them. Its expansion was operationally disciplined.
By the early 2020s Rokt had accumulated a client list that included some of the highest-traffic e-commerce properties in the world. That roster is what generates the data advantage. The more transactions Rokt processes, the better its models become at predicting which offers convert for which customers in which purchase contexts. This flywheel is the same logic that underpins most successful ad networks, but applied specifically to the transactional moment.
Rokt has raised substantial venture funding over the years, with investors including Tiger Global. In 2023 the company was reported to be processing more than 5 billion transactions annually. That figure puts it in a different category from most Australian-founded adtech businesses, which tend to stay regional or get acquired before reaching that kind of global footprint.
Commerce media and where Rokt fits
The broader category Rokt operates in is now called commerce media. Retailers with large customer bases are monetising their audience data and on-site real estate by selling advertising to brands that want to reach shoppers in buying mode. Amazon's advertising business is the most prominent example. Retail media networks run by Woolworths (Cartology) and Coles (Coles 360) are the local equivalents. Rokt sits slightly differently: it works across retailers rather than building a single retailer's network, acting as the infrastructure layer for post-purchase monetisation.
This positioning gives Rokt access to aggregated transactional data across many partners simultaneously, which is commercially valuable but also creates questions about data governance. As Australia's Notifiable Data Breaches scheme places tighter obligations on how customer data is handled and disclosed, any company processing transaction-level data at scale faces increasing compliance attention. Rokt processes data under US jurisdiction primarily, but its Australian engineering operations and heritage mean local regulatory developments are relevant context.
AI as the next competitive layer
Rokt has been public about embedding large language model capabilities into its platform, using generative AI to personalise offer copy and creative in real time. The idea is that a generic offer headline converts worse than one tuned to the specific purchase context, and that AI can generate enough variation to test and optimise at a speed no human copywriting team can match.
This is consistent with where the broader digital advertising industry is heading. Personalisation at the creative layer, not just the targeting layer, is becoming the standard expectation. For Rokt, it's also a way to deepen the moat: retailers and advertisers who are embedded in a platform that learns continuously from billions of transactions are harder to displace than those using a static tool. Australian enterprises evaluating AI's role in their own revenue operations should note how companies like Rokt are operationalising AI model drift management in production, keeping models accurate as consumer behaviour shifts across seasons and market conditions.
What Rokt's trajectory says about Australian tech
Rokt's story is one of patient, unglamorous company building. It isn't a company that dominated Australian tech media during its growth years. It didn't take a splashy ASX listing or run the kind of funding rounds that generate press. It built a specific capability, proved it in the hardest market, and scaled it into a global business.
That pattern appears elsewhere in Australian enterprise tech. Companies like WiseTech Global built deep vertical software in freight logistics for years before the broader market paid attention. The common thread is a focus on proprietary data and network effects rather than product surface area. Rokt's transaction data across thousands of commerce partners is not something a well-funded competitor can replicate quickly, even with superior technology.
Whether Rokt pursues a public listing is a question the company hasn't answered publicly. The business has the scale and revenue profile that would make a Nasdaq listing viable. It also has the option of continuing as a private, profitable operator in a growing category. Either way, it stands as one of the more substantial software businesses Australia has produced in the current generation. Not many Australian-founded companies can claim to sit inside the checkout flow of global retailers processing billions of annual transactions. Rokt can.