How Paddle turned a stringent FinTech rulebook into a scalable, AI-mapped qualified market

The challenge: rules too strict for off-the-shelf data to map

As a FinTech company focused on SaaS, Paddle needed a clear idea of which markets and territories to pursue, and how many in-market accounts were a true reflection of its qualified market. With stringent rules on who could become a customer, it was a challenge to reflect those requirements when quantifying future opportunities and coordinating go-to-market efforts. Paddle needed a platform that could accurately detect and refine data for both ideal and bad-fit accounts.

A limiting sales criteria

Unlike most industries, Paddle’s in-house rules for the right account are extremely tight. Due to the nature of the FinTech sector, a customer doesn’t just need to match Paddle’s product capabilities — there’s also a legal component built into identifying the right accounts, a serious responsibility alongside the pressure of winning them.

Despite having clarity on who Paddle couldn’t support, mapping the market of “right customers” still proved a challenge. Historically, Paddle tried to identify accounts using technographics and industry filters, which produced false positives and negatives from mislabelled software companies, and poor technographic accuracy and coverage from other providers.

Even once Paddle identified a software company, the qualification became stricter still — accounts were only eligible if their product sold via the kind of online checkout Paddle’s infrastructure is built to support.

Unclear market estimations

Beyond having specific rules to follow, Paddle needed a better understanding of how many qualifying accounts existed, and how many net new accounts would emerge year on year. Without that visibility, it was difficult to confidently hire or scale commercial teams, or to work out how the product needed to adapt as that market grew. The result was stagnated numbers, unusable accounts and under-serviced team members.

The solution: a truly refined approach

Defined customer criteria

BDRs were the first touchpoint in finding customers, and very often the deciding factor came down to professional instinct — which wasn’t scalable, and not referenceable.

Working with Goodfit, Paddle built scalability in two ways. First, the team defined both good-fit and bad-fit accounts against three questions:

1. What needs to be true for an account to be accepted?
2. What needs to be true for an account to be rejected?
3. What are the data points that define both?

With the criteria defined, Goodfit built two natural language processing models — the first to identify software companies, the second to identify whether a software company was selling offline, via invoices, rather than through an online checkout.

Working with Goodfit, Paddle can now supply accounts to its Sales team with a trustworthy, benchmarked level of accuracy.

The results

Improved accuracy and predictability

Using the NLP models, Goodfit mapped the accounts that “passed” both, giving Paddle a single, reliable view of who to work. In the 12 months before working with Goodfit, only 30% of the accounts being worked by the commercial team met the necessary product and legal requirements. In the month following Goodfit’s implementation, that rate grew to over 70%.

With improved accuracy, clear mapping and net new accounts being created, working with Goodfit has given Paddle overall trust in its market understanding and account distribution going forward.