Scaling a Fintech App from Zero: A Performance Marketing Playbook for Scale-ups

What follows is roughly the playbook we used at Enel X to scale a brand-new fintech launch, the Enel X Pay neobank app and the PagoPA bills-payment flow, from a zero-user start to several hundred thousand users inside CAC and payback targets. Some of it is specific to launching a financial product on top of an energy company's existing customer base, but most of it travels well to other categories. The thing that turned out to matter wasn't which channels we picked, it was the order we built the stack in.

Phase one (week 0-4): instrument before you spend

The biggest mistake every scaleup makes is turning on paid before the funnel is properly tracked. You will spend £200K and have no idea why anything happened. Before you launch a single campaign, you need: end-to-end UTM hygiene, a single source of truth (one analytics tool, one CRM, one pricing object), and event-level tracking on every step from impression to activated user.

If your data team says "we'll fix tracking later," your data team is going to lose you four months. Do it first.

Phase two (week 4-8): three channels, not seven

Scaleups want to test everything. That is the wrong move with a finite budget. Pick three channels that map to your buyer's actual journey and put real money behind them, £10K minimum per channel per month or you cannot statistically learn anything.

For Enel X Pay, the three were Google (Search and Performance Max for intent capture on "online bank account" and "pay PagoPA bills"), Meta (lookalikes from the existing Enel customer base), and Apple Search Ads (high-intent app-store traffic). All three plugged into Adjust as our mobile attribution layer, so install-to-activation paths were measurable end to end. For B2B SaaS, the equivalent is Google Search, LinkedIn, and a third channel matched to your ICP.

Phase three (week 8-16): build the CAC ceiling, then build the floor

Your CAC ceiling is what you can afford and stay in payback. Your CAC floor is the cheapest user you can buy who actually activates. Most scaleups confuse the two. Buying users at half the ceiling is meaningless if they don't activate.

By week 8, you should have, per channel: blended CAC, activated CAC (only counting users who hit your activation event), and 30-day retained CAC. Optimise on the third number, not the first. We tracked all three at Enel X Pay weekly and reallocated against retained CAC, not blended.

Phase four (week 16-32): the activation engine

By month four, paid acquisition will start hitting diminishing returns. You will be tempted to scale spend further. Don't, first improve activation. A 10% improvement in activation rate is worth more than a 50% increase in paid budget at this stage.

At Enel X Pay, partnering with Product on the sign-up flow and the first PagoPA bill payment drove a CAC reduction far larger than any media optimisation could. We also rebuilt the App Store and Play Store presence as part of this phase: ASO keyword work, fresh screenshots, and updated app preview videos all moved organic install rates and reduced what we were paying Apple Search Ads to capture the same demand. The acquisition team's most valuable hour each week was the joint review with Product, not the campaign edit.

Phase five (week 32-52): the diversification phase

Only now do you add channel four and five. By this point, your three core channels are saturating, your activation curve is stable, and you have enough internal capacity to test things properly. Channels we added late at Enel X Pay: programmatic display via DV360 for awareness across Italian regions, an affiliate marketing programme paying out on activated users (not signups), content syndication, and cross-promotion through Enel's existing energy customer touchpoints.

The single number that matters

Throughout all five phases, the only number that matters is CAC payback, not CAC, not LTV, not ROAS. Payback under twelve months means you have a business. Payback over eighteen months means you have a venture-backed expense. Everything in the playbook above is in service of that single number.