How Fyle Increased Demos by 34% While Cutting Cost per Demo by 27% in Nine Months of Paid Search

About Fyle

Fyle (now Sage Expense Management) is an expense management platform that simplifies expense reporting and reimbursement by integrating directly with tools teams already use: QuickBooks, NetSuite, and corporate credit cards.

Their core audience is finance teams and operations leaders at small and mid-sized businesses looking for a modern alternative to legacy expense tools.

Fyle had already built a solid paid search foundation with significant monthly spend and steady lead volume. But they had hit a point where more spend was not producing more results.

Over February to October 2025, TripleDart audited and tightened the existing campaigns, then added new keyword themes, competitor campaigns, a portfolio bid strategy, and daily guardrail reporting.

This PPC case study shows how Fyle increased demos by 34% with an expense management Google Ads strategy built for an account that had hit its ceiling. Demos booked rose 34%, cost per demo fell 27%, and paid search revenue grew 30%, including through the mid-flight rebrand after Sage acquired Fyle in July 2025.

TripleDart’s Approach: Stabilize the Account, Then Expand Search Intent

Timeframe: February to October 2025 (nine months). PPC and paid media.

The engagement started by making the existing campaigns more efficient. From there, we expanded into new keyword themes, competitor campaigns, and eventually new channels, layering growth on top of a stabilized foundation.

Phase Overview

‍

Phase 1: Getting More from What Already Existed

Before adding anything new, the first move was figuring out where existing budget was being wasted.

We audited every campaign and ad group. The ones consistently running above target CPA were paused, and that budget was moved into campaigns that were already converting well. We also cleaned up conversion tracking so Google's algorithm was optimizing toward qualified demos rather than generic form fills.

The goal was a leaner, more efficient account that could absorb future investment without dragging down blended performance.

Phase 2: Going After Untapped Search Intent

Fyle's paid search coverage was concentrated around a relatively narrow set of business keywords. Effective at a certain scale, but it meant large pockets of relevant search intent were being left on the table.

Coverage was expanded in a few targeted ways:

• Dynamic Search Ads opened up queries the team had not manually targeted, especially around expense tracking for verticals like non-profits and construction.
• New keyword themes around receipt tracking and expense reporting filled coverage gaps across the full buyer journey.
• Three keyword cohorts were established to organize the account: Cards (QB card searches), Expenses (general expense management), and Receipts (receipt tracking).

Each new campaign type was selected based on where there was clear, untapped commercial intent. 

Portfolio Bid Strategy

To manage budget allocation across a growing number of campaigns without micromanaging each one, we introduced a portfolio bid strategy. Google's algorithm could now distribute spend toward campaigns and ad groups most likely to convert at or below the target CPA.

A set of leading metric guardrails kept things in check:

When these leading metrics stayed within guardrails, MQL outcomes were consistent regardless of day-to-day CPA swings. 

This shifted the operating model. Instead of reacting to lagging CPA data, daily “6 AM reports” on impression share and click share enabled corrective action before spend was wasted.

Phase 3: Capturing Competitor Search Intent

With the foundation stable, the next layer was competitor campaigns targeting users actively evaluating alternatives. We launched campaigns across seven competitor brands, prioritized by US search volume and conversion potential.

The highest-intent competitor terms converted well from the start. Newer targets were added after evaluating search volume and bidding viability in the US market. Those SaaS paid search competitor campaigns became a steady source of demos while cost per demo kept falling.

Solving the Search Term Waste Problem

A big chunk of search query spend was going to queries that drove zero conversions.

The most common offenders were:

• Generic navigational searches (people looking for integration partners, not the product itself)
• Consumer finance and budgeting app terms with no B2B relevance
• Broad brand queries with no buying intent

A structured negative keyword program was put in place, with blocklists updated multiple times per week. The cadence was adjusted based on spend levels and new campaign launches.

Operational Decisions That Moved the Needle

A few focused operational changes contributed to cost performance beyond campaign structure:

• Weekday scheduling: Fyle's B2B audience converted during business hours. Pausing weekend campaigns reduced wasted impressions and concentrated budget on peak-intent windows.
• Integration-led messaging: QuickBooks and NetSuite integration campaigns resonated with Fyle's core audience and became a consistent performer.
• Automated monitoring: An alert system built via Google Ads API and Make triggered notifications when key metrics fell outside guardrails. Real-time alerts paired with daily evening performance updates meant problems were caught early.

Navigating the Sage Rebrand

When Sage acquired Fyle in July 2025, the campaigns needed to transition to the new brand identity by October. This kind of mid-flight rebrand creates real complexity in paid search: trademark policies change, brand search behavior shifts, and CPCs can move unpredictably.

The team managed the transition by running multiple brand ad variants in parallel to isolate what was working under the new identity, diverting spend toward non-brand cohorts that were unaffected, and building daily performance checkpoints to catch and address shifts quickly.

Extra budget but no clear way to deploy it? Book a strategy call and we will start by auditing where current spend is wasted.