pdfFiller, an online document editor from Ukrainian unicorn airSlate, has more than 64 million users globally, including those who want to get to know the product with a 30-day free trial. Dmytro Pastukh leads the team responsible for attracting more new users — and converting these customers to paid subscribers.
For subscription-based tech companies like pdfFiller, maintaining consistently high user acquisition rates has become a growing challenge. This is partly due to the rising number of competitors, as research shows that the PDF editor software market is expected to expand at compound annual growth (CAGR) of 9.37% to reach $4.72 billion by 2030.
The market growth is driven by adoption of PDFs as a standard file format and advancement of cloud-based technologies that facilitate secure data sharing. As a result, tech giants invest heavily in developing innovative PDF editing and management tools, offering a wide selection of options.
Source: Global PDF Editor Software Market by Operation (Compress & OCR, Convert To PDF, Convert From PDF), Subscription (Annual, Monthly), End-User - Forecast 2024-2030.
While convincing a potential customer to take part in a free trial is no easy feat, attracting more paid users is an even greater challenge. People have more possibilities to learn about new products and channels to buy than ever before. In times of uncertainty, their needs and preferences are evolving rapidly, making it difficult to reach relevant customers with the right message at the right time.
That’s why we continuously test different strategies and scale up what works best. For example, at pdfFiller we prioritise testing ad copy headlines to understand what resonates best with our audience — and we learnt that mentioning competitor names in headlines is ineffective.
By taking these steps, we move closer to our goal of reaching our ideal customers; those most likely to convert into paid subscribers. Previously, distinguishing between free-trial seekers and those ready to commit was a challenge due to people using similar search queries. To overcome this hurdle, we turned to the power of AI.
Source: Google data. Numbers represent search interest relative to the highest point on the chart for the given region and time.
Starting by getting audience signals right
At pdfFiller, we streamline document management and help users minimise errors with features like fillable forms and custom branding. With one of our experimental campaigns , we aimed to increase product and brand consideration and encourage more people to try our platform. While our ads effectively conveyed product features, they lacked the visual elements to spark emotional engagement. Crucially, we also wanted to prioritise converting those who are most likely to become loyal customers .
Last December we ran a four-week experiment with AI-powered ads. We paused our Dynamic Search Ads (DSA) campaign, which generated text ads based on our website content. We then launched an AI-powered Performance Max (PMax) campaign. While both DSA and PMax use AI, the latter one employs more advanced solutions, like Google’s technology focused on conversion based auto-bidding strategies and audience signals, reaching people who are more likely to become high-value customers. These features strongly aligned with our goals.
We used identical settings for both campaigns, including the same message for creative assets , landing page, bidding strategy, and bid and daily budget . We then compared the PMax campaign's performance in December to the DSA results from November.
The AI-powered ads proved to be a game-changer as they allowed us to reach customers with an immediate purchase intent more effectively.
We were initially hesitant to try PMax, thinking it might compete with our search campaigns for user clicks, inflate costs, and limit our control over performance. Fortunately, our concerns proved to be misguided.
By optimising audience signals with first-party data about users who subscribed to paid plans in the past, we were able to train the algorithm to reach similar people. Notably, unlike other campaign types, PMax allowed us to reach broader audiences beyond those we’d initially chosen. The algorithm can expand its reach if it determines it will benefit the campaign's goals.
Our core challenge was aligning with consumers’ goals and interests at the moment an ad is displayed, a critical window that can determine people’s purchase decision within seconds. The AI-powered ads proved to be a game-changer as they allowed us to reach customers with an immediate purchase more effectively.
While both DSA and PMax campaigns generated free trial users, the latter brought 3X more paid subscribers at an 8% lower customer acquisition cost.
Driving engagement with visuals
The shift to advanced AI solutions allowed us to explore all available ad formats, including images and video, in addition to text ads. Our DSA campaigns relied solely on dynamic text ads that customised messaging based on user search queries. During the experiment, we added images and videos to showcase pdfFiller’s key features, such as PDF editing and e-signing.
This shift to visual storytelling proved to be effective in driving further user engagement. People exposed to ads in the experimental campaign, including visual assets, spent 25% more time on average on our website per session compared to those who only saw dynamic text ads from our original campaign.
While this experimental PMax campaign focused on a general product overview and a “Try it out” message, we are now preparing to showcase pdfFiller's advanced features to reach people specifically interested in these solutions. For example, we have an extensive library of ready-to-use templates tailored for professionals across various fields, from education and insurance to real estate. We have created a separate landing page for each form, and we plan to use AI to attract new users with separate campaigns dedicated to different PDF forms.
We are also testing other AI tools, like broad match, to discover new and better user intent, reaching an even wider audience. We are excited about further exploring AI potential to understand how our audience works, what drives their interest, and influences their decisions.
