Tıkla Gelsin & SEM – Beyond the 5 iPhone 16s Draw with TV2D Data-Driven Strategies

In our iPhone 16 sweepstakes campaign, we leveraged the power of digital channels, data-driven insights, and first-party user data to elevate spending thresholds and maximize campaign participation. Through advanced technology, AI-powered solutions, and data-driven insights, we not only boosted sales but also enhanced brand engagement. Throughout the campaign, we delivered personalized messages at the right time to the right users.

Live Period – Digital  

By conducting an in-depth analysis of past purchasing data and inflation rates, we identified users placing orders below the target spending threshold and assessed them with a unique perspective. We indexed users’ historical spending habits to current inflation levels, calculating that the 250 TL sweepstakes eligibility threshold equated to 124 TL in 2023 and 80 TL in 2022.

However, we didn’t stop at inflation data alone! We incorporated proprietary metrics such as purchase frequency, total spending amount, order distribution above and below 250 TL, visit frequency, lifetime active days, and order count. Using AI, we scored users based on their likelihood to place orders above or below 250 TL. Rather than focusing solely on purchase-driven segments, we implemented a multi-layered targeting strategy. By utilizing BigQuery’s advanced Machine Learning capabilities and our in-house sophisticated data science algorithms, we classified users based on retention, churn, loyalty, and activity within the “Pickup” and “Home Delivery” services. We built 10 distinct audience segments and their lookalikes, aligned with different advertising objectives.

With this detailed segmentation approach, we identified churned, active, loyal, and high-retention potential users by channel and order threshold interest. This enabled us to develop precise and highly targeted engagement strategies. The core dynamic of our campaign was integrating data and technology to reach the right user, at the right time, with the most effective messaging. As a result, we implemented a powerful, data-driven marketing strategy that increased conversion rates and strengthened user engagement for both Pickup and Home Delivery services.

Following the launch of our digital advertising campaigns, we aimed to enhance campaign awareness in real-time without additional spending. Unexpected technical constraints prevented us from announcing the campaign and sweepstakes within the app, meaning users only became aware of their participation via SMS notifications triggered after placing orders above 250 TL. This challenge led us to explore creative ways to inform users about the campaign while they were still in their shopping carts, ultimately encouraging higher basket values.

Our primary objective was clear: to inform users in real-time while they were actively considering a purchase and effectively communicate the benefits of the campaign. However, achieving this without additional costs required an intelligent solution. This is where Firebase’s In-App Messaging (IAM) feature came into play. This tool enabled us to target specific user segments and track conversions effectively. Ensuring the sustainability and long-term efficiency of our strategy required defining optimal cart value thresholds. Through meticulous data analysis, we identified that basket values at 250 TL, 500 TL, 750 TL, and 1000 TL had the strongest impact on user behavior. Since orders above 1000 TL were rare and users earned an additional sweepstakes entry for every 250 TL spent, we focused on these specific threshold values.

To better understand users’ shopping tendencies, we examined cart signals in detail. By analyzing past user data, sub-brand pricing, and menu structures, we determined the most critical ranges:

• 210–249 TL for the 250 TL threshold
• 415–499 TL for the 500 TL threshold
• 650–749 TL for the 750 TL threshold
• 850–999 TL for the 1000 TL threshold

We dynamically defined these thresholds in-app using GA4/Firebase’s Create Events feature. Utilizing our existing integrations, we created four new events and leveraged Firebase IAM to target users in real-time, maximizing engagement. Instead of a one-size-fits-all messaging approach, we developed a personalized communication strategy. Users with cart values below 250 TL received informative messages about the sweepstakes benefits, while those at 500 TL, 750 TL, and 1000 TL thresholds were shown creatives emphasizing their increasing number of entries.

To determine the most effective messaging format, we conducted A/B tests comparing two communication strategies: explicitly displaying the number of sweepstakes entries versus using a “+1” format to emphasize incremental rewards. The results were clear—the “+1” format significantly outperformed in both user engagement and conversions. Consequently, this became the foundation of our messaging strategy across all campaign communications.

Through in-app messaging, real-time targeting, and smart segmentation, we not only announced the campaign but also transformed the shopping experience into a powerful user engagement strategy. Our data-driven optimizations increased basket values and ensured that users participated in the campaign more consciously and enthusiastically. This innovative approach not only secured the success of our current campaign but also laid a strong foundation for future strategies. With a data-centric, user-focused, and dynamic approach, we pioneered a new era in conversion optimization.

Live Period – Media 

In our campaign, we went beyond traditional media by leveraging innovative channels that enhanced brand awareness. In addition to performance campaigns, we developed strategies prioritizing awareness and engagement. Furthermore, we created AI-driven content tailored to our target audience’s interests, aiming to establish a deeper and more meaningful connection with them.

As one of the first brands in Turkey to utilize digital screens on the back of motorcycles, we designed data-driven routes based on historical data, traffic patterns, day, time, and location analyses. Through dedicated couriers, we maximized visibility by strategically pausing motorcycles in front of competitor stores and locations aligned with our brand.

Moving beyond conventional audio advertising, we delivered AI-powered, personalized jingles tailored to users’ music preferences. Whether they listened to arabesque, jazz, or electronic music, we ensured that our message resonated emotionally through customized versions within their preferred genres.

In digital out-of-home advertising, we implemented time- and location-based targeting, ensuring our ads aired during peak traffic hours. Additionally, we reinforced engagement and penetration by retargeting users via push notifications, capturing those we might have missed on screens.

We rotated the “Click & Collect” and “Eat & Earn” creatives in our broadcasts, ensuring our messages reached consumers in different contexts. This approach kept user interest consistently high and provided multiple touchpoints for delivering varied messaging.

Through these innovative media strategies, we strengthened our campaign’s media reach with a multi-layered approach. By pushing the boundaries of technology, data, and insights, we executed a cutting-edge campaign communication that redefined media engagement.

Post Period – Television  

After the campaign period, our primary goal was to sustain and even enhance the high performance and visibility we achieved through various data, insights, technology, channels, and media. While doing so, we aimed to show consumers that we are always by their side, build deeper connections with them, and maximize brand awareness and performance by engaging them at every stage of the classic e-commerce funnel.

Since we were already reaching our users online, on the streets, in stores, and even at competitor restaurants, we decided to expand our “continuation and reminder” strategy to television screens and households—places where consumers make the clearest decisions, feel secure, and where we could reach a larger audience with a single targeting effort. Instead of covering all cities or just the three largest ones, we leveraged our own data-driven insights. We analyzed restaurant counts, opening/closing hours, brand awareness and research reports, competitor distribution, online shopping behavior, and revenue distribution across cities. By incorporating TV Kantar data on household reach and targeted TV coverage, we strategically selected 10 cities. Additionally, we determined the most effective broadcast hours—between 7:00 PM and 9:00 PM—based on restaurant operating hours, app order and revenue trends, TV viewing rates, and daily rating analyses.

Instead of using conventional formats like basic on-screen banners or QR codes, we focused on gamification. Our approach reassured users that they hadn’t missed out on our iPhone 16 giveaway and reinforced our commitment to offering continuous opportunities. When users pressed the green button on their remote, they were welcomed with a personalized opening screen tailored to their city.

Rather than showing the same advertisement to everyone, we highlighted 10 iconic structures unique to each city within our creative assets. Users were then prompted to answer three questions, with the chance to win a grand prize if they answered correctly. To maintain engagement, even incorrect answers led to exciting surprises, presented with vibrant visuals and playful messaging.

Recognizing the importance of data-driven insights and personalization in our consumer relationships, we structured our game questions to enhance awareness of our iPhone 16 campaign, increase visibility for both our main and sub-brands, and highlight key app functionalities. Moreover, we tailored the discount coupons awarded based on each user’s in-game performance and responses, leveraging machine learning to determine optimal discount values, brands, and channels for each city. By analyzing raw data, such as average order value, order frequency, and revenue potential, we identified the key metrics to improve in each city and designed AI-driven incentives accordingly.

To maximize reach, we carefully selected TV channels. Our primary choice was TV8, given its gaming-oriented audience, while our secondary selection was ShowTV, based on high ratings and insights from Kantar indicating that households consistently watched “Bahar” and “Kızılcık Şerbeti” at least once a week.

Since we aimed for long-term impact, we planned a two-month execution period. However, knowing that our sales typically drop during the first two weeks of Ramadan, we divided the campaign into two phases. This approach allowed us to analyze performance in the first phase, optimize city-based TV ad distribution for the second phase, and allocate the budget strategically—40% in the first phase and 60% in the second—ensuring data-driven decision-making.

After the first phase, we developed new metrics by analyzing both TV performance and Adjust-based sales data. Based on these insights, we discontinued ad placements in three cities, reduced exposure in two cities, and reallocated impressions to the top-performing five cities. We further optimized budget and placement distribution based on TV channel CTR performance.

To push beyond gamification, we integrated advanced measurement technologies. We embedded impression tracking codes within different campaign layers on Adjust, allowing us to capture and analyze user interactions at multiple touchpoints—including green button engagement, correct and incorrect answers, and QR code scans leading to app visits. This enabled us to track users’ journeys across IPs and cities and assess post-ad purchase behavior.

Did we stop there? Of course not. Although Adjust data provided valuable insights, it did not allow us to create audience segments for future retargeting. However, we wanted to continue engaging users beyond TV. To achieve this, we leveraged the impression data collected from TV campaigns and mapped user interactions to device IDs via IP addresses. This allowed us to identify and retarget users who saw or interacted with our TV ads in real-time.

To further refine our digital strategy, we established automation on Google Cloud, enabling seamless cross-platform integration. This system automatically synchronized audience data with Google, Meta, and TikTok daily, ensuring that our campaign insights translated into optimized digital and programmatic advertising efforts.

Post Period – Digital 

We segmented our campaigns across three different platforms based on distinct target audiences and their characteristics. Beyond standard advertising inventory, we implemented Playable Ads for each platform, coding them in HTML to maintain our game mechanics and engagement outside of TV. These ads featured city-independent versions of the questions asked on TV. To further personalize the experience, we integrated machine learning to dynamically determine offer values based on users’ responses. Instead of using a universal coupon code that could be shared and affect campaign performance, we developed a technological solution within our HTML code to ensure each code was generated uniquely and could not be reused.

Leveraging our data and measurement capabilities, we created various audience segments for precise targeting. Our goal was to encourage users who had seen our TV ads and earned coupons to complete their orders, re-engage those who saw the TV ads but did not interact, and involve those who had neither seen the TV ads nor participated in the main campaign. By incorporating Playable Ads, we extended the gamification experience beyond TV, enabling users to earn coupons digitally and encouraging them to redeem them.

Our audience segmentation included:

• Users who saw the TV ad but did not engage in the game
• Users who played the TV game and won a coupon with the correct answer
• Users who played the TV game and won a coupon despite an incorrect answer
• Users who did not see the TV ad, did not place an order above 250 TRY during the campaign period, and missed the sweepstakes
• Users who played the game via Playable Ads on Google and Meta and won a coupon

We analyzed these segments using platform-specific tools to determine their characteristics and strategically assigned each audience to the most suitable platform based on data-driven insights.

Platform Strategy:

• Google: Users who played the digital game and either answered all questions correctly or won a coupon despite incorrect answers were targeted with Demand Gen campaigns. Meanwhile, users who engaged with our TV ads and won coupons were retargeted through conversion-focused App Engagement campaigns. Playable Ads were exclusively used for App Install campaigns, with an in-app action strategy.
• Meta: Playable Ads were integrated into App Conversion campaigns to re-engage users who saw the TV ad but did not interact. Additionally, we launched a chatbot strategy to strengthen engagement by reminding users of their earned coupons via ads.
• TikTok: Given the platform’s active and gamification-friendly nature, we targeted users who had not seen the TV ad and had not participated in the sweepstakes, directing them to Playable Ads to win coupons.

To maintain a dynamic and automated approach, we updated these audiences daily using TV ad data and real-time coupon usage reports. This automated data pipeline ensured seamless audience segmentation and optimization across platforms.

Results 💫

As a result of the digital campaigns we ran using our audiences during the campaign period:

• On Google, we achieved 57% lower CPA, 160% higher ROAS, and 258% incremental revenue.
• On Meta, we achieved 39% lower CPA, 61% higher ROAS, and 162% incremental revenue.
• On TikTok, we achieved 54% lower CPA, 121% higher ROAS, and 224% incremental revenue.

With our real-time communication strategy, we targeted users at critical cart values; by capturing users approaching +1 entry to the raffle (value-based) in the cart step, using only our measurement/integration power and Firebase IAM tool, and informing them with a popup (without spending any media budget), we increased the average cart value by 38%, generating an additional revenue of +3.2M ₺.

This strategy accounted for 7% of total orders, contributing 16% to the overall revenue. The incremental impact on sales reached 104%, highlighting the importance of integrating real-time access to users through the right platforms and leveraging our integration power.

When comparing the average increase in “threshold carts” to previous periods, we saw an overall 131% incremental impact in both sales and revenue.

Users who saw the banner increased their cart average by 38%, while users who didn’t see the banner had a 12% decrease in their cart average.

Looking at channel segmentation, the order impact for “Gel Al” users showed a 177% increase, and the revenue impact was 166%, while “Sana Gelsin” users showed a 113% increase in orders and a 116% increase in revenue, proving the higher impact on our BO goal for GA sales.

Thanks to this strategy, we secured +13K raffle entries, resulting in an additional 7% of users qualifying for the raffle.

Based on the results of the TV commercials we ran after the campaign, when we took the previous TV campaigns as a benchmark, we achieved an unprecedented ROAS, 8 times higher than the previous campaigns. Additionally, we achieved 1.6 times higher CTR and 60 times higher CR compared to the benchmarks, making us proud. By reaching 1.6M households and a total of 6.5M people, we saved a total of 2.2M TL compared to traditional TV media spending.

With a session time of 17.2 seconds, we reached 2 times the benchmark, and our engagement benchmark was 5%, while we achieved 12.3%, which is 2.5 times higher.

From this TV campaign, we generated 380% incremental revenue by increasing our average cart value by 180%, alongside 5K additional orders. We acquired 8K new users and successfully re-engaged 18K users into our app. While only 3% of the coupons we created were used, the impact on sales and revenue shows how crucial brand awareness is and highlights that what truly matters is personalization, the importance of measurement, and how it’s possible to create wonders from data with the right technologies and ideas.

From this TV2D (digital) campaigns, While achieving an 80% incremental revenue on Android with a 2x ROAS difference on Google, we broke records on iOS with a staggering 13x ROAS difference, delivering an additional 1802% incremental revenue with 85% lower order and visit costs.
On TikTok, we increased conversions by 1.2x, achieved a 3x ROAS difference, and generated 170% incremental revenue with 65% lower CPA.