TurkNet Multiple Location Targeting

TurkNet is one of Turkey’s leading internet service providers in the telecommunication sector, making all sales and promotions through call center and digital channels. It provides its customers with a “freedom internet experience” with unlimited, non-committal and high-speed internet advantages.

By promoting the fiber product, which provides faster internet service than Turkey’s average, we aimed to reach users in locations that can benefit from this product.

OBJECTIVES:

The people we need to reach in the campaigns are the people in locations with fiber internet service and internet infrastructure, whose sales we want to increase. We planned to act with the right location targeting in order to reach the right people in search campaigns. In this way, it was aimed to spend the budget on users who will convert and to increase the conversion rate of the campaigns. We expected an average monthly improvement of 2 times in performance metrics, the details are as follows;

• Increasing the Conversion Rate
• Increasing the transaction x2
• Keeping ROAS at the same level

STRATEGY & IMPLEMENTATION

In order to target only users with “Fiber internet” infrastructure at their location, we obtained detailed data on internet infrastructure neighbourhoods/districts/provinces of all of Turkey from our brand. From the data covering all internet infrastructures, we separated only the data of those who have the fiber internet infrastructure of 100 Mbps and above that we need.

Since targeting around 2000 locations within a 3-mile radius cannot be achieved manually, we imported and imported this data. Since the number of locations where a campaign will be targeted is limited to 500, we have revised the data to divide it into provinces. We rearranged the Search campaigns to target only these locations.

By rearranging the keywords and ad texts of the campaigns, we increased the ad rank rate to a higher level and targeted users in the right location.

Since Impression share is our target, we proceeded with the strategy of focusing on keywords. For this reason, we have determined only 15 keywords among the search terms that users search for their internet needs and have a high potential to turn into sales, and we have grouped them as follows. Each keyword phrase and exact match were added to the name groups.

CHALLENGES:

• Thousands of Location Data
  We requested infrastructure data from our customer with a breakdown of provinces, districts and neighbourhoods. Location data with approximately 49K neighbourhoods came from the customer. From these locations, 4325 locations with fiber infrastructure (with a speed above 100) were separated. Because 4325 locations were too many to target, we prioritized neighbourhoods based on possible conversion rates.
• Import Format Docs
  Google ads do not allow bulk uploading of this volume of data. Since the format in Google maps is not standard, we have observed in our research that there is no tool that we can bulk upload.
  As a result of dozens of unsuccessful attempts, we have brought our data into a format that we can use in Google campaigns and update when needed, using Google excel macros.
• Campaign Limitations – Radius Limit
  Unlike our plan, Fiber 100 locations would be targeted under a single campaign, but we ran into google limitations. We had to categorize neighbourhoods according to provinces, as radius targeting (close location targeting) does not allow targeting more than 500 in a campaign. At this stage, 2300 neighbourhood locations were activated by dividing them into 5 campaigns according to the provinces.
• Target Audience
  The audience we want to reach; primarily consists of people residing in regions with fiber internet infrastructure. As can be seen in the table above, there are 4,325 neighbourhoods with this infrastructure. Therefore, our target audience is users living in fiber regions and looking for fiber internet.
  The user who searched the relevant keywords in the search campaigns of the project and needed the internet was our target audience. In order to target the most relevant users, we analyzed the old campaign data and created our keywords from the words we want to have the best performance.

RESULT:

We have reached users who can actually receive this service, by launching the search campaigns we use to reach users who search for the words “fiber internet” only to users who have access to fiber internet infrastructure. Thus, although we narrowed down the locations, we increased the number of our subscribers 10 times.

• The user is increased by 3.3x!
• The conversion rate increased by 2.7x!
• The ROAS increased by 4x.
• The number of transactions has increased by 10x!

ROAS:

Ecommerce Conversion Rate:

Users:

Transactions: