Data Analysis

The traffic sources of online store are roughly distributed as 30% Organic and 30% Adwords and 40% Other Sources. At first, it was necessary to understand the online consumer behaviors. Because of that, we have started with implementing analytics enhanced e-commerce integrations. In this process, we took the necessary precautions for minimizing retention, and we have applied funnel optimization steps. Using the onsite search query, we mapped the customer interest on various products and reshaped our in-site navigation along with AdWords campaigns.
We have also realized (via Analytics account) that more than 70% of the customers were using at least two-channels to get to the purchasing level.
Because of that, we knew that multiple traffic sources contributed to the successful purchase. We explained our customer about the benefits of switching to a non-last click model and started to use data-driven attribution model. In this method, we managed to measure the effectiveness of different traffic sources to the contribution of conversion.

One of our biggest problems on was that we could not get satisfactory results from shopping ads. More specifically, the XML feed form was not optimized for Google Merchant Center. Another problem was the low ROI levels in shopping ads. In other words, shopping ads were not used effectively and properly by our client when we first took over the online store. We had worked in coordination with our client’s software team to optimize XML feeds as a first step to tackle this issue. Starting from the basics, we have improved the necessary information fields such as description and title of our Merchant Center products. After that, we have adjusted our XML feed functionality to allow displaying only certain products which have enough stocks, so that we could prevent unavailable or low-stock products from appearing in the XML Feed.
We have continuously discussed the importance of omnichannel marketing and multi-device strategy on the retail side with the team. During this period, Google Turkey team helped us with global researches during E-commerce Acceleration program. We did AdWords Search Conversion Attribution analysis for Cross-Device Attribution.
We have also reached a conclusion that; the number of conversions which are finalized on desktop devices. However, user-interaction happened in mobile devices were more than double of the conversions which are finalized on mobile devices but user-interaction happened on a desktop.
Looking at this data, we have improved our mobile and our bid adjustment strategies. We have also started to use AdWords smart bidding and portfolio bidding strategies like target ROAS. Thanks to bidding strategies and the analysis of “cross-device interaction” by AdWords, we did full funnel conversion optimization (with machine learning), and our ROAS figures have been improved by up to %100.

What have we done?

Penti customer service has created a scoring system for our keywords based on incoming lead quality. We have fed this information to our weekly optimizations and improved our “cost per transaction” values, as well as our quality score.
We are also a firm believer of the importance of cross-channel synergy in the optimization of social media channels like Facebook and Instagram. We used the search intense powers of our clients from AdWords search and shopping ads campaigns to improve the performance of our Facebook and Instagram ads. In this respect, our non-Google campaigns have improved by up to %300.

We have linked Search Console and AdWords, in addition to our optimizations in XML feed for words that have received a minimum one impression in organic but could not catch a good organic position. Using the rules feature in Google Merchant Center, we highlighted our product and product groups with high ROI and high profitability. We tracked the clicks, organic clicks and combined clicks reports using the combined ad and organic stats report in AdWords. In this regard, we have made a great contribution to the profitability targets of Penti brand.

Using Shopping ads reports, we have improved the order of product listing on the website. We also did on-page optimization and landing optimization. Thus, our conversion rate ratios have increased by up to 70% in all our traffic sources.
Using Google Merchant Center rules, we’ve made sure that the campaigns and products that need prioritizing, have come to the forefront, in shopping ads.
We have created different audience sets for more effective use of Audience targeting and deep audience targeting. We created different audiences with Analytics Audience targeting features, such as browsing product listing pages, browsing listing pages, using the search function, listing products at cheapest prices, especially those overlooking campaign products, purchasing, shopping abandoners.
We’ve added these audiences to our AdWords campaigns. We have not used any bid adjustment feature, and thanks to our simultaneous bidding strategy and machine learning; the ladder mass has been optimized with higher or lower bids, by considering the quality of the Adwords audience. As a result, our shopping ads improved by up to 350%, and the online penetration of the brand increased.
Using the Auction insight report, we have improved our competitive analysis. We found answers to questions such as “which categories have impression share loss” and “which areas we should invest more” by using the insights we have. For our prioritized categories, we set high importance campaigns on shopping ads, and we pushed certain product groups to maximize revenue within our budget.


We have started to use event-based marketing tools and WebPush technologies very heavily on We have created new audiences on analytics to collect separate lists of consumers who are sensitive to inbound marketing strategies whom we have brought by e-mail marketing.
We used the predictive ad audience feature for various event-based marketing companies on the market. With this feature, we have created different remarketing lists for end-users with a high propensity to buy ratio.
We are, Boosmart, a company that develops technology on platforms like Analytics, Search Console, YouTube, Facebook, Instagram. There are hundreds of pages on our site that we use for different campaign groups such as long tail, generic, brand + generic, shopping, GDN. However, these pages occasionally caused a “404 – not found” error as a result of the depletion of the product. To solve that situation, we used AdWords API technologies to get a list of pages where we send traffic with AdWords every day. Using the Microsoft Azure server infrastructure, we sent requests to this page every day and checked the http status. In this way, we made sure that if any page had a 404 error, we were immediately informed, and we improved the performance by editing the destination URLs.
Using Analytics “add-ons” in Google Sheets, we have automatically extracted product detail pageview figures and product revenue figures for product page URLs. We regularly followed the “Product Revenue / Product Pageviews” values and automatically tracked the list of productive products through a Google sheet. We have also used this list to optimize the single products and product groups that needs prioritizing in Search, GDN and shopping ads. On the Analytics side, we’ve done integrations such as user ID, enhanced e-commerce, and improved funnel measurement.
Using data analysis tools and machine learning algorithms; we analyzed the search terms to achieve higher conversion rates. At this point, we have identified which words such as “buy”, “comment” or “compare” are more productive for us.
We continuously compared the list of words triggered by shopping ads but not included in search ads or the search queries of our search ads. Besides, we also compared the organic search terms data obtained through Google Search Console with API solutions and we have designed automated processes for new word discovery. We used these search terms on negative keyword, search keyword, product listing ads side and product priority side.
As a result of all these improvements, we have increased the impression figures of our product listing ads campaigns by 350%. In addition, the achieved transaction figures increased up to %250. Thanks to growing the brand on the side of shopping ads; the client wants to increase the budget we spend with AdWords, along with our offer. Hence, we have also achieved budget increases up to %100 on our budget for generic campaigns that we normally use for discovery purposes. We have also done a full funnel optimization by displaying more commercials while they are still in the discovery stage. In May 2018, we spent % 350 higher AdWords budget compared to YoY (year-on-year) benchmark and provided approximately 150% ROI improvement.


“” is a leading retailer in Turkey for socks, underwear and beachwear categories. They serve to more than 2 million users/monthly with more than 400 physical stores in addition to their online stores.