Advertising campaigns were launched November 2017
Reduction of the conversion cost, increase in ROMI, and increase in the number of conversions. Customer attraction from social networks.
During the initial audit of the Google Ads account, typical errors were identified: the confusing structure of the advertising account, a lot of intersections in search queries, the absence of some advertisement extensions, and errors in the configuration of the promotion campaigns.
First of all, we corrected all technical errors in the configuration of the promotion campaigns and updated the structure of the advertising account. We made the structure of the Google Ads account more transparent and logical, in accordance with the structure of the site itself.
An important point was the analysis of the profitability of the promotion campaigns by regions and adjustment of the rates by various parameters: by region, by device, and by impression time. Our decisions were made based on the statistics we obtained from Google Ads and Google Analytics for last 12 months.
We also studied the negative keywords in details: we expanded the number of queries, removed some of the existing queries that cut off the target traffic, and performed “cross-minus” by the groups of advertisements in the campaigns. We redesigned the banners for the dynamic remarketing. We launched promotions.
After launching of the updated campaigns, we analyzed the associated conversions and corrected problems in the configuration that had an effect on decrease in the overall sales volume.
After obtaining the first results, we identified the categories of the leading products and focused just on them. We scaled the most profitable campaigns: we expanded their semantics and increased their budgets.
A month later, we analyzed the promotion campaigns that showed lower results and reallotted the budget in favor of more effective campaigns.
As a result of comparison of the statistics on our campaigns and on previous ones the following results were obtained:
A little later, the use of the automatic rules was started to reduce the daily budgets on weekends and for their return on Monday, as well as the use of the rules for automatic adjustment of the positions of the advertisments in search results was started too.
With the help of split tests of the advertisments, we identified the most conversion elements: calls for action, titles, and texts, and we use them in the context of each group.
We carried out the Google Ads experiments with use of various conversion strategies. The “Maximum conversions” strategy and the classic manual setting of the rate were the best in terms of acquisition value.
As a result, we reached a stable ROMI level of more than 400%. And the average cost of the customer acquisition is 216 UAH.
Comparison of ROAS indices before and after optimization of the promotion campaigns:
In March 2018, with the introduction of the Google Shopping campaigns, this option was immediately enabled for the client’s account. During the first several months, the campaigns did not almost spend the funds and did not bring any results, but starting from July, the volume of sales sharply increased. As a result, the effectiveness of the shopping campaigns was higher than the usual campaigns on search results. It was decided to reallocate the budget in favor of the Google Shopping campaigns:
Since May 2018, we also started working on customer acquisition from the social networks for the online store.
Initially for the baby and children’s clothing store the target audience was determined, and the content strategy and the content plans were developed, taking into account the specifics of each of the social networks (Instagram, Facebook). Thanks to the well-developed strategy, a regular posting of the publications was organized using the Later service, which allows to plan publishing of the posts, and the single style was chosen for the Instagram feed.
To make it easier for users to go to the website from Instagram, we enabled the service that allows to make multiple links – linkin.bio. We also started actively to use Instagram Stories with built-in links leading to the site that were not previously published in the account.
Regular product drawings were held to increase the involvement in the accounts and the audience loyalty. The most popular among the audience were drawings of sets of bed clothes – such competitive posts obtained more than 4,000 comments during the period of 5-6 days, during which the drawing was held.
Opinion leaders whose main blog topic is motherhood were selected for cooperation. We also paid attention to the activity on the bloggers ‘ pages — the involvement coefficient (the ratio of the number of likes to the number of followers of the page) should not be less than 4-5%.
Among them such famous bloggers are as Tanya Prentkovich (289K followers), Nastya Stadnik (305K followers) and Mila Baraeva (158K followers). They made reviews on the products we sent, what resulted in increase of coverage in the account and in increase of the number of followers on the page in Instagram. For example, on the day when Nastya Stadnik published the stories about Berni, the activity in the profile increased more than three times. The number of profile views increased to 6,200, while the average daily profile traffic is 1500-2000 thousand persons.
We also started publishing screenshots of reviews and photos made by the customers, what increased the level of confidence in the store.
Later, we added the feature that makes it possible to publish Shoppable posts in the Instagram account. This again increased the conversion rate from Instagram, since the user no longer had to go to the additional linkin.bio page, and he or she went directly to the page of the selected product:
Since it was decided to consider the number of orders on the website but not the level of the user’s involvement in social networks as the main task when maintaining the account, so we also evaluate the results by the number of transactions on the site from social networks. The growth of transactions from social networks after starting of the work is shown below:
As a result of our work, the number of sales from paid search and social networks continues to grow, and we continue to optimize the performance of the accounts and the results.
UPD: April 2020
More than one year has passed since the case was written. A lot of things changed, and we decided to make a small update.
During this period, a huge number of tests were made in advertising. Dozens of hypotheses have been put forward. We used all available Google Ads and Facebook Ads tools to scale the results. Let’s study some of them:
It turned out that different strategies work for different campaigns. Depending on the results obtained as a result of the tests, the strategies were distributed among the campaigns with selection of the best one in terms of its performance for each of them.
As a result, for the shopping campaigns it was succeeded to increase ROAS by almost 20%, the number of clicks by 44% and in addition to decrease the expenses by 7.4%. The number of sales and the return increased by 15%.
We also tested combined audiences by days and number of pages viewed, what allowed us to identify an active audience with an acceptable cost per each conversion for some certain categories of the website.
How to view conversion statistics from the first visit of the website in Google Analytics:
Here are some examples of the audiences:
15000Increase in Instagram followers during the period of 4 months
12000Increase in Facebook followers during the period of 4 months
7,5%Increase in conversions from social networks
27%Growth of transaction coefficient