Google Ads Automation: 2.4x Cost Reduction, 8.5x Increase in Conversion Rate

Client

MС “Dobrobut” is the largest network of medical clinics in Kyiv with an integrated approach to patient treatment: from laboratory diagnostics to surgical intervention.

Medical Network “Dobrobut” has been operating since 2001, offering treatment in more than 35 highly specialized areas in 13 units located in 9 districts of Kyiv. 2,000 highly specialized employees 24/7 help people of all ages.

Work Period

Active work was carried out from 01.01.2019 to 31.05.2020.

Our Task

Increase the effectiveness of advertising campaigns in Google, increase the number of records in the medical center, inform the target audience (TA) about innovations in the clinics’ work.

Preamble

Since October 2015, Google advertising campaigns for the client have been being run by one of the primary advertising agencies.

During the initial audit of the advertising account, our experts found several shortcomings:

• the semantic core included all queries that related to both diseases and treatment (including very broad informational keywords);

• advertising campaigns (AC) did not have a well-thought-out structure that meets the client’s objectives;

• advertising campaign settings were made without taking into account the specifics of the client’s target audience;

• as landing pages for advertisements, the pages of the site’s medical library were used, which were not prepared for user conversions;

• accurate tracking of goals (user actions on the site) was not configured.

A common mistake when launching advertising campaigns is the lack of structure: combining informational, commercial and branded keywords into one advertising campaign. This distorts the statistics and excludes the possibility of more subtle and effective cost management.

Setting goals in Analytics was done at a superficial level. Imported conversions to an ad account had no categories, which significantly reduced the efficiency of the algorithm when automatically setting bids.

All campaigns used only automatic bidding strategies.

Client Expectations

For the client, advertising on Google is one of the crucial communication channels with potential patients. Below there are his expectations from campaigning:

• provide all clinics with a stable number of visits;

• optimize the cost of advertising campaigns;

• inform about new methods, service packages and directions of work of medical centers;

• implement the ability to redirect potential clients from one clinic to another if necessary.

Advertising Campaign in Google Search

At the first stages of setup, our specialists performed standard work based on the “two pillars”:

• implemented the account structure;

• fully configured site analytics.

Please note: full and correct setup of site analytics is the basis of any advertising campaign!

Ad Account Structure

Brand advertising campaigns

Over the years, the client’s medical network created a strong brand. Therefore, working with branded traffic must be mandatory and scrupulous, especially in the following:

• we separated Russian-language and Ukrainian-language requests into separate campaigns;

• divided requests by locations and doctors.

Campaigns by destination

At the setup stage, the client identified 7 priority areas:

• Oncology;

• Pediatrics;

• Surgery;

• Cardiology;

• Gynecology;

• Emergency;

• Stationary department.

Later on, our specialists added new categories and subcategories of clinics — and in May 2020, we worked out 18 areas of work for the medical centre.

Working with negative keywords

When collecting semantics, we unloaded search queries for existing advertising campaigns. Broad keywords in the account, such as “vaccination”, collected a large number of search queries, thus obtaining valuable statistics on negative keywords and conversion requests.

Consider Google recommendations

We prepared the updated advertising campaigns for launch, taking into account the main recommendations of Google:

• AC was supplemented with all relevant advertising extensions (clarifications, additional links, structured descriptions, addresses, phone numbers, promotions and links to the application);

• three expanded advertisements were added to each ad group, as well as supplemented with adaptive search ads;

• we adjusted the rotation of advertisements;

• we detailed the geography of the display — not just the city of Kyiv and the Kyiv region, but added districts of the capital and, if necessary, set geotargeting within the radius of the clinic, which allowed us to collect more accurate statistics on users and fine-tune rates depending on location.

Setting up site analytics

Our specialists paid particular attention to setting up macro and micro goals for the site in Google Analytics. Due to the peculiarities of the site, each of the 10 types of submitted forms had a unique identifier, which complicates the setup process.

To solve this problem, we created stationary identifiers (elements) of js-code for events based on pop-up windows “We will call you back soon. Have a nice day!”, with the help of which we managed to debug an apparent fixation of submitted forms.

Having set up goals in Google Analytics, our company’s employees determined the main performance indicators. As with most Services business categories, site form submissions and calls are the goals.

We imported these goals into the ad account based on their categories (lead, contact, signup, page view, etc.) and chose either Position Based or Data-Based attribution model. Choosing an attribution model other than the Last click captures a complete picture of the conversion to account for Google algorithms.

Due to the incorrect fixing of goals in Analytics earlier (a click on the form call button was defined as a macro conversion of the submitted form), it was impossible to rely on the statistics collected by the previous contractor.

Google Display Campaign

Given the nature of medical advertising, running remarketing is against Google’s advertising policies because it crosses the user’s personal boundaries.

We implemented the launch of advertising in the Display Network with the help of an interested audience: we created intent audiences. We launched advertising banners (taking into account all legal requirements — disclaimers on each banner), as well as adaptive display ads.

When creating intent audiences, it’s a good idea to include high-frequency keywords in the direction so that there is no double meaning in their meaning. To specify the audience, we recommend using competitors’ domains together with CS.

From December 2019 to March 2020, in the areas of Gynecology, Oncology, and Pediatrics, we tested the launch of a smart advertising campaign in the Display Network.

Optimization Work

For effective scaling and development of the project, it is necessary to use the artificial intelligence (AI) of Google in the Republic of Kazakhstan. For its use to give a tangible positive effect, it is necessary:

precise implementation of all settings (both from the side of the advertising campaign and the side of conversions);

a clear definition of realistic goals (step-by-step “guiding” the user through the sales funnel, where each stage has its own goal);

a smooth transition from the historically possible to the desired goals of the Republic of Kazakhstan.

The launch of the updated ACs took place in the manual bidding mode and with a phased transition to automatic ones, depending on the collected statistics for each AC and the goals that we set for it (maximum conversions, a specific cost of conversions, or the percentage of received impressions).

Project Niche Features

According to Google Trends, there is a seasonal decrease in demand for clinic services in late December and early January.

The pandemic in March 2020 also harmed the efficiency of most businesses.

Results

As a result of working with an advertising account, our team managed to achieve impressive results*, which are discussed below.

1. The overall conversion rate shows a stable positive trend, and we increased it by 381.66% (from 2.05% to 17.52%):

2. The total number of conversions shows positive dynamics as the scaling and optimization of the AC: (Except for seasonal downturns in December-January and the situation with the pandemic in March-May 2020).

3. At the same time, after switching to automatic bidding strategies and completing the main training period, we reduced the cost of conversion by 2.41 times (by 58.56%):

(Our PPC specialists used labels 1 and 2 to indicate the stages of transition to auto strategies).

4. Due to meticulous work with keywords, negative keywords and bid optimization, we increased the percentage of received impressions in the search network from 35.7% to 88.83%:

5. We also recorded an increase in conversions for form submissions:

• the “Call Me Back” form submission:

• the “Clinics” form submission:

• the “Doctor” form submission:

• “Call” (due to technical reasons on the part of the call tracking service, correct target fixation became possible only from July 2019):

*Comparative analysis is presented for the same period of the previous year.

Project Participants:

Project Manager: Juliya Myrna

PPC-specialist: Iryna Khomenko

PPC-specialist: Oleksandr Shylin

PPC-specialist: Iryna Kyriy

PPC-specialist: Andriy Kolomiets

Head of PPC department: Halyna Lyman.