“If your business is not on the internet, your business will soon be out of business” – Bill Gates
This article encapsulates the fundamentals of web analytics that thrives at the intersection of the information economy and analysis tools. The walkthrough is facilitated by my involvement in one such web analytics campaign. I have tried giving a brief overview of the process and for the sake of the length of the blog, walkthrough for setting up Google Analytics for your website can be covered in a separate blog post.
With the customer at the center of the marketing concept, it’s the marketer’s job to anticipate where the customer will appear and what they need to read. Analytics is integral to analyzing and interpreting data to make business decisions about what you should be doing online to understand and connect with your customers.
Modern web analytics tools like Google Analytics, Adobe Analytics, Oracle Analytics, are enabling the marketer to analyze, understand and act on the data that a consumer/business generates while browsing for products or services.
I have been really fortunate to have signed up for a course that helped me apply simple concepts that one learns through ECON and MARKETING 101; facilitated by the ever powerful, and the growing institution of analytics.
Google Online Challenge and Analytics at Robert H Smith School of Business is designed to cover all bases that are required to engage and use Google’s web analytics suit comprising of Google Analytics, Google Ads and Google Tag Manager. Through this course, the student gets to venture out into the real world to find a client by pitching this concept that has the potential to create great value for the client involved. The campaign unfolds in three stages: (1) Finding a client (2) Feeling the pulse of the clients’ market and consumers, identifying KPI and setting goals for the campaign (3) Running a 21-day campaign with a budget of $250 to achieve goals defined in step (2)
After much deliberation and selecting from a pool of multiple clients we finalized on a child development center based in Mumbai, India – Ummeed. They have an online presence and their motive is to create awareness around child developmental disabilities. They accept registrations for their services through their website – https://ummeed.org
After interacting with the team at Ummeed we identified three main verticals that would help augment their online rationale through the campaign – Employment (Professional, Volunteer, Interns), Therapy/Training session sign-ups and Awareness around child developmental disabilities.
As a part of the next step, we had to engage with the developer of the website to embed the Google Analytics code snippet within their website code. Detailed information about how-to’s for this step can be found here – https://support.google.com/analytics/answer/1008015?hl=en
Once the code was embedded we had access to the performance of the website. We could map the demographics of the audience interacting with the website, the source of traffic, along with the performance specific to different pages of the website. Through this, we could tabulate the current performance of the website and align the numbers with the defined KPI to set goals for our campaign.
We used the simple concept of a purchase funnel to track how users interact with the website and measures that would help them percolate down to high-value conversion on the website which was valuable to our client.
Since our client was an NGO we couldn’t attach a dollar value to the conversions on the website, however, we tracked every conversion by deploying tags on PDF download links, Subscription links and Video play button along with scroll tracking using Google Tag Manager. More information about this can be found here – https://support.google.com/tagmanager/answer/6102821?hl=en
Onwards to the heart of the campaign, our tryst with Google Ads dashboard.
WEEK 1 : We began our campaign with 28 Ad-groups and 300+ keywords which we articulated by brainstorming so as to create a long-tailed keyword. The idea was to cater Ad-groups to a specific client. Since Ummeed is an NGO that caters to children with disabilities, many of our keywords were met with a high competition in trying to achieve a decent average ad position. We set our average CPC $0.15 since we were targeting a region were bids hovered around $0.05-0.5. In the first week we got appreciable impressions and clicks, however, the quality score wasn’t coherent with our success at luring potential clients to our ads, which was indicative of the scope of improving further. We critically assessed our ad-groups, ad-copies, and keywords along with our landing pages; By the end of the first week of the 21-week campaign we garnered 4551 impressions, 141 clicks click-through rate of 3.10% with an average CPC of $0.16.
WEEK 2 : As a result of critical assessment of each component of our campaign we identified new high potential areas that could be targeted via our campaign in addition to the fine-tuning that was needed as regards our ad-copies and grouping of keywords. Multiple keywords with a Quality Score of 7+ were indicative of steps taken in the right direction. We also paused ad-groups with search terms that showed poor performance, keeping the Quality Score central to our decision-making process. At this time the traffic of users referred to the website through our ads increased considerably. We could see a considerable increase in client engagement through events that were triggered via the tags we had deployed at potential areas of conversions on the website. At the end of week 2 we recorded 6790 impressions, 600 clicks, click through rate of 8.84% with an average CPC of $0.13.
WEEK 3 : We explored Google Analytics to identify areas that could be tapped into to increase our client engagement. We noticed that a large number of users bounced off after an average of 2.5 pages after landing on our website. We also noticed an anomalous matching of keywords through ‘search terms’ functionality of Google Ads which was causing the wrong user being shown our ads thus increasing the bounce of rate and low conversions. We made note of this and promptly added ‘Negative Keywords’ to our ad-groups to prevent this from happening. We tried to refine our Ad-copies by monitoring the performance of the multiple copies of Ads and retaining the ones with better performance. Our efforts to improve quality score were bearing fruits, we had 50+ keywords with a quality score of 7+. By the end of the 3rd week we had 11300 impressions, 1090 clicks, click through rate of 9.71% with an average CPC of $0.15
Owing to the limited budget, we had to continuously monitor the spending and performance of each ad group and keyword every day (technically night since the client is from India). After 21 figurative sleepless nights, we were successful in accomplishing and surpassing the goals we set at the inception of the campaign.
Given above are the metrics and their respective performance that we achieved. In total, we recorded 105 keywords with a quality score of 6+, of which 37 had a score of 7+, 23 of them with a score of 8+ and 7 with a score of 9+.
For a comprehensive summary of the campaign, I created a Tableau Dashboard wherein I queried data from Google for defined measures and dimensions.
In summary, the campaign was successful in capturing the essence that drives web analytics – “get into the shoes of your consumer and let analytics do the magic”
The ad campaign received 20100 impressions and 1701 clicks resulting in an average CTR of 8.39% and an average CPC of $0.15. We recorded 723 unique events which include PDF downloads, video views, and email link (registration) triggers. The campaign yielded many positive milestones – 2,898 conversions (139% increase) including 1,475 location page-views (73% increase), 21 feedback form engagements (200% increase), and 1,260 new users (58% increase).
That’s about the overview of my Google Online Challenge and Analytics, Until next time…
Be sure to visit my Tableau Public Profile to see many interesting visualizations – https://public.tableau.com/profile/viraj1076#!/
Data Vizardry 