Web Analytics Driven Online Marketing (Part 1 of 2)

Raja Bhat, VP – Consulting Services, Nabler Web Solutions
With rising inflation resulting in an economic slowdown, the first thing in an organization that gets cut is the marketing budget. In the coming months we will be seeing more and more organizations reducing their marketing budgets.

Since marketing budgets are directly linked to the sales revenue, a fall in sales will result in a reduction in the marketing spend. This, across the board, cut of marketing budgets happen due to the lack of measurability of marketing programs. As the famous saying goes, “I know half my advertising budget goes waste but I don’t know which half”.

It is becoming increasingly difficult on the internet to reach the right audience, acquire the right visitors, convert them into your customers and retain those customers. The biggest challenge for us today is to determine which interactive channels to prioritize, how much to spend and how to measure the results.

Why invest in web analytics?

In times like these, when greater measurability and accountability is needed to measure the performance of the web as a marketing channel “Web Analytics” plays a crucial role in providing that much needed intelligence.

Online marketing is much more accountable than traditional media in terms of ROI calculation, but surprisingly, few organizations in our country actually use such web intelligence to optimize their online advertising and marketing spend.

Web analytics has evolved from the days when it was just used to measure hits and page views. Today, it forms the integrating platform around which all online marketing initiatives are deployed and optimized for better visibility into the performance of each and every online initiative.

How can web analytics improve your marketing ventures?

To measure and improve the performance of the web as a marketing channel, the following three essential requirements have to be taken into consideration:

1. Have clearly defined website objectives and create key metrics (KPIs): Clearly defined website objectives and KPIs (Key Performance Indicators) can help you measure the performance of the website against the defined objectives. The goal is to improve the KPIs over a defined period of time.

2. Consider both the website and the website promotion together: The website and website promotion cannot be measured and optimized in isolation. This is because the visitor activity on the website must be tied to its source to understand how each media segment is performing.

3. Create goal pages and assign weights and value: All visits to your website are not equal and only about 1-5% of the visitors register or buy from your website. So, what about the rest of the visitors (95-99%)? Surely most of them came to the website with a purpose and may or may not have accomplished those. Goal pages are key pages in the website and visits to which reflect the level of engagement with the website. Goal pages could be important pages such as product, resources or about us pages. Visits to these pages show a definite interest and intent of visitors to your offerings.

Once we have these three things accomplished we can measure and optimize the website and promotions.

What can web analytics data be used for?

The insights from web analytics data can be used to optimize the following:

- Marketing campaigns
- Conversion funnels
- Segmentations
- Landing pages and other important pages

Optimizing marketing campaigns

1. Optimizing display advertising campaigns: A typical brand campaign is based on a plan, impressions purchased across media vehicles and combinations of various creatives and sizes are placed. The campaign is monitored and the success or failure is measured through commonly used metrics such as CPC (cost per click), CPM (cost per thousand impressions), CTR (click through rate) and CPL (cost per lead generated). These metrics, on their own, are insufficient to measure the true value of display ad campaigns. With hundreds of placements with varying combinations of website/property, creatives, ad size and landing pages, how do we know which combination works the best in achieving your campaign objectives?

The metric to be used to measure brand campaign is CPEA (cost per end action). A typical brand campaign is aimed at moving a visitor through the buying process of Awareness/Engagement – Brand consideration – Purchase intent. The website content can be categorized into the above three stages. The goal page values and weights can be assigned to the relevant content pages. The sum of the goal pages viewed is the end actions completed by the visitor. The end actions measure the “interaction levels” of a visitor within the website.

By using CPEA as a metric, you can achieve the following:

- Perform overall campaign efficiency analysis
- Identify the best performing combination of site/property, creative, ad size and landing pages in achieving the campaign objectives
- Perform creative analysis and identify the best performing creative
- Measure the impact of view throughs (impact of non-click traffic on overall campaign efficiency)
- Conduct post impression analysis of brand campaigns

2. Optimizing email marketing: In an earlier article, Vikas Tandon of Indigo Consulting had written exhaustively about the email marketing opportunity.

Email remains as one of the most effective marketing channels. In spite of low email open and response rates, the low cost per contact makes this channel so effective. However, bulk emailing with no consideration to “relevancy” has negative impact and puts the company brand at risk and all future mails are “instantaneously deleted”. (All of us delete many mails everyday from our inbox, without even opening the emails, once we brand the source as “no value”.)

How can web analytics help optimizing email marketing?

Traditional marketing segmentation divides customers into different groups based on demographics, purchase history etc. Web analytics provides a third component – “Behavioral data” of the visitors to the website influenced by the email. This data gives insights to create ‘relevant’ emails.

Consider the following examples:

Action – Visitor downloads a whitepaper or other documents
Reaction– Send informational mails to the visitor that will educate and provide opportunities to learn more

Action – Visitor visits the product pages and exits from the website
Reaction – Send mails with product comparison specs, warranty, trial offer etc.

Action – Visitor puts an item in the shopping cart but abandons the purchase process
Reaction – Send an email two days after the abandonment that includes incentives to purchase the abandoned product and other high margin products

Let us say a promotional HTML email was sent to 100 people, out of which 35 visited the website and eventually two bought the product/service. In the absence of behavioral data, we are blind to the 33 people who showed some interest in the product but due to some reason did not choose to continue. I am sure you will agree with me that we have a far better chance of converting these 33 people into customers, if we know the reason for abandonment and communicate relevant information rather than causing fatigue through a general email blast again.

Even if there is a 10% conversion from among the 33 people, you will have an overall conversion rate of 5.3%. In other words, an increase of 165%.

3. Optimizing SEO and PPC marketing: Both the earlier articles, Integrating SEO (search engine optimization) and PPC (pay per click) by Vinod Nambiar, Position 2 and Of Long Tails and Paid Search Campaigns by Nimish Vohra, Regalix, brilliantly cover the finer points of search marketing.

I am going to focus on how web analytics data can be leveraged to optimize SEM. We all know that top 20% of the keywords account for 80% of the traffic. Keywords/phrases can be classified into 1) branded and 2) category/generic keywords. By using web analytics tools we can look at the keywords that brought visitors to the website. You will find a small head of branded keywords and a very long tail of category/generic keywords contributing a small amount of traffic to each.

By using the intelligence gained from web analytics data, we can reduce the SEM spend by optimizing the website through SEO for branded keywords and PPC advertising for the low bid value category keywords.

Paid search campaigns need not necessarily be a direct response channel, it can be effectively used for brand campaigns too. Since the conversion rates are so low, we need to understand the “interaction levels” of visitors who came through various keywords. By assigning goal page values and weights (mentioned earlier), we can effectively measure and track the value of individual keywords/phrases in bringing the most “engaging audience”.

Most of this “engaged audience” may not buy from the website in their first visit, they may however bookmark the site and visit again to purchase/register (latent conversions). By using web analytics data we can gauge the effectiveness of a PPC campaign by tracking the conversions at a later date which can be attributed to the source keyword/phrase. [To be continued…]