Avinash Kaushik kicked off the London SES 2012 Conference with an eye opening presentation on how we need to think 
differently when it comes to online marketing today. He barely scratched the surface on several key issues faced by internet marketers today; touching on each issue just enough to open our eyes to how we now need to take a different perspective with our approach to managing online marketing campaigns. He explained to us that we need to take full advantage of the control we have with internet marketing that lacks with traditional forms of advertising. With internet marketing we are in control. The way we influence is not by shouting at people (TV, Radio, Print), but is by having conversations. This is done by driving optimal traffic to where it belongs with data, and opening the doors for conversation with the use of social media.
Know which metrics matter
The marketing sales funnel is like the animal food chain. Each step of the sales funnel could be compared to the next species on the food chain, devouring the one that came before it. Disruptions to any of these steps propagate down the line, all the way to the paid conversion. We need to stop looking at metrics of the past, metrics that Avinash dubbed as “lame.”
| Lame Metrics |
New “Awesome Metrics” |
| Clicks |
Loyalty |
| Page Views |
Time since last action |
| Visits |
Conversion Rate |
| “Touches” (people touched) |
Number of Days and Visits from an action |
| Emails Sent# |
Unaided Brand Recall |
| Number of Press Reports Released |
Economic value |
| Likes, followers, "The latest flair of the month” |
Task Completion Rate |
The new, “Super Awesome” metrics represent how a combination of “lame” metrics interact with each other. Instead of focusing on page views, focus on user loyalty. User loyalty is a combination of page views and other “lame” factors. Every “lame” metric possible needs to be tracked, regardless of how insignificant it may seem. On its own it may look worthless, but combined with other metrics, it could become a gem.
Optimize for Users That Don’t Convert
Web analytics software enables us to look at click paths to identify points in the sales funnel where users are dropping off. By optimizing for these users we are giving attention to those who would not otherwise convert. 
Value needs to be given to micro conversions. Examples of micro conversions are qualified leads, email/blog subscriptions, and catalog requests. Macro conversions are the goal. They are sales which produce revenue. However, total economic value is determined by the combination of the many micro conversions with the few, or the one macro conversion.
Influence & Evolve to Win with Social Media
If a person values the “lame” metrics (listed above), than they look at social media as yet another channel that can drive business. The reality is that social media provides much greater opportunities.
Incorporating social media into a marketing portfolio presents an advertiser with new metrics to gage social participation. Giving economic value to these micro conversions gives an advertiser a clearer view which levers deliver which economic value.
- Conversion Rate – Audience comments Per social Contribution
- Amplification Rate – “Forwards” per social contribution
- Applause Rate – “Positive Clicks” Per Social Contribution
- Social Economic Value – Sum of Short & Long Term Revenue + Cost Savings of Social Contributions
As of recently, Google started displaying social media results along side search results. This means that if a website comes up in a search result, it will be more visible to the searcher if the searcher’s friends have +1’d their site. This is giving us first glance into the future of search; a future which will display results based not only on its relevancy for the search terms used, but also on the social interaction of the result.
Understand More (be less wrong) with “Intelligent” Attribution Modeling
As a company grows, it ads new marketing channels to its portfolio, and advertises more aggressively. The result is an expanded reach which then opens up the doors to new challenges.
Answering the question of who gets credit for a conversion when that conversion came from multiple campaigns over time is a difficult one. The only clear answer Avinash was able to give on the topic is that there is no clear answer. Every business has a unique way that it needs to use to approach multi channel conversions.
Avinash went over the simplified approaches that companies frequently use when dealing with multi-touch conversion analysis.
- First Click Takes All – This strategy gives all of the credit for the conversion to the first campaign that sent the user to the site.
- Last Click Takes All – This strategy gives all of the credit for the conversion to the last campaign that sent the user to the site.
- Equal Credit – This strategy gives equal credit to each campaign that sent the client to the site before he converted
- Time Decay – This strategy gives credit to each campaign, but the least to the touch point that delivered the conversion the furthest back, and the most to the touch point that is the most recent.
Each of these models have flaws. The “time decay” model is the least flawed. Finding the perfect model for a business requires painstaking analysis. This analysis is often wrong, but still less wrong than the alternative.
When producing a model, it is important to consider as many variables as possible. The analysis is only more complete with more data. Avinash showed the following graphic in his presentation, representing just some of them.
To understand how many of your clients come from multi touch conversions use Google Analytics. Google provides us with an easy to understand multi-channel conversion visualizer. It is a powerful tool to visualize the interplay between channels.
As Avinash put it, in order to win with attribution modeling you need to:
- Understand
- Test
- Be Less Wrong
Multi Channel Portfolio Optimization
Avinash raised an interesting question. What would happen to a companies’ PPC search campaigns if all social media efforts were turned off? What would happen to a companies’ display campaigns if all PPC search campaigns were turned off? He was in no way seriously suggesting it, but used it to provoke some thought.
Understanding the dynamics between different marketing channels is key in being able to determine the right mix. Perhaps social media isn’t directly producing many sales on its own, but removing it would impact the results of other marketing channels. In this case, social media needs to be given value for the assistance it is giving those channels, and perhaps, deserves more resources and optimization to produce better results elsewhere.
Conclusion
Avinash’s presentation brought up many more questions than those answered. He made us aware of challenges that we did not fully understand. It is now up to us to to find our own solutions to these challenges, as very few internet marketing solutions have a “one size fits all” answer.
Avinash helped me realize that though marketing is divided into many different departments, and those departments are divided into campaigns, which are then divided into other even more granular layers; their performance and optimization needs to be done by looking at the whole. We often go into granular details to be more thorough in our work. We often forget from looking so closely that we need to step back and see how each granular point may be effecting the other’s and thus the big picture.
