Web Analytics & E-commerce (En)

Using Data to Improve Results - Lessons from the Real World of Optimisation by Matthew Tod, Logan Tod & Co

This article is a summary of Matthew Tod's presentation during the Omniture Summit in London on April 21 2009. Matthew is Chief Executive at Logan Tod & Co

 

Key messages:

• 3 elements for successful optimisation: technology, process and people.

• Setting the right objective: PROFIT.
  
  
• We've got to look beyond the traditional web analytics metrics: traffic and conversation rate, and to focus on profit.

• Gross Profit Optimisation Drivers:
  1. Direct income : traffic, conversion rate, items per order, average item value
  2. Relevant costs : product cost, shipping costs, promotional discounts, marketing costs and return& charge backs.

• It is possible to configure SiteCatalyst to capture / import the data you need to look at Gross Profit in your reports.

• A Gross Profit model to aid optimisation: you have to understand the revenue levers and cost levers, if you want to start to do the right thing.

• Plan for good data to enable the process:
  1. doing the initial deployment is just the start, if you fail to keep up with changes in your infrastructure or business needs you won't have the data,
  2. review your deployment and upgrade the tags,
  3. there is significant managment ressource needed to train and coach people (internal and external),
  4. there needs to be some budget to maintain data quality and integrity.

• Plan analysis based upon how people consider your offering before engaging / buying: last click is credited with the goal while there may be multiple campaigns that create the initial engagement. It is much bigger and longer than the single session on the web site.

• Campaign roles change with engagement: campaign usage changes with consideration. It is essential to look at campaign types (paid search, advertising, etc.) over time.

• Case-study on Kitbag, Europe's leading online sports retailer:
  1. First implemented HBX in 2005 as the first step to really understand what was happening across the sites.
  2. Training was done on several occasions, but it had no impact as it was too remote from people's day job.
  3. The solution was to implement standardized reports emailed to account managers daily/ weekly / monthly.
  4. A single calculator was also implemented in Excel in order to ensure profit focus.
  5. A 1:1 mentoring program was implemented with higher satisfaction, real action and benefit. The plan goign forward is to continue this program to develop the skills of the account managers further.
  6. Automation is the way to scale further: Mercado (now Omniture Merchandising) was implemented in 2008, and it is planned to link Mercado to SiteCatalyst to drive merchandising automation.

• And in injury time:
  1. Link analytics to Gross Profit
  2. Consideration is rising fast and you need to adopt your analytics processes to reflect this change.
  3. Work on optimization by focusing on the key constraints and testing them to find and deliver improvement. Repeat.
  4. People make things happen, so don't settle for basic training, figure our what it really takes to get people using analytics every day and execute a program to make it happen.

Note: this post has been written live. Please excuse any typos and brievety.

Developing a Cult of Analytics

This article is a summary of Steve Jackson and Vincent Kermorgant's presentation during the Omniture Summit in London on April 21st 2009.

Steve Jackson, a well known blogger and Web Analytics consultant to some of the worlds’ biggest companies, took us through an 8 step process designed to show what’s needed in an organisation to get the best from the Analytics investment. In big companies, it’s a change management programme and Vincent Kermorgant a Senior Web Analyst at Nokia discussed how Nokia built up their own culture and came up with the same processes as described by Steve Jackson.

Key messages:

• The following maturity model has been used at Nokia:
  Level 0: customer data ignored.
  Level 1: some value of concept data acknowledged (initial visibility via KPI reporting)
  Level 2: management based on facts i.e. consumer data (touch point reporting established and service personalization)
  Level 3: consumer-analytics driven management and operations
  Level 4: some opportunities leveraged Nokia-wide (cross-nokia targeting, up/cross sell and traffic generation)
  Level 5: major opportunities leveraged Nokia-wide (personalized content)

• To move from one level to another level, you must comply with all levels (Business use, data set, people and process, tools and architecture, data quality & privacy compliance).

• The 8 step process toward a data-driven culture:
  1. Urgency,
  2. Led top down,
  3. Vision,
  4. Communication,
  5. Actionable,
  6. Quick wins,
  7. Continuous Organizational Improvement
  8. Routinely used.

• Urgency has to be present. Without a sense of urgency, you won't get management buy-in. Create it. For instance: what are your competition doing while you don't do analytics? IF you don't do this in the next months/years, what will happen?

• Has to be led top town. Speak the CEO's language: $$$ instead of bounce rates. Transmit urgency. Create awareness with the tools. In 2008, Nokia established the Customer Data & Integration Program (CDIP) to build consumer data into a strategic asset: 3 years program, centrally funded and reporting to top management, guided by a steering group including all business parties, more than 50 persons (internals and externals).

• Vision (with analytics at the core): the vision shouldn't be about analytics, the vision should be great than that. But, in order to support and deliver the vision, you need to have an analytics approach based on the customer needs and behaviors.

• Internal communication: communication is of course important, but it should be targeted (for instance, communicate only on the KPIs that are relevant to the users). Communicate the urgency "we need improvement in these results in 3 weeks". Communicate also what is happening in the market place.

• Actionable: you need to build a cycle of analysis of analysis and action. Nokia is using the AECR model (Acquire, Engage, Convert and Retain). Scorecards are preferred to reports (that nobody reads anyway).

• Quick wins: these should be planned and monetized.

• Continuous organizational improvement: create roles that support analytics. Reward passion and create incentives to find business insights.

• Routinely used: one of the main challenges is to ensure that the web analytics approach becomes part of the day-to-day processes of the company on the long run.

Note: this post has been written live. Please excuse any typos and brievety.

A Thousand Points of Interaction: Measuring Multichannel Online, Offline and Everywhere in Between by Robert Viebahn, Nintendo Europe

This article is a summary of Robert Viebahn's presentation during the Omniture Summit in London on April 21st 2009. Robert is Director of Online & CRM at Nintendo Europe.

Key messages:

• The Nintendo platform was redesigned in 2006/2007 with new website layout, new web architecture, new CMS and new analytics, mainly for new target audiences: seniors and women.
• The European platform is in 10 countries and 7 languages.
• Nintendo's policy of data protection: very strong data protection policy. No names and personal data are collected during the purchase process. No data can be collected from minors. Nintendo want users to register voluntarily to get premiums (Club Nintendo).
• The behavioral and social demographic data (age, gender, country, registered products, crossownership) are being merged into Discover On Premise, in order to perform detailed analysis.
• The data is then sliced, diced and analyzed with Discover On Premise, which is a very powerful tool for analyzing rapidly a tremendous amount of data in a very visual manner.
• Some aggregated data is then exported back to the CRM and CMS systems in order to enrich the users's profile (for instance: type of content viewed by a user) as well as to push personalized content.
• Main outcomes of the Discover On Premise implementation:

1. Faster customer analysis (1 day instead of 7)
2. Nintendo's web visitors have a more personalised experience
3. Technology is no longer a limitation, easy to use.
4. Nintendo now receives insight as and when it needs it, very flexible.
5. Nintendo can instantly see the success and failure of changes on the Websites.

Note: this post has been written live. Please excuse any typos and brievety.

The future for Customer Experience by Ed Thompson, Gartner Group

This article is a summary of Ed Thompson's presentation during the Omniture Summit in London on April 21st 2009. Ed is VP Distinguished Analyst at Gartner Group.

Key messages:

• 3 types of behaviors with this crisis: panickers, worriers and attackers.
• Cliché of our times: "A downturn is a terrible thing to waste" Frank Blake, Home Depot, 2009
• Business leader priorities are very clear : cutting operations costs (68 %), increasing revenue (53 %)
• Deeper sales and marketing strategies sort winners from losers in downturns.
• CEOs see value in sales, marketing technology and data-driven analytics.
• Focus is less on customer satisfaction but on enhancing cross-sell and up sell opportunities.
• Digital marketing is getting to start more funding than traditional marketing.
• Top 4 processes for improving operational efficiency: brand management, budgeting and financial management, marketing operations management and strategic planning and marketing performance management. Web analytics has a key role to play with all these areas.
• Organizations are now trying to move faster in the web analytics process maturity model from Bill Glassman (Gartner).
• Challenges for Web analytics leaders: competitive, attitudinal, behavioral, skills and operational.

• Predictions: by 2011 the online community will directly influence 33 % of all B2C online and offline purchases instead of 8 % in 2008.

• Social applications are relevant to each stage of the customer's buying process. Example: Wells Fargo connects with difficult-to reach demographics on Second Life. Newegg.com aids community feedback on every page.
• Customers expect more, will switch more and tell more.
• Is your company still thinking in terms of products or in terms of customer experience?
• Executives understand customer experience but organizations struggle to respond. Executives seek differentiation through customer experience, but complications arise (employee motivation, IT issues, etc.).
• How to measure customer experience? Different ways within the same organization: customer satisfaction (Customer service), loyalty (churn, retention), advocacy (referrals, customer refs), brand/reputation (marketing communications), etc.
• Your challenge is to drive emotional loyalty to your brand.
• The more business you move business online, you can drive the customer experience up (when funded properly).
• Avoid the middle ground. Customer intimacy, operational efficiency and product leadership are key types of customer experience. Choose the type of customer experience that better fits your industry instead of trying (and failing) to be the best in all 3 types.

• Recommandations:
  1. Digital:  is replacing traditional, accelerate the transition,
  2. Social: look beyond Facebook and implement what's already proven
  3. Customer experience: if Amazon can reach number one in customer satisfaction in the US, then any pure digital business can.

Note: this post has been written live. Please excuse any typos and brievety.

Accelerating Revenue Growth with Behavioural Driven e-mail Marketing by Nick Heys, Emailvision

This article is a summary of Nick Heys' presentation during the Omniture Summit in London on April 21st. Nick is CEO of Emailvision.

Key messages:

• Emailvision provides behavioral-targeted marketing tools, which integrate with the Omniture Suite.
• Example #1: recency of last click segmentation for GHD.
• Example #2: automating shopping cart abandonment Email. Sending a reminder (Pinnacle).
• Example #3: 3 Suisses has integrated customer data from their web site into their Email with personalized promotional Emails.
• Example #4: Siblu has developed a culture of testing by creating A/B split tests for every Email campaign.
• Example #5 : Brands4Friends are automating their testing (MVT) in their Email marketing campaigns with personalized promotions on the fly depending on the behavior.
• Emailvision's growth has been fueled by customers switching from in-house systems.
• 12,2 billion messages sent by Emailvision customers in 2008.
• More talent involved with Email marketing: on average 2,5 users per Emailvision client.

Note: this post has been written live. Please excuse any typos and brievety.

Building an analytics driven online strategy at John Lewis Direct by David Walmsley, John Lewis Direct

This article is a summary of David Walmsley's presentation during the Omniture Summit in London on April 21st 2009. David is Head of Web Selling at John Lewis Direct, a leading UK retailer.




Key messages:

• John Lewis Partnership founded in 1929.
• Partnership is owned in trust for the benefit of 68,000 partners.
• Johnlewis.com launched 2001.
• 2007 sales reached £268M
• 45,000 products online for sale.
• Seamless shopping experience between online and offline.

• 3 building blocks: 
  1. having the right organisation - a structure focused on the customer,
  2. developing an analytics-driven culture
  3. obsessing with the customer experience.

• Approach:
  1. getting the worldview right,
  2. test and learn (small changes with hypothesis)
  3. insight into action

• Example #1: optimization of the check-out. Several minor changes which make a huge difference at the end.
• Example #2: a battery of testing for the delivery options wording testing has been done.
• Example #3: understanding behavior for gender shopping (for instance: women shopping for men's clothes)

• Future trends:
  1. increasing focus on personalizing the experience with better targeting, merchandising, etc.
  2. fragmentation of messages/conversations rendring classic segmentation and data methodes largely redundant
  3. customers collaborating over what they see, where and when, diminishing role of push/broadcast targeting.

Note: this post has been written live. Please excuse any typos and brievety.

New Product Launch and Technology Showcase by Matt Belkin, Omniture

This article is a summary of Matt Belkin's presentation during the Omniture Summit in London on April 21st. Matt is SVP and General Manager, Emerging Business, Omniture.

Key messages:

• SearchCenter (bid management) and Test&Target (A/B and MVT testing) integration: you don't have to change your code on your landing page. Once the testing is completed, you can select to push the winning campaign to the regular traffic.

• Recommandations (cross and up-selling) : see JJ. Esquire demo site at www.jjesquire.com. You can select several recommandation types : Top sellers, most viewed, people who bought this bought that, etc. No need for IT involvement. Test&Target is also integrated with this tool and you can profile visitors based on previous behavior for further visits.

• Video Sharing: SiteCatalyst can now search for any videos based on keywords in Youtube, MSN Video, etc. and you can monitor how much this video spreads over the Internet (number of times a specific video is viewed, average rating, author and sharing site). You can do this on your own videos as well as your competitor's. You can then create a Viral Watchlist on your dashboard in order to tie together what's happening in the viral video space and your website.

Note: this post has been written live. Please excuse any typos and brievety.

Opportunities in Digital Marketing by Mark Read, WPP Digital

This article is a summary of Mark Read's presentation during the Omniture Summit in London on April 21st 2009. Mark is Director of Strategy & CEO of WPP Digital.

Key messages:

• Every client wants to spend more on digital. Digital represents today 26% of the revenues of WPP.
• WPP started investing in technologies with the acquisition of 24-7 Real Media 2 years ago.
• Social, mobile and analytics are key areas for innovation and investments.
• CMOs are looking for more accountability as well as for help in locking budgets, increasing profits and reducing costs.
• WPP is trying to bring creativity and technology together.
• 13% of all ad spent is now on the Web (UK: 27 %).
• Mobile is going to become the natural device for accessing the Web.

• Predictions: when we come back out of recession, the current trends will have accelerated and all things will be a step further. There will be a massive shift online. Big growth in social media. Optimistic with MySpace and Facebook usage and ad opportunities.

Note: this post has been written live. Please excuse any typos and brievety.

"Lead the change" by Josh James, Omniture

This article is a summary of Josh James' presentation during the Omniture Summit in London on April 21st 2009. Josh is CEO & Co-Founder of Omniture.

Key messages:

• Omniture contributes optimizing customer interactions, wherever they may happen (store, Web site, cell phone, wireless device in a car, etc.)
• The potential for the mobile is incredible (for instance in Japan where you can type coupons codes from billboards and get the content directly in your cell phone)
• In Japan, some corporations are doing more business with their mobile-based web sites than with their PC-based web sites. So we have to think the way we approach websites today.
• The world is changing so rapidly. If we don't have a strategy for the mobile Web, we are going to be left behind by our competition.

• Interactive marketer, search marketer, marketing VP, Web analytics, business analyst and developer should all be leading this transformation.

• Integrating different channels and marketing initiatives is key to get opportunities of optimization.

• Omniture defines itself as an optimization company focusing on CMOs and marketers.
• Omniture made $309M of revenue in 2008 with 5,100 customers worldwide.
• 17,000 servers around the globe.
• 200+ employees in EMEA with 1,100 customers.

• The next wave: Web / Mobile / Social

• Omniture is managing 250,000 transactions per seconds and 1,000,000,000 per quarter on behalf of their customers.

• Each of the 250,000 transactions per second is an opportunity for optimization. According to Omniture's estimate, less than 10 % is optimized.
• You can optimize on the previous navigation patterns, searched keywords, etc.

• Product annoucements:
  1. OnlineMarketing Suite (launched today in Europe),
  2. Recommandations (Amazon-like product and content recommandations. Target visitors on any channel: web, Email and mobile)
  3. SiteCatalyst for Viral Video (measure the viral spread of video, determine video campaign effectiveness and monetize video with advertisers).

• Recent product innovations: Merchandising (acquisition of Mercado) and DiscoverOnPremise for Retail.

• Omniture Developer connection: Yahoo Widget, Web Services, Geo Visitor App, WordPress plugin, Conversion Event Calculator, iGoogle Gadget.

• What's next:
  1. Omniture SearchCenter (optimize landing pages against keywords and ad copy)
  2. Test&Target (maximize return on ad spend)
  3. Omniture DigitalPulse (audit site and identify opportunities of optimization)
  4. CMO dashboard (integration of various data sources in a My Yahoo-like interface).

• Together, we lead the change: your ability to listen and optimize every customer interaction is your game changer.

Note: this post has been written live. Please excuse any typos and brievety.

London Calling - Live blogging from the Omniture Summit on April 21st

I will attend the Omniture Summit in London on April 21st, where I will perform live blogging on the main keynotes and presentations.

Stay tuned!

04/21/09 update: the blog posts for the Omniture Summit are available HERE.