WebTrends Connect ’08 Seattle – Data Warehouses and Customized User Engagement

Yesterday I attended WebTrends Connect ’08 in Seattle, Washington with Ryan Summers of ISITE Design.

The event was three hours long, and fast-paced. We arrived at 9:00 Am to find a speaker from WebTrends discussing the future state of news. He asked us to consider what could happen if one used online traffic to drive what is written about in analog newspapers.

As the seminar progressed, the ability to stitch together a holistic view of customer experience became a primary area of concern. One of the first major points was that Marketing and IT need to be working together.

“There is a lot of information locked behind closed doors”, said the next speaker, who was from TeraData. “There is no way to get the data out of the web analytics solutions and into the reporting dashboard.

He proposed the idea of business intelligence tools that could access this online visitor data and put it into an enterprise data warehouse.

He pointed out that one of the current difficulties of using data is that we are making a transition from 2nd Generation Enterprise to 3rd Generation Enterprise systems.

The 2nd Generation is Closed + Proprietary, whereas the 3rd Generation is Open standard-based. You simply can’t inegrate data systems when they are separated by proprietary, closed systems.

As O’Reilly once stated, “The internet is becomnig an enormous database that can be quiried, sorted, and applied to existing models and practices to change things”. WebTrends, TeraData, and other new systems seek to integrate many different systems with Analytics. The only way to streamline the spaces between data and change is to remove the closed doors betwen that data.

He stated that tech solutions should be open source based. These technologies seek to implement solutions that bring the two together and erases the nodes between them. He also pointed out that standards need to be in place that everyone agrees about across the organization.

Data Flow and Analytics

Online data can influence customer marketing, call centers, data warehouses, CRM and merchandisers.

Business Centric
—-
Business strategy
Performance Management
People — Processes
Analytics Processes
Analytics Applications
BI Platforms
Information Management Infrastructure.
——————

The next speaker was Paul Barrett
Global Director, Customer Management and DIgital Advertising

“How does one create one vIew of your customer?”, he asked, responding that the solution was “an Interactive marketing Intelligence across the enterprise”.

“Marketing an IT don’t like to play together — they have completely different mindsets.” A solution is to create Integrated Data warehouses so that the website and the customer service can blend into each other. Bringing that relationship closer together allows a more holistic view of all of the data coming into a company.

One of the most difficult issues is getting the recommended changes implemented in a shorter period of time every time.

It’s not about the data you can get at, but the risk of not knowing the data you’re missing.

The metrics are cheap, bu the metrics you don’t know are not.

Case Study: Travelocity

Tavelocity took the idea of using the CRM to drive customers to offers through the website. They had to avoid things like, “if you get a flight to NY for $500, and you log back out, when you next visit the site you can’t be shown an offer for $300, since you just bought the $500 one”.

They have a lot of dynamic decision making since you’ve already placed an order for $500.

Whne you become able to share the data between these systems, you become able to provide customized experiences for your customers with data that revolves around them.

You can also begin to bring unstructured data, such as the data on blogs into analytic understanding. For instance, when people blog about your website you can bring that data into your data warehouse.

The next speaker was Kevin Bobowski: Vice President, Client Services at Statigent

Stratigent is Partnering with Exact Target to provide a variety of KPI’s, such as benchmarkeing, competitive intelligence, visitor engagement measurement, optimization, proactive reporting and analysis, website testing and optimization.

“Don’t start with strategies that are too high-level that you don’t see value from them in the short term”, Bobowski said.

Primary challenges facing organizations today:

  • So many initiatives and not enough time to implement all of them
  • Picking and choosing what to do in a timely matter reduces liability
  • Intense pressure to produce results
  • Increasingly complex demands (coming from your business stockholders — their demand for information, customer insights)
  • New marketing technologies create more data and more silos
  • Balance to Building out these Capabilities

Strategy + Business Process

Do you have a unified strategy and clear goals that are measurement?

Infrastructure

Is realible and flexible technology in place to meet the evolving needs of ke sakehodes?

Value Creation Tactics

(Any processes a business uses to glean value from your data, testing, campaign analysis, conversion testing, customer segmentation).
What actions are you taking — on a consistent basis – to drive ROI?
Are the building blocks in place for this?

Isolated Successes vs. Long-term success

An organization with isolated successes – look at the successful campaign. This is often an indicator that there’s an executive sponsorship.

Where did you get your data, how did you get your data? instead of how to analyze your data.

Building a ‘culture of analytics’

The organization needs to invest in value creation tactics.
Demonstrate short term results that allow you to gain greater sponsorship and credit for lager projects — with the long term strategy and goals in mind.

He coined the term ‘Stratactical’, which he defined as, “of or relating to a strategy driven-approach using value based tatics”, adding that “while it’s great to have a long term strategy, you also have to balance it with results on the short term. You need to develop reliably, stable success.”

43 Percent of organization say they’ve started the process — but they’re not seeing any value.

How can you get hte results? How can you guarantee that those results will net short term wins

Key Preformance Indicators

It is difficult to show ROI from a seamless cross-channel customer experience with personalization and customization in place.

It is less difficult to Show ROI from a trigger-based communications program with customer profiling and predictive modeling.

4 tactics for a successful business case

-Build an actionable strategy
-Connect your marketing data
-Establish relationships (where are the shared pain points across the organization? Data silos make it difficult to have everyone on the same page).
-Optimize, Test, and Repeat…
-Incorporate resting to amplify the value you generate

Building an Actionable Strategy

-Optimize the Media Mix (return on ad spend)
-Increase the most productive spend
-Increase E-mail spend if cost per acquisition is less than direct mail and other marketing tactics
Reduce acquisition costs and increase profits per customers
Find and understand the total cost per Acquisition.

5 Types of KPI’s

-Outcome/Business driver

Return on Ad spend (Answers the question of”how well am I doing?”

-Diagnostic metrics (Helps you answer the question “how can I do better?”. An example is a conversion rate — click through to a landing page).

-Smoke alarms (Helps you anticipate potential problems that may exist. Example is number of unopened E-mails. That’s an indicator that something might not be right — might be sending E-mails to the wrong audience. Allows you to dig deeper).

-Predictor KPI’s (Allows you to look into the future — Answers the question of will I do better tomorrow? A client may invest in a banner campaign. Customers may need an amount of time to evaluate the purpose. Banner clicks might not convert immediately — in a day, week or month. Some organizations know up front, and they can prepare for revenue and stock — for how well they’ll be doing in 45 days).

-Latent KPI (the most valuable of all. Helps you answer the question of where are my marketing opportunities. Can take the form of customer surveys on the websites. This data sometimes sits solely in the marketing department and is not let out, but the data there should be shared across many channels — because it can help every department understand how others are seeing their organization).

Business Philosophy

Don’t let the KPI’s change every week or month, or else everyone in the organization will have a difficult time synching with each new idea, method, or direction. Focus, and slough off things that don’t match that focus.

Connecting your Marketing Data

The question is where are you as an organization? If your organization is not advanced enough

ExactTarget Email Integration

You can exact a customer’s E-mail address and the Product SKU’s they’ve purchased/looked at, as well as sales funnel abandonment info.

You can use ExactTarget automation to pick a file up, bring it in, and send and email leveraging that data.

Creating a targeted,one-to-one message using our proprietary scripting language.
Press the “Start” button and go about you your daily business.

Promote product A, but you know that customers who purchase product A also purchase products B and C. So you can include those in dynamic E-mails based on their interest in product A.

Promote product B and C automatically, but only if that product is in stock.

Optimization

Optimization is bigger than testing but testing plays a key role in the organization. Testing can drive short term wins.

Optimize emails creatives to increase click-throughs

-Headlines
-Offers
-Message/Copy
-Images
-Call to actions

Try different calls to action — when you begin to use multi-variate testing, you increase the capability to really increase your ROI.

Increase lending page conversion rates by multi-variate testing in the same way

-Headline
-Form field
-Color scheme
-Calls to action

1. Be Stratatical. Make sure that strategy is actionable.
2. KPI’s. Abolsutley essential.
3. Breaking down data silos allows data to flow into larger areas. These data areas combine into one bigger view of the customer. Which allows a richer view of the customer.
4. Optimization testing. It cannot be said enough how important optimization testing is. How else can you know what is successful in the site and wha epople are looking at.

KPI’s can exist on every level.

For copies of the Slides, E-mail:
Kevin.Bobowski@Stratigent.com

Many Ways to Get One View

Barry Parshall, Director of Product Management

The more plugged in the rest of the organization is, the more successful that organization can work together.

We see more and more organizations taking the data out of WebTrends and turning it into their own

Microsoft has categorized all of their key metrics into dashboards.

WebTrends 8.6 Demo

The UI was taken from Microsoft Outlook.

OTBC Drivers

A means for getting direct access to WebTrends data bins with an ODBC-compliant application.

Same driver for WebTrends Analytics and WebTrends Marketing Warehouse

Easy to install and use.

Demo, WebTrends ODBC Driver

-Open Excel 2007
-Click Data from other sources —> from Microsoft query
-WebTrends Demo — the data source for WebTrends already set up.
-Click ‘ok’
-This will connect you to the webtrends backend — select the profile. double click the ones that one cares about. Can grab multiple data channels and pull them into the same Excel data sheet.
Next –> you can sort by anything you want.
-Choose ‘sort by revenue’, descending.
-Click Finish.

Limitations

Ryan Summers informed me that, “you can only quiery one time period at a time. You can’t query two time periods”.

Datalinks for Microsoft Excel

From Business Intelligence, Inc. (a Portland, Oregon Company).

You can join all of that data with other data sources. Provides a real simple way to join that data with other data sources. Will also export the data intelligence to Excel — so that you can go to Excel, hit refresh, and Excel will auto update all of the fields.

Webservices API —> can directly implement data in and out of the data warehouse. Bi-directional data transport for WebTrends Marketing Warehouse
Based on SOAP.

TechnologyLeaders (New York)

Dynamic Alerts lets you know when to act — when a variable changes dramatically, you are sent a notification E-mail.
Event envelope based on historical norms. Alerts are sent when activites exceeded historical norms.

DataMind (Seattle, Wa)

They produce a very slick custom scorecard offering the capability to export all of your analytics to one place. There’s also a browser overlay application.

Demo

All Excel based. Data arranged by tabs. Capturing a thumbnail of your website and overlaying the analytical data over it. Consultants customize the scorecard for your business.

Data Scheduler
Allows the scheduling of an export for the underlying analysis tables that contains hundreds of thousands of data sources.

The WebTrends Marketing Warehouse

-A true data warehouse, based on Microsoft SQL Server 2008.
-Stores and maintains discrete records of visitors, and all “events” taken by them.
-Every campaign click-through, ect.
-Limitless correlations and audience segmentation capabilities
(Get me a list of visitors that met all of this criteria).

The fundamental difference between aggregate web analytics, and customer centric web analytics is the ability to run queries.

Case Studies

Orbitz, a TeraData customer, is using it for:
-Web behavior connected with offline transactions
-Combine data analysis, reporting in its own warehouse.
Polaris
-Web scoring accurately segments diverse audiences
-Use WebTrends Score to assign point values to particular onine actions (as a mean to quickly asses consumer interest in some products vs. others)
-On a nightly basis exact
use this to populate further interaction .

The Microsoft SQL Database is automatically populated with data and assigns each user action with a number.

Polaris Home Page (rule sets)
New vehicle interest: 1 (user clicked here but left after a few seconds)
Racing Interest: 0
Parts and gear interest: 15 (user clicked here and watched a video on parts and gear)
Engaged (how interested is your user?): 10 (video watching)
Total Score: 26

————

Conclusion

While some of these techniques are not new, it was nice to see many of them presented in three hours. It is apparent that bringing data from different sectors into common areas will help many to understand how users and companies are interacting. Analytics are becoming essential for companies to efficiently connect and deal with many customers at once while providing them with customized experiences.

I look forward to watching industries and products that help reduce the data silos that affect many current companies. The technologies are there — it is just an issue of getting these technologies into companies so that more users can be understood. I am sure that interest in these tools will only increase in the future.

———–

About

Amber Case is a Cyborg Anthropologist and new media consultant living in Portland, Oregon. She likes to attend events and meet people in the industry. You can follow her on Twitter @caseorganic.

Leave a Reply