April 1997 to July 1998
Intel Corporation, Hillsboro, Oregon, USA
Central MarCom Services
Web MarCom Specialist
|
Key Accomplishments
|
- Developed and deployed one high-volume, high-exposure marketing web site for a Fortune 500 company.
- Developed several prototype HTML client/server interfaces for the divisional product group.
- Played my part in launching a company wide web authoring mentorship and training program.
- Achieved my personal operational goal in Web Operations: Lifetime Customer Value and Cost of Goods Sold data in real-time!!
|
Summary of Skills Used
|
PROJECT EXPERIENCE
|
TECH EXPERIENCE
|
Mentored Web Authors
Multi-Divisional Web Consulting
Liaison to Corporate Marketing
Liaison to Corporate MarCom
Liaison to AOL Corporation
Cross-Divisional Leadership
Marketing Programs
Contractor Management
Internal Trainings
|
HTML 3.x to 4.x
Cross-Platform JavaScript
Web Applications (Client/Server)
User Interface Design
Internet Telephony
Internet Chat Technology
AOL Publishing Environment
Ecommerce Applications
Statistical/Logging Techniques
Quality Assurance
|
HIGHLIGHTS
About Central MarCom Services
The Central MarCom Services group was assembled to provide a better way to share resources amongst the many small divisional marketing groups at Intel's Oregon plants. The group had two teams: Print/Production Services and Interactive Services.
The Interactive Services team consisted of about half-dozen interactive specialists who were divided amongst the small divisions as support and liaison to corporate resources. I was fortunate to be 100% devoted to one of the larger divisions, Internet Services Operation (ISO).
Refining Web Project Management
The Web was a new medium. Many different industries are touched by it. In web site creation all the traditional media agencies were trying to get a handle on how to develop content for it.
It was entertaining to watch the ad agencies try to fit the web into their printing paradigm or the multimedia publishers try to make the web fit their idea of a multimedia CDROM. Every single standard had to be broken and redefined. Every person with a tradition behind their profession was dragged kicking and screaming along the way.
Here are some examples:
- The famous "hits" versus "page views" debate.
- The never-accurate results of online ad performance.
- The Mac/Unix/DOS/Windows communications breakdown between agencies and vendors.
- The long filename debate.
- Trying to get management to understand the complexities of executing web programs.
Really, Web Design and Site Architecture is about the convergence of software development complexities with the out-of-box creativeness of graphics design. Except with more limitations.
Imagine putting engineers and marketers in the same room and locking the door. My colleagues and I were put in the position of making peace. We had to liaison between all these diverse people and pull together a final product that was compelling, interesting and useful.
To this end, I developed new ways of statistic analysis. I invented new ways of using client-side technologies. I worked with vendors to define a way of building web sites in a "modular" way (software design ideology) which allowed them to be maintained later after launch.
I had to hire and fire several different agencies (even the prestigious ones) over a period of time because they couldn't "get it". But eventually, I found a core group of flexible agencies and produced some great results!
User Interface Design
 I built several prototypes of the Intel AnswerExpress Service. Initially the plan was to build a customer service interface which you accessed directly through your web browser.
The division had already contracted with a very prestigious and very expensive creative agency (Studio Archetype -- now known as Sapient). When I came aboard the agency was mired in design indecision on the division's part. In the end, I had to code at least two different revisions of the prototype myself because the budget was tightening and Studio was just too expensive to keep, even though they were a fantastic group.
Online PC Service Portal
The original service plan was to develop a Web browser-based client (essentially a web site) which would require a mostly server-side solution using CGI and databases. We almost closed a deal with BroadVision to use their technologies under the hood. I spent a great many hours with BroadVision engineers who were assigned to me acting as the lead architect on the job. The hour before we were able to show our functional prototype to executive staff the plug was pulled on the deal due to financial issues.
Intel AnswerExpress Support Suite
The division hired-up their engineering groups and built a user-interface using Microsoft Internet Explorer v4 as the guts. It was sold as a client-side application on CDROM with a recurring monthly fee schedule for the online services. The key benefit of the service was the ability to get real-time response from online helpdesk people.
The rest of the service offering was about online content. We had assembled a huge library of online content ranging from online tutorials to PC tech notes. We partnered with ZDNet.com for most of the content. I helped liaison to Ziff-Davis to gather and integrate their content into our system.
Content Evaluation and Management
After the BroadVision project was canceled, I had to find other ways to be useful to the division. I helped architect a short-term content evaluation solution. Our Online Library feature needed to be managed and it was huge. I worked with a contractor to build a simple database using Microsoft Access and a web CGI interface to gather meta-data about all the content we were including in our Online Library. The information could then be output using standard SQL. The meta-data was then tied into the service offering allowing the Online Library to be browsed and searched.
My system got us through to version 1.0 in time to launch the service. But then became outdated because the engineering group had decided to build the rest of the system using new technologies for their databases, very cutting edge, but not SQL-compliant.
The engineering group decided to build custom tools and replace the simple CGI-MSAccess solution with custom built C++ programs. (Every time they needed to change the user interface for the content editors they had to engage an engineer and spend at least a month in the development cycle).
The Big Launch
Intel AnswerExpress was Intel's first business to consumer service offering ever. We pulled out all the stops. I negotiated star treatment for our divisional marketing programs all across the Intel Corporate web site. I had to spend a lot of influence points at the corporate level. We nearly reached the million hit milestone in our first months. This performance metric was a big deal for a small divisional effort. Me and my team received praise from the head of corporate Marketing Communications.
We had online ad banners, print advertising, and an online store to purchase it on. (The second ecommerce site in Intel's history). I had front page exposure on intel.com and held it for a month.
Ecommerce At Last
Intel AnswerExpress was Intel's second Ecommerce success ever. The first was the online shop for buying Intel mugs and trinkets. Due to the inability of Intel's internal operations to support fulfilling online orders (let alone taking online money) I had to contract an outside ecommerce handler (CyberSource) and use their online store embedded on the Intel.com web site. I also helped smooth the integration point between CyberSource's online order system and our divisional operations group by handling fulfillment integration and account management. We were selling a subscription service, so this was NOT easy.
AOL Presence
Our target customer was almost exactly the same as the typical AOL user. So we worked with AOL to build a presence there. In fact, I even (although I don't like to list it as an available skill) was sent to training to learn to edit and manage AOL content within the proprietary AOL environment.
Web Management Mentorship
 Along with a web author certification program that I helped design came the responsibility of mentoring new authors and helping them get certified. After the bureaucracy got their hands on the certification program it became a nightmare task for authors. It was so complex that at one point there was serious discussion about sending authors away for a week long training and a test at a Sylvan Testing Center.
Part of my role as a senior webmaster was mentoring new authors to help them study for and pass a certification test. I was also a resource for all Oregon and Utah authors during this period.
Mastering The Medium
Thanks to Java and JavaScript we can do many things with a Web browser and there are a fantastic number of possible applications you can build when you combine the client-side features of the browser with the server-side features and databases of the web servers. As Intel became more "organized" and process driven in its web server farm management I began to see the client-side solution set as my only option on a number of projects. During this tenure I invented methods of using JavaScript to pre-cache web site graphics during initial site load to speed up the experience for the users. I pioneered using JavaScript, Java, and GIF Animation as a medium for online demos. This effort resulted in an online "flash-like" entry point to the Intel AnswerExpress web site and an online tutorial complete with audio voiceover narration -- all with Java and JavaScript.
My divisional management was so excited about my pre-caching techniques they talked about patenting it. I quickly quashed that idea rather than cause one of the largest companies on earth to set a pretense for patenting web script code.
Being isolated from the server farms by the bureaucracy made a web designer crafty. My crafty solution to the nearly unusable statistics gathering being done on the servers was to implement a Java Counter on our web sites.
In essence my Java Counter was a simple Java applet that loads itself once per session and triggers a "hit" count on the server logs. The difference between a normal "hit" to an HTML page and a "hit" to a java applet is that the applet only loads once per browser session. That means that I could tell exactly how many individual users were visiting my sites -- a previously unheard of claim. I was able to prove that I had over 70,000 unique customers visit our online store in April of 1998.
Real Statistics for Planning
Thanks to techniques like my Java Counter and the ability to get to the financial data from our divisional controller I was able to prepare an actual statistical analysis of our success with our online store. I could project the Lifetime Customer Value (LCV) AND even more importantly the Cost of Goods Sold (COGS). Once again, something that had never been attempted in the history of the Web.
Moving Forward
I made my mark at Intel. I helped establish the hierarchy of web design and the practices we followed as authors. By the time I was ready to leave, bureaucracy had caught up with the innovation and put every thing under "control". A Corporate Electronic Marketing group started to take over and the divisional business units were being "rifled" through layer after layer of process.
Intel was also cutting back manpower during the end of this period. Whole divisions were being slashed and man-power was being tightened. It was clear that I was one of the "keepers" and that I'd soon be feeling the crunch of someone else's workload on top of my own.
During this period, I got a call from a colleague at RuleSpace, Inc. that said they needed a hard-working webmaster really fast. I needed a different challenge anyway and at RuleSpace, Inc. I could actually direct the whole web show, so I left Intel.
 previous experience
|
