Developing a Business Case for an XML Authoring System

Featured Paper from XML 2001 Conference Proceedings

By: 
Jean Mercedes Hamilton, Valley Forge Information Systems 

Email: jean.hamilton@vftis.spx.com

1. Introduction

XML-based technology today provides solutions to many of our real-life business problems. The benefits of structured language (XML) authoring include single-source, multi-channel publishing, improved information quality and consistency and enhanced functionality of electronic output. But after we have informed ourselves about the technology and the products and convinced ourselves of its value to our business, we still need to convince senior management to invest budget dollars in our project.

In its basic form, an XML authoring system consists of an XML editor integrated with an XML repository or Content Management System (CMS) plus a publishing module for automated print and electronic output. Extended systems tend to include workflow and translation management components.

This paper will review the corporate approval process and the players involved, discuss ways in which an XML authoring system can increase revenue and cut costs, and provide an example for analyzing the break-even point of an XML authoring system project.

2. The Corporate Approval Process

Capital expenditures in the corporate world need to be planned, budgeted and approved. The approval process for an XML authoring system often runs like this:

1. Technologist "Tim" in the publications department gets tired of the inefficiencies in the current process and receives some first information about XML authoring.

2. Tim convinces his manager "Martha" of the need for XML. Martha, and possibly her manager, agrees to let Tim assemble a small group to do more investigation.

3. The new "XML team" sets out to learn about the technology. They attend conferences, ask product suppliers for demos. At the same time, they may start writing reports to Martha, in which they try to explain their current authoring process problems, user needs for the future and possible solutions which XML can provide.

4. At some point, Martha starts thinking about her budget for next year and asks the XML team what the new XML authoring system will cost. Tim provides her with product pricing information.

5. The XML team goes through the RFQ process: RFQ prepared by the team and sent out to potential suppliers; proposals and cost estimates received and evaluated by XML team; supplier short list chosen. After a 6 to 12 month courtship, the winning supplier is chosen.

6. After the proposals from the suppliers have been received, it becomes clear that the XML authoring system will be a major capital expenditure. Martha and her manager start taking it up the ladder.

7. Unfortunately, the information from the XML team has provided Martha with only one side of the coin: the costs. No one has tried to estimate the savings that the system will bring. As the capital expenditure request reaches the top echelons of the corporation, it is compared with other IT spending requests. More than likely, new product development systems, e-business initiatives and hardware upgrades are also on the agenda.

8. If Martha and the XML team are lucky, they will be asked to develop a business case and come back next year. Otherwise, their budget request will be nixed and funds will be appropriated to "core business" interests like product development or sales. In any case, the whole team, including the suppliers who fought hard to win this new customer, becomes disillusioned.

What could Tim and Martha have done differently to turn this scenario into a happy ending? There are three main things which need to be done to develop a business case: (1) understanding the players, (2) analyzing current costs and (3) calculating estimated savings from the new XML authoring system.

3. The Players

The publications department. The pubs team typically consists of subject matter experts, writers, editors and graphic artists who are challenged to create more documentation for more products in more markets every year. Depending on the size of the team, they may have one person or only 50% of a person dedicated to hardware/software support to keep things running.

The IS/IT department. Most companies have an IS team for hardware/software support, but they may also have IT specialists for developing new software systems. The pubs team support person may report into this organization. If so, the IS team will have general knowledge of the issues facing publications. If not, we need to prepare ourselves for quickly educating as many IS/IT people as possible to get them on our side. IS usually has a list of hardware and operating system platforms that they prefer to work with as well as a list of preferred software (read "database"). It behooves us to choose the preferred products for the XML authoring system.

The CIO. As head of the IS/IT department, the CIO is looking for ways to increase synergies between the hardware and software systems within the organization. Thus we can expect to hear "Just use the xyz repository that was recently installed for engineering / r+d / marketing (choose one). It handles versioning." Or even "XML? The new MS Office package supports XML. Just use that." At the same time, most CIOs want to get large, impressive systems in place to "leave their mark". Foremost in their minds may be ERP, PDM or e-business systems. An XML authoring system will need to compete for their attention and support.

The finance department. Both the purchasing and the accounting functions of finance can help us with the XML authoring system. Purchasing, of course, should be involved when developing the RFQ and while evaluating the cost estimates and potential suppliers. Purchasing will also be responsible for negotiating a final price with the chosen supplier. Accounting should be able to help us develop the business case by analyzing current costs, estimating increased revenue with the new system and developing a complete ROI analysis.

The CEO. As the economy tightens its belt, more and more capital expenditure decisions are making their way up to the CEO. If given the choice about where to invest the company's hard-earned cash, a CEO will most likely choose core areas like product development, R+D or sales before choosing publications. To counteract this, we need to show that the new XML authoring system will either increase revenue, cut costs or both.

4. Increasing Revenue

XML authoring systems can increase a company's revenues in various ways.

• Customer satisfaction and retention

  Customer retention is a competitive advantage which can be improved by offering more services to a customer. In the documentation world, this means personalized content. Product instructions, guides and repair information which are product specific have more value to a customer. In general, it is very difficult to estimate the impact of increased customer satisfaction on revenue, however, the costs of losing a customer are clear: acquiring a new customer costs six times as much as retaining an existing one.

• Products to market faster

  Most products must be delivered with some kind of documentation. If the documentation (for example vehicle owner guides) is not yet finished, it may delay product launch. An earlier product launch, on the other hand, will increase revenues. While this increase is quantifiable, it is not clear how large a role the documentation plays. Just because the instructions can be completed faster doesn't mean that the product is ready to roll out the door.

• New markets

  To enter new markets, product information and documentation must be translated into new languages. The XML authoring system, by decreasing translation costs, reduces the cost of a new market entry, thus reducing the entry barriers. A company's cost accountant or export product manager should have access to market entry costs figures and the resulting increased revenue. Decreased translation costs may make it more enticing for a company to enter a new market.

5. Cutting Costs

An XML authoring system not only increases a company's revenues, it also will produce cost savings in 6 areas:

• Increasing authoring productivity

  In traditional desktop publishing systems, 20-50% of an author's time is spent doing page layout - just formatting the page. XML authoring systems provide automated page layout. The author can now concentrate on being a subject matter expert and give up the tedious task of choosing fonts, font sizes, indentions, etc.

• Reducing publishing effort

  As a structured mark-up language, XML supports automated processes, including publishing. Once the XML documents have been authored, automated publishing processes take over to generate print, CD ROM, internet, PDA and mobile phone output.

• Increasing information reuse

  An XML authoring system with a CMS provides the ability to store and reuse information objects which are smaller than a typical document. (The choice of chunk size or granularity is an important decision, but is outside the scope of this paper.) Each object gets stored together with metadata which defines the range of products for which an object is valid.

  Metadata includes descriptors such as product model name / number, build date ranges and other attributes which describe the product. For automotive documentation, metadata typically includes model, model year (or VIN number range), engine type, transmission type, body style, right-hand-drive or left-hand-drive.

  A major goal of the CMS is to make it easier for an author to find and reuse an existing information object than to create new text. Using a CMS, information reuse can be increased by 30-60%.

• Reducing translation costs

  Translation costs can be reduced in two areas: (1) by reducing the amount of new text created or changed and (2) by eliminating the manual page layout of each translated language.

  There are three ways to reduce the amount of new/changed text: (1) Using a glossary and/or controlled vocabulary; (2) Information object reuse via the XML authoring system; (3) Using a translation memory system to find matching sentence pairs below the object level.

  Manual page layout of each translated language may account for a significant amount of the total translation costs. For example, with only 5 languages, language layout may account for up to 40% of the total translation bill. Since layout is automated in XML, these costs can be totally eliminated.

• Reducing maintenance costs

  As the XML authoring system provides a higher level of reuse, fewer new objects are created and thus fewer objects needed to be maintained or updated.

• Future-proofing data

  Companies which in the past have moved their data from WordPerfect to Word to a desktop publishing format to the next file format, etc., know the costs of data conversion. Since XML is a non-proprietary and standards-based markup language, native XML data is operating system and software product independent.

6. Developing the system vision and goals

The vision and goals for the XML authoring system should be developed early on and should include quantifiable objectives.

Here is an example of project goals which are difficult to add metrics to:

The goals of the ACME-Authoring System are to

• Provide our authors with an efficient XML authoring system

• Provide our distributors with a new and highly integrated information system.

Compare the above to quantifiable goals:

The new ACME-Authoring System will

• Reduce authoring costs by 30%

• Reduce translation costs by 40%

• Improve quality and consistency of our documentation

These are both real-life examples of goal statements. (Names are not used to protect the innocent.) If you were a senior manager, which one would you prefer?

7. Evaluating current costs

Most companies track the parts and labor costs of every screw which goes into their products, but only a few companies know how much their publications cost. If you belong to the former, you will have an easier job compiling the needed information. If you belong to the latter, the analysis will take longer, and while you're at it, you could consider introducing cost tracking methods.

To evaluate current costs, we start by breaking down the documentation process into the main groups: Authoring, Reviewing, Translation, Publishing. The following is a typical cost breakdown of the documentation process:

The following statistics are needed to evaluate the current costs:

• Number of authors

• Number of pages developed per author per year

• Hourly cost of average author (including salary and indirect overhead costs)

• Per cent of author time spent on page layout

• Per cent of author time spent on other duties (if applicable)

• Annual working hours

• Number of reviewers

• Number of pages reviewed per reviewer per year

• Hourly cost of average reviewer (including salary and indirect overhead costs)

• Translation costs per page

• Language page layout costs per page

• Total annual number of pages translated

• Number of translated languages

• Number of publishing experts (to generate paper, CD, internet output)

• Number of pages published per publisher per year

• Hourly cost of average publishing expert (including salary and indirect overhead costs)

The following table lists typical values from a current cost analysis:

The total current costs for this example are then:

8. Estimating cost savings and increased revenue

Our business case will focus mostly on the cost savings associated with using an XML authoring system.

Estimating cost savings

Now that we have calculated our current costs, we just need to choose a few saving parameters to calculate the estimated savings. XML authoring eliminates the need for page layout, both during authoring and later for language layout. These costs will disappear. Information object reuse can be increased between 30-60%. Publishing routines can be developed to automate print, CD and internet output. The table below lists some typical cost savings parameters:

Adding these cost parameters to our calculations results in the following estimated future costs:

The total costs have been reduced 50%!

Estimating increased revenue

Of the three ways in which an XML authoring system can increase revenues (customer satisfaction, products to market faster and new markets), the latter is the easiest to quantify. If your company is planning on entering new markets in the near-term, you should definitely include the reduced translation costs which will result from the new XML authoring system.

Using the data from the example above, the costs of adding a new translation language without the XML authoring system would be: $850,000. With the new XML authoring system, they costs would just be: $650,000. The savings of $200,000 results from the automated page layout which XML brings.

9. System rollout and timing - or - How quickly can savings be realized?

It typically takes 24 months for a new XML authoring system to be developed and implemented. But by using a phased delivery project management methodology, we can deliver parts of the system to the users earlier. After an elapsed project time of 12 months, authors should be able to start working in XML and thus start accruing savings. For more information on phased delivery project management, see the XML 1999 Philadelphia Conference proceedings or visit http://www.vfscout.com and look at the "Presentations - Saturn SPMS Case Study" section.

Here is an example of a typical XML authoring system roll-out and the savings that can be associated with it.

Project timing. The complete project will take 24 months. After 12 months, the XML editor will be installed with basic repository functions. 18 months after project start, the first reviewers will receive XML data. 21 months after project start, the first XML files will be sent to translation. Automated publishing will be in place 24 months after project start.

Each of these events generates cost savings. Our cost savings calculations from above provide us with average monthly savings amounts:

• Monthly authoring savings: $48,000

• Monthly review savings: $32,640

• Monthly translation savings: $47,917

• Monthly publishing savings: $50,400

To make our calculations even more conservative, let's reduce each of the above monthly cost savings figures by about 50%. We will use the following cost savings figures:

• Monthly authoring savings: $25,000

• Monthly review savings: $15,000

• Monthly translation savings: $25,000

• Monthly publishing savings: $25,000

In addition, let's assume that during the month where the new functionality is released, that only 50% of the team is affected. That is, in month 13, only 5 of the 10 authors will start using the system. The other 5 will start up in month 14.

Based on the assumptions given above, the accumulated savings will reach $2 million 41 months after project start.

10. Break-even analysis

To complete the business case and develop a break-even analysis, we need to map the cost savings to the project costs as they are incurred. For this example, a 10-seat XML authoring system, we will assume a total project cost of $1.5 million, paid in equal installments over the 24 month project period. Plotting the accumulated savings versus the accumulated costs results in a visual overview. We can quickly see that the break-even point is 36 months after project start, or just 12 months after the project is completed!

11. Return-on-Investment (ROI)

Return-on-Investment (ROI) is a financial investment analysis tool which is used to compare potential projects. A full ROI analysis must include the company's cost of capital, the corporate tax rate, investment tax credits, depreciation, cash flow and the time-value of money (net present value).

The description of a compete ROI calculation is beyond the scope of this paper. However, the business case and break-even analysis will provide your company's accountant with the information needed to to perform a complete ROI analysis. (An advanced Excel template for computing the net present value of a capital investment will be included on www.vfscout.com.)

12. Upgrading to a next-generation XML authoring system

For companies which are already using XML authoring and are planning to upgrade to a next-generation system, developing the business case may be more difficult. Essentially, your first-generation XML authoring system should have already given you the "easy wins" from automated layout and publishing in both the original language and the translations. However, you may still have inefficiencies in the system, areas which could still be automated, and you can always increase the reuse amount.

The basic procedure of analyzing your current costs and estimating future savings remains the same. Unfortunately, your cost analysis may now be twice as big. To do the job right, you should compare your costs during the pre-XML time to your current costs as well as comparing to your estimated future costs. After you had moved to XML, did you gain all of the efficiencies expected? What are your main cost drivers today? These are the types of questions you will need to examine.

13. Conclusion

XML authoring makes good business sense, even if the economy is slowing down. However, to get a new XML authoring system approved, we need to present the project to senior management in terms that they understand: quantifiable goals, reduced costs, break-even point, ROI. To get the project approved, you will need to make the project economics as important as the user functionality and the technology. Add a company accountant to your team early on to help with the number crunching.

A business case consists of a comparison between the savings and the project costs. The most difficult part of the exercise is to gather the current cost data. Using simple calculations in a spreadsheet program, it is easy to generate the break-even point of a project. A complete ROI analysis should be left to the professionals.

The example depicted in this paper develops a business case for a 10-seat authoring system with project costs of $1.5 million over 24 months. The authors today spend 30% of their time doing page layout. In the future XML system, page layout will be automated. In addition, information object reuse will increase by 40%. Given these basic parameters, the break-even point of the project (without considering the time-value of money) will be 36 months.