By Mark Lewis, special to The Content Wrangler
This is the second installment of the DITA Metrics series which examines the cost and reuse values for a DITA project to determine DITA ROI. The concepts and ideas discussed are based on the cost model introduced in the first paper, DITA Metrics: Cost Metrics – Part 1.
This paper looks at the savings trend when reusable master topics are used to document similar products. How much does it cost to document each additional similar product?
We’ll discuss how we derived the following savings trend (Figure 1) from our model so that you can modify the model to fit your content and your business requirements.
To benefit from this article, you should be familiar with the cost model discussed in the first paper in this series and you should have at least an intermediate level understanding of DITA including topic structure, elements, conrefs, child maps, and filtering/conditional processing.
In the first paper in this series, DITA Metrics: Cost Metrics, we created a model to determine the cost of creating user guides for three models of a fictitious personal digital assistant (PDA). We also looked at the technique of creating reusable master topics. This type of topic is created when the content for all PDAs is similar rather than unique. Using filtering metadata, content that is unique is marked as belonging to a specific PDA. When the user guide for a specific PDA is published, content that is specific to the other PDAs is filtered. The filtering feature in DITA is also known as conditional processing.
Since the model for the reusable master topic includes the cost of documenting three PDAs, in this paper, we separate the cost of adding a single PDA from this model and determine the savings trend for adding additional PDAs to the user guide. We show that if we are documenting similar products, using reusable master topics drastically reduces the overall cost.
Cost Per Product With Unique Topics
Table 1 shows the cost estimates for creating unique topics that we developed in the first paper.
For a given PDA feature, if we create unique content for each PDA, then we create a unique topic for that feature for each PDA. For example, if we have three PDAs and we need to document “How to add a contact,” we must create a task topic for PDA One, another for PDA Two, and another for PDA Three. Assuming that each task topic requires an accompanying concept topic as supporting material to describe background concepts, we also need to create three unique concept topics.
For each product that we add (Table 2), the cost of documenting a feature increases by the cost of one unique task topic, or 5.60 hours.
And for each product that we add (Table 3), the cost of documenting a feature increases by the cost of one unique concept topic, 5.20 hours.
Now we have the cost to document a feature using a unique topic per product.
This is really a worst case scenario, because most likely, some content would be copied and pasted from one topic to another. But copying and pasting comes at a cost as well. There is the cost of locating the content to copy, the cost of accurately copying and pasting the content, and the cost of customizing the content to a given product. Keep this in mind when we look at the savings trend later.
Cost Per Product With A Reusable Master Topic
Let’s determine the cost to document a feature for a group of similar products using one reusable master topic. First, we must determine the cost of the unique content to be created for each product that will be added to the reusable master topic.
We’ll begin by reviewing the cost ranges for creating reusable master task topics and reusable master concept topics that we developed in the first paper (Table 4), taking note of where unique content exists.
Table 4: Cost of a Reusable Master Topic”
Table 5 shows the total cost for each type of reusable master topic and averages those values.
Assuming we already determined our filtering strategy and know what filtering metadata to apply to product-specific content, we do not include the cost of developing that strategy here. However, it will take extra time to write a topic that is generic as possible and that implements the strategy, so add 1 hour for “writing with filters.” In addition, add the cost to test/review each filtered version of the published topic to ensure that we have applied filtered metadata correctly and that the filtered topic is accurate for each product. So, add 1.5 hours for “testing.”
Table 6 shows the effect of adding the “writing for filters” and “testing” costs to the average total cost values from Table 5.
The model includes the cost to document three PDAs. So, how much new content is added to the reusable master topic each time we add a new product? Table 7 shows the cost of content specific to a new PDA, PDA Four, in the reusable master task topic.”
For a task topic, Table 8 starts with the cost total for three PDAs from Table 6 and adds 1.45 hours for each new product.
Table 9 shows the cost of content specific to PDA Four in the reusable master concept topic.
For a concept topic, Table 10 starts with the cost total for three PDAs from Table 6 and adds 1.5 hours for each new product.
Now we have the cost to document a feature for a group of similar products using one reusable master topic.
Cost Comparison – Per Product: Unique Versus Reusable Master Topics
When we compare the cost of creating unique topics for each product to the cost of using one reusable master topic, the savings trend emerges.
To determine the savings trend for task topics, let’s merge Table 2 (Cost of Unique Task Topics for Each Product) and Table 8 (Cost of the Reusable Master Task Topic Per Product) . Table 11 shows the result.
Charting the data in Table 11 provides a graphical view of the savings trend for task topics.
Using the data in Table 11 in the following formulas, Table 12 shows the cost of the reusable master topic as a percentage of the cost of the group of unique concept topics. As the number of products increases, the relative cost decreases and the savings increases.
Cost = (Cost of One Reusable Master Topic) / (Cost of Unique Task Topics)
Savings = 1.0 – Cost
Our model helps us see that the cost reduction / savings can be significant.
To determine the savings trend for concept topics, let’s merge Table 3 (Cost of Unique Concept Topics for Each Product) and Table 10 (Cost of the Reusable Master Concept Topic Per Product). Table 13 shows the result.
Charting the data in Table 13 provides a graphical view of the savings trend for concept topics.
Using the data in Table 13 in the following formulas, Table 14 shows the cost of the reusable master topic as a percentage of the cost of the group of unique concept topics. As the number of products increases, the relative cost decreases and the savings increases.
Cost = (Cost of One Reusable Master Topic) / (Cost of Unique Concept Topics)
Savings = 1.0 – Cost
And our model helps us see that the cost reduction / savings for concept topics are also significant.
Customizing the Model
You can modify the model to fit your documentation projects by changing the components of the unique topic model to more closely match the DITA elements you typically use in your own task and concept topics. Add cost estimates for any new elements. Use the same technique to update the reusable master topic models and cascade the new costs through the model. With this custom model, you should be able to predict your own savings trend.
In our model, we see that the overall cost of documenting a given feature for a group of similar products using a reusable master topic is drastically lower than the cost of creating separate unique topics for each product. This is true for both task and concept topic types.
This paper focuses on creating content, but the savings in maintaining content would be similar, because we would update only one reusable master topic rather than a group of unique topics.
In summary, the trend is that as the number of similar products increases the average creation cost per publication decreases when you take advantage of reusable master topics.
About the author
Mark Lewis has received Society for Technical Communication (STC) awards for Distinguished Chapter Service and the Florida Technical Communications Competition. Mark is the DITA Product Manager for Usability and a product evangelist for Quark. He has presented on technical writing, DITA, and object oriented design topics at DocTrain, STC, DITA North America, and other national conferences.
Mark is a member of the Organization for the Advancement of Structured Information Standards (OASIS) DITA technical committee. He and John Hunt are co-chairs of the OASIS DITA for the Web subcommittee.
Return on Investment is a hot topic in technical writing. Mark has authored several white papers on DITA Metrics that prove the savings and high content reuse percentages possible with DITA’s structured, topic-based architecture. His DITA metrics model was a Center for Information Development Management – Rare Bird Award 2009 Competitor. Mark manages The Content Wrangler Community groups: DITA Metrics and Writing OBJECTively.
Send questions or feedback to firstname.lastname@example.org or email@example.com.