Chapter 3
Turn Knowledge into Wisdom
In This Chapter
▶ Discerning business intelligence
▶ Making reports useful for all
▶ Getting the scoop on analytics
▶ Letting trends tell a story
▶ Driving your business with dashboards
Extracting information from the data within your many systems isn’t easy, but it’s critical for making good business decisions. Many businesses today are both data-rich and knowledge-poor. In this chapter, we explore reporting and analytics, note the differences, and demonstrate why they’re so important to effectively managing a project-based business.
Getting Smart with Business Intelligence
It’s awesome to have mounds of data, but data is nearly use- less if you can’t turn it into actionable information. That’s why your company needs a business intelligence application (that’s BI for short) to make sense of all of the data. A good application is a one-stop shop for assessing results by project, department, account, customer, organization, vendor, and so forth.
All BI systems have some components in common:
✓ Data: All your functional areas probably have their own applications, databases, and spreadsheets. Businesses maintain systems for accounting, time and expenses, human resources, customer relationship management, budgeting, and manufacturing (if that’s what your busi- ness does).
✓ Metadata: Think of metadata as the way that your BI system makes sense of data. It’s kind of like a library catalog, categorizing data by all the attributes that are important to your business. Metadata is what makes a good BI application really shine, but building metadata yourself takes time, often months.
✓ Report-authoring interface: You need access to a straightforward graphical user interface for your report writers. Your authors will choose elements from the metadata (for example, revenue by project) and structure the report in a manner that’s easily understood.
✓ Consumer interface: This is the “face” of the application that most of your organization will see, and it can make or break you. Your users want easy access to the infor- mation they need so they can run reports quickly.
✓ Administrator interface: The administrator must be able to secure data, control the environment, and monitor use.
When selecting a BI system to help your business make sense of your data, evaluate each solution’s power as well as ease of use. If you’re considering a homegrown tool, be sure to assess the total cost of ownership — sometimes saving some money upfront can be costly down the road.
Creating the Most Useful Reports
Good reports make everyone happy — bad reports waste people’s time. Following are some simple tactics that will help make BI reports valuable for your business:
✓ Keep it simple: Consider what users really want and need to see. Reports should fit the users’ “care-abouts.” Overcomplicating the report will just prevent people from using it.
✓ Be current: As with fruits and vegetables, the freshest possible data is the best. Your business can’t make criti- cal decisions as effectively if the information isn’t current.
✓ Communicate to the report users: When you’re creating reports, be sure you understand the requirements, share drafts of the reports, and then monitor usage of each report.
✓ Stay consistent: Try to use the same reports for every- one as much as practical. In a perfect world, you may change the delivery method but not the report or the underlying data.
Getting to Know Analytics
There’s knowledge, and then there’s wisdom. They’re two dif- ferent things, but the former can lead to the latter. Think of your company’s data as knowledge. You’ve got lots of it, but do you have the wisdom that comes from understanding that data? That’s what analytics is all about.
Analytics is best defined as the use of data and analysis to identify trends and make business decisions. Although analyt- ics may be easily confused with reporting, there are notable differences, which you can see in Table 3-1.
An analytics platform is deployed as a series of dashboards, with each one typically tied to a specific key performance indi- cator (KPI) or metric. Because different members of the orga- nization may be interested in different KPIs, your executives may follow one set of dashboards while your project manag- ers follow another.
KPIs are the measures that drive your business. Although each business is different, there are several KPIs that apply nearly universally to project-based businesses:
✓ Revenue
✓ Profit
✓ Backlog
✓ Labor utilization
✓ Indirect rates
✓ Proposal win rate
✓ Days sales outstanding
Looking at Budgets, Forecasts, and Trends
If someone told you a division’s labor utilization for the first six months of the year was 75 percent, what kind of assump- tions would you make and what actions would you recom- mend? Hard to say, because it’s difficult to make sense of that statistic without any perspective or comparisons. Now imagine you also learned that the division’s utilization for the prior year was 78 percent and that the budget for this year is 82 percent. Once you know that, you have a better idea where the division stands, and you’d probably be concerned. Stir in a little more analysis, and you could start to suggest correc- tive courses of action.
Comparisons to budgets, forecasts, or targets are critical components of an analytics environment. As the example highlights, metrics in a vacuum are virtually meaningless.
Evaluating performance against the baseline budget is often the best way to gauge whether the organization is meeting expectations. In Figure 3-1, you can easily see that the organi- zation’s revenue for March and April 2011 fell below budget, while results for the preceding several months were very positive.
Forecasts typically represent the “best guess” today of how the organization will perform in the future. Combined with actual results, forecasts can provide valuable visualizations of trends. In Figure 3-2, the shaded columns represent the funded backlog for the prior eight months. Performance has been relatively flat, and that information alone is fairly benign. To the right, however, are the forecasts for the coming months, and you can see that there’s reason to worry. Based on Figure 3-2, backlog is expected to fall dramatically over the next eight months, which means it’s time to investigate what can be done to change course.
These examples highlight how analytics can provide insights that drive informed decisions and illustrate the difference between reporting and analytics. Reporting nearly always focuses on looking in the rearview mirror. That can be impor- tant, but it’s also history, something that can’t be controlled. Chart out trends that incorporate both history and your future projections, and you can start to impact where you’re headed.
In the backlog example, the next step is to drill into that trend and isolate which divisions, program areas, or project manag- ers are most responsible for the anticipated slide. The prob- lem could simply be that part of the organization hasn’t fully updated its backlog forecast. But perhaps a key program area is failing to find any new funding or work is drying up.
Examining the Metrics That Matter Most
Every organization measures itself in a slightly different manner, but certain key metrics are relevant for a high per- centage of project-based businesses. For additional informa- tion on project-management-based analytics, see Earned Value Management For Dummies, Deltek Special Edition, or Integrated Program Management For Dummies, Deltek Special Edition. Here are some of the metrics likely to be most useful to your project-based business:
✓ Revenue: The revenue recognized by project-based busi- nesses is shaped substantially by the contract type(s) in effect for each project. Revenue is calculated differently for time and materials, cost-plus, and firm-fixed-price contracts — and countless variations exist for each pri- mary type. Evaluate your revenue analytic by organiza- tion, project manager, customer, and specific project.
✓ Profit: This is arguably the most important metric for any company and is the ultimate measure of your success. As with revenue, profit is impacted largely by contract type. Companies should be able to assess profitability trends across contract type to ensure that they’re pursuing the right type of business.
✓ Backlog: Your backlog analytic helps track how much work remains for your organization and allows you to measure whether you’re operating above or below your budget. Make sure that this analytic includes not just existing contracts but also those you’re proposing and hope to win.
✓ Labor utilization: This metric evaluates how efficiently your employees are being applied to direct, or billable, projects. The labor utilization analytic can offer insight into which employees are overperforming or underper- forming and let you know whether your staffing levels are appropriate. It’s imperative for management to be able to review both direct and indirect components of the metric.
✓ Indirect rates: Many companies track at least two dif- ferent versions of rates for indirect pools such as fringe, overhead, and G&A. The target rate is the estimated rate based on its budget, while the actual rate is calculated based on incurred costs. Comparing both as the fiscal year progresses is a critical function and a valuable com- ponent of an analytics system.
✓ Proposal win rate: No need to limit your analytics to financial data. For example, as a project-based organi- zation, your ability to win new business is paramount to your success. Setting targets for proposal win rates across different parts of your company will allow you to evaluate the performance of your business development function.
✓ Days sales outstanding: The processes of quickly and efficiently recording costs, billing the customer, and receiving payment have a dramatic impact on your com- pany’s success. Errors in coding vouchers or timesheets or delays in generating invoices can be devastating to cash flow. Days sales outstanding, or DSO, is a measure
of the time it takes to collect on an invoice, converting a receivable into cash.
✓ Projects at risk: Perhaps the greatest potential of an ana- lytics application is the ability to draw attention to areas that need corrective action. Project-based businesses want to quickly know whether they’re operating proj- ects that have the potential to generate losses or leave the company out of compliance with contract terms. Examples of risk categories include:
• Billing in excess of the funded contract value
• Costs incurred after the project’s end date
• Revenue recognized in excess of the contract value
• Revenue recognized below the budgeted amount
Designing Your Dashboards
Now that you’ve decided which analytics to track, you need to decide what sort of format should be used for the individual dashboards. For instance, what charts, tables, and visualiza- tions would give your executives the best sense of project profitability across your enterprise?
Start the same way you did with reporting. Ask your manag- ers what type of interface and information would be most beneficial to them in their decision-making process. Base your conversations and your dashboard design on the following considerations:
✓ Preferences for information consumption: Executives tend to opt for more graphical displays of information than other users, but that’s not always the case. Find out what level of detail is the “right” level.
✓ Quantity of information: With analytics, you can defi- nitely have too much of a good thing. Managers may ask for 20 different revenue charts, and they may be able to justify each. The end result, though, will be an exceedingly crowded screen that’s both confusing and overwhelming.
✓ Consistency in dashboard design: Executives aren’t hired based on their ability to interpret a dashboard, and they don’t have time to become familiar with a bunch of different dashboard styles. Build some consistency into each dashboard in your analytic application to minimize the learning curve of your users.
✓ Timeliness of information: Make sure the dashboards are refreshed as often as reasonably possible. Analytics, by their nature, tend to be summary-level measures that don’t change dramatically every minute. It’s usually acceptable to update your dashboards with new data every night, but make sure your executives are aware
of when the last refresh occurred.
✓ Control of dashboard look and feel: For years, dash- boards were designed and controlled centrally, with very little flexibility for users to modify the appearance of the interface. These days, users have more control and can interact more powerfully with the dashboards and even add their own objects.
Displaying a Thousand Words
Charts that once required days of work by a graphic designer can now be rendered in minutes or less, telling stories more effectively. They’re often visually stunning, but are they the right fit for the analytical needs of a project-based company? Here are some thoughts:
✓ Keep the dashboard simple: Restrict your display to the information that is relevant to the decision-maker. Similarly, avoid using chart types that are cluttered and require extensive explanation.
✓ Match the chart type to what is being measured: Different chart types are inherently better at displaying different types of information. Line graphs work best when measuring progress over time. Pie charts effectively com- pare percentages of various components to the whole.
✓ Avoid the bells and whistles: Technology makes it tempting to overload dashboards with objects that flash, spin, and jump off the screen. They look great in demos, but if they distract from the message of the dashboard, leave them out.
✓ Remember the function of the visualization: Each object on the dashboard should have a purpose. One component may exist to explain how revenue trends over the course of a year. Think about what type of object would best serve that purpose (a line chart, per- haps) and what options should be available to the user (analyzing the data by organization, contract type, or project manager, for example).