Wednesday, March 24, 2010

ROI for Data Quality Initiatives

Gaining the support of senior management for spending on data quality is an essential but sometimes difficult task. Almost all well run companies use some form of capital budgeting for their investment decisions. Thus, a sound financial justification for data quality projects is the key to building a compelling case for spending on data quality. However, the financial benefits of data quality improvements can be difficult to measure. Nonetheless, if your proposal is factual and backed up by hard data, it stands a far better chance of getting approved. A proposal based on facts will be perceived as less risky and its projected returns will be considered to have a higher probability. The greater the projected financial return of a data quality project is; the greater is the likelihood that it will be approved.

If you are considering spending on an upstream data quality tool and are currently spending on downstream data quality solutions, your task can be considerably easier than for a de novo project. If you already have a downstream data quality system, you likely have data on how much employee time is required to cleanse your data bases or if you are using an outside service you have received invoices that document the cost of that data quality effort. Upstream data quality tools will eliminate much of the downstream costs. Thus, making an estimate of the amount of benefit provided by upstream data quality tools can be relatively easy and will be based on hard facts not supposition. You can use Ikhana’s ROI calculator to create a quick return on investment (ROI) for your data quality needs.

Return on investment is a quick, but somewhat crude, measurement of the financial worthiness of a project. It is simply the benefits gained from an expenditure minus the expenditure with that difference divided by the expenditure. In the above case, the ROI is the cost savings from an upstream data quality tool less the cost of the tool and that difference divided by the cost of the tool.

Another simple financial analysis that can be used is the payback period . A payback period is the amount of time it takes for an investment to pay for itself. If an investment has a payback period of as little as a few months, it will be very attractive to company management for a couple of reasons. One reason is that the internal rate of return (IRR) (more about this later) will probably be very high because the investment is recouped so quickly. Another reason is that, because the investment is recouped quickly, the risk is less. The more time that passes after an investment is made before the returns pay for it, the more time there is for something to go wrong. If an investment pays for itself immediately, it is a “no brainer” to make the investment. The IRR of the investment is almost infinite. There are some who argue that the payback period is an essential metric for evaluating IT projects.

The IRR and net present value (NPV) are perhaps the most common and most trusted capital budgeting tools. For projects that require multiple periods to implement or multiple periods to recoup the initial investment, the IRR and NPV are the most appropriate analytical tools. The reason for this is the concept of the time value of money. Basically, the notion of the time value of money is that a dollar today is worth more than a dollar a year (or any period of time) from now. The rate at which the value declines over time is known as the discount rate. For a company the appropriate discount rate is the company’s weighted average cost of capital (WACC). Many companies add a risk premium to the WACC to arrive at their discount rate.

To calculate the NPV of an investment proposal the future benefits from the investment are discounted back to today using the company’s discount rate (WACC). Then, when the investment costs are subtracted from the discounted benefits the result is the NPV. If the NPV is positive, it indicates that the project could be a good investment and should result in an increase in the value of the enterprise because the future benefits exceed the costs of the project even when discounted at the company’s cost of capital.

The IRR gives financial analysts another metric to judge the merits of an investment. The IRR uses the same concepts as NPV. However, the IRR is itself a discount rate. It is the rate that when used to discount the future cash flows of an investment will result in a discounted value equal to the cost of the investment. It is the return on the investment over time. (No wonder they call it the internal rate of return.) If the IRR of an investment is greater than a company’s WACC, the investment will have a positive NPV. The greater the spread of the IRR over the WACC, the more attractive the potential investment will be.

This is a very cursory overview of financial analysis for capital budgeting decisions. Please refer to the hyperlinks for more detailed explanations.

In a future blog we will discuss how you can estimate the benefits of a data quality initiative.

Tuesday, June 9, 2009

The web user experience vs. data quality.

I sat in on a very interesting meeting with a client today that I wanted to share.

My client is a large multi-national service provider. They maintain an enterprise wide automated marketing system that is used primarily for communications (as opposed to lead gen).

They have several web development groups who are responsible for various corporate web sites. All of these sites feed contact data to the marketing system. However, the group responsible for the data collected (my primary client) has no control over how the response forms or landing pages are built by the various departments which own the sites and none of the other sites employ the same upstream data quality and standardization tools used by my primary client. Some of the problems that result are:

Company name is the primary focus for corporate data analytics, but most of the response forms do not ask for company name;

Some of the response forms ask for delimited names (i.e., first name, last name) and others ask for the
respondents name in a single field;

Some forms ask for email and phone, some email only;

Some data points required on one sites form are omitted by other sites; and,

Without universal requirements and data standardization, duplicates become a problem in the marketing database.


The web development groups are focused solely (at least it was apparent to me from our dialog) on the user experience. They have very little interest in the data received from their respective response forms.

I’m sure other organizations with multiple web sites offering different types of content share these same problems. But there is very little discussion on this topic that I can locate on the web.

Can anyone steer me towards others dealing with this topic?

A question for marketing data management professionals….

I have been involved in marketing and CRM data quality initiatives for a number of clients since early 2000. In all my encounters with companies trying to manage marketing and CRM data feeds from web sources, I have found they employ a number of individuals whose primary job roles is to manually edit and “cleanse” the data they have received after it is in the database. Clients who typically receive over 1,000 contact records per day often employ a number of people to manually cleanse data after receipt.

So my question is this: does your organization dedicate staff to manually cleaning and standardizing data? If so, please share the number of people you use in this capacity and the average number of records you receive on a daily basis.

Data Quality - Upstream or Downstream?

We are in the upstream data quality software business, so I keep wondering why data quality processes are still run once in a while, rather than as a normal part of the data capture process. Why do most companies start worrying about data only when it’s already dirty, already in the database, and in use? How come it doesn’t occur to them that the quality of data needs to be addressed when it’s actually captured? Since many data quality issues can be addressed at the point of data capture, why don’t more companies use upstream processes to improve their data?

A recent Forrester paper titled It’s Time To Invest In Upstream Data Quality suggests that when companies realize short-term data cleanup ROI immediately, it’s hard to justify front-end investments that may take years.

At the same time, Forrester says, IT budget planning committees tend to avoid the existing data quality (DQ) products that allow integrating downstream data hygiene rules into front-end processes, justifying this by solutions’ cost and complexity.

The result? I&KM (Information and Knowledge Management) pros quickly reach diminishing return on data quality investments, requiring even more investments later on to catch up with missed opportunities like verifying customer contact information, standardizing product data, and eliminating duplicate records.

Read the paper to find out more.

f you are thinking about implementing an upstream data quality solution, or if you already have, chime in here and let us know your thoughts.