Ark-iving Information

Last week I wrote about importance of treating the information as an asset. The bottom line is that due to its intangibility, its value is difficult to measure, and thus more often than not, totally neglected and ignored. However, unless the value is shown, the managers will continue sidelining information management projects, unwillingly leading the organizations further into information overload chaos.

So is there any practical advice how to approach the value information estimation?  There is no simple answer as every organization perceives and uses information in different ways. However, like with Generally Accepted Accounting Principles, certain set of rules could be worked out allowing building foundation for estimation model.

Why do we need it?

With hundreds of projects, that organizations need to allocate limited financial and human resources, information management projects, usually are low on the list. Exceptions are fancy pet projects, like for example – implementing latest trendy applications. Often such projects bring limited benefits to the organization comparing with costs and efforts invested. Usually their business cases have financial models built on shaky numbers, basing primarily on soft benefit estimations, often discounted by accountants.

The reason why we want to estimate value of information is to make sure that the scarce resources are channeled to these projects that address needs related to what is most important to the organization. The segmentation of the information value could be done from the following three perspectives:

–          Business need – when information is part of the business workflow; related to improvements in productivity; when information is taking direct part in marketing strategy or it needs to be visible to external business stakeholders. Key role here also plays ensuring single version of truth – when the organization wants to make sure that the decisions are made based on latest version of information. The business need also includes improvement of the productivity by allowing users to find the information faster, but also to spend less time managing information that does not have to be managed to the same degree (ex. transitional records). Pareto rule applies here pretty well – spend 80% of your time on managing 20% of the most important informational artifacts.

–          Risk – organization needs to comply with legal, regulatory or statutory requirements; needs to provide evidence of business decisions, activities and transactions

–          Costs – to replace the information or costs related to acquire information, licensing and subscriptions

The value of information is realized only through its use and this should be criteria for its measurement. However – the ‘use’ is totally subjective. Attempts to measure information value by representing it through more tangible dimensions like availability through business intelligence tools or data volume cannot be successful. For example, in case of BI – the way how well information is aggregated cannot determine capabilities of organization’s management. On the other hand, volume of data does not reflect its quality, content or ability to find information within it.

Valuation approaches

As mentioned above, the valuation of the information is often complex due its dependency on many, often intangible factors. However there are some situation that the straight valuation could be done. But first let’s put some groundwork. There are two types of approaches:

–          Qualitative – tending to be subjective, describing information in terms of some categorization, often informal

–          Quantitative – based on hard numbers and as such, more objective and reproducible

The information valuation fits somewhere in the continuum between purely qualitative and purely quantitative. The degree of how close it will be to either side of the spectrum will depend on type of information, how it is used, its purpose, type of business organization, risks impacts, organizational culture and so on. The organization needs to develop set of classes for its informational assets and categorize the assets accordingly. Once this is done, various strategies to manage the information and prioritization of related projects will be possible.

The development of such set of classes should be governed by some basic principles.

Information Valuation Principles

1. The value of information increases with extent of its use
2. The value of information is constantly changing
3. The value of information is increasing with its accuracy
4. The value of information is increasing when synthetized with other information
5. The value of information changes in time

Valuation Dimensions

The above principles could help to develop few dimensions that could be used for measurement:

1. Business need and meaning, depending on type of information (plans, design documents, patents, contracts, invoices and so on).
2. Costs
   1. Historical (cost of replacement of information)
   2. Market value – capitalization when sold (for example patents), or what if information is available to competitors
   3. Present value – related to current usage in organization’s business processes
3. Time

Measurement Process

Depending on which class given informational asset was classified, various measurement approaches could be used. For some assets the measuring of value represented by one or two of the dimensions, could be fairly straightforward. For example – in case of patents, when sold, their market value could be easily estimated. For some others, no financial value could be directly assigned, and qualitative process would be required. For most however, it will be combination of both.

Who and when

Information valuation is part of data governance, and should be led by appointed business representatives, accountable for data in particular business area of the organization (Data Stewards). Because value of information is time dependent, the governance model needs to define a process where Data Stewards with team, periodically review the monitored assets, and update their value. Some of the assets will end their lifecycle, or their value will diminish, so they might fall away from the review scope, some new appear and will need to be included, and some others value will change impacting their priority rating. Data Stewards will need to work with several stakeholders to value the information:

–          Business stakeholders – business owners of the information, could be represented by business analysts

–          Risk/compliance stakeholders – representing business continuity and legal groups when required

–          Finance stakeholders – to help with value estimation and its relevancy (this works in both ways, helping educate the finance people about the informational assets values)

–          IT stakeholders – to help understand current challenges and costs of maintaining of the information. This could include data analysts and data administrators

How

The Data Stewards should develop (if they don’t have it already), rough information value chain, based on their business processes. Roughly 20% of these informational asset will be important and impacting the business performance or representing significant risks to the business.

Once this is done, the Data Stewards should define asset value classes, related to dimensions as described above. This will be an iterative process, and will require some adjustments later, once more information is gathered. As mentioned earlier, some of the classes will have direct quantitative attributes like financial value. For some however, although related to financial dimension, it cannot be estimated directly. In such cases, reversal of the process could be suitable – in other words what would be the cost to the organization, if the information is not available, would have to be recreated, what would be time and effort to reconstruct, or if it would be available to competitors.

Some information will need to be estimated by a qualitative approach, for example:

–          Importance could be represented by following ranking set:

1.  No economic value, little process impact
2. Useful, but loss means small impact on business
3. Important but replaceable, with medium consequences to business and costs
4. Important and serious consequences to business/legislative, regulatory or statutory requirements
5. Critical information, like permanent records (for example articles of incorporation), vital records for business continuity

–          Impact:

1. Transitory
2. Tactical
3. Operational
4. Strategic

–          Age:

1. Old, close to end of lifecycle
2. Medium, still needs to be preserved
3. New
4. Permanent

The process that I believe is most suitable for estimation – is Wideband Delphi technique. This process uses expert opinion and assists in achieving of a consensus. Data Steward (or appointed facilitator) describes the informational asset, providing as much information as possible, and dimensions to be estimated. The participants write down (without any discussion) their estimates. Data Steward asks participants whose values significantly differ, to explain their viewpoints. All are asked again to write their new estimates, and the process is repeated. Usually 2 to 3 iterations are sufficient. If not, it is indicating that particular asset requires more analysis, before the valuation exercise could be repeated.

Once the assets value is estimated, it should be made available to groups responsible for projects portfolio management as well as included in information architecture governance documentation.

Summary

The process of informational asset valuation is not simple and needs to be customized for each organization. Key role play Data Stewards who need to develop set of qualitative/quantitative/hybrid sets of criteria and classify the assets accordingly. Data Stewards should include other stakeholders in valuation of the assets and should do it periodically as the assets value changes over time. The output should be part of the information governance process and made available to project portfolio management groups as well as it should be part of information architectural governance.