Stop Trying To Put A Monetary Value On Data – It’s The Wrong Path

I recently received a client inquiry that I see flavors of a few times per quarter. The client said that they are trying to explore ways to establish information value within their enterprise. Because people often frame data and information as an asset, then shouldn’t we be able to establish its value?  

What I share with my clients is that trying to place a monetary value on data and information itself is a red herring, an effort that I highly recommend all avoid – unless you enjoy philosophical exercises that don’t translate to actual business value. (Apologies to those that fit in this camp — have fun!)

 The “data is an asset” rhetoric doesn’t translate to putting a monetary value on a customer record, as an example, because data in and of itself has no value! The only value data/information has to offer — and the reason I do still consider it an “asset” at all — is in the context of the business processes, decisions, customer experiences, and competitive differentiators it can enable. 

For example, a customer record doesn’t have value unless you can sell, market, or service that customer. So for each customer record, many customer intelligence analysts calculate lifetime value scores, the potential share of wallet available, the customer’s propensity to buy certain products and services, and even the cost of servicing the customer. But that doesn’t put a value on the customer record itself: It places the value based on the sales, marketing, and service processes the data supports. And that’s where the data value should live: in the consuming processes.

That’s why my process data management approach recommends that all data management, data governance, data quality, and MDM efforts be put into context of the most critical business processes that consume and depend upon trusted data. The alternative is attempting to boil the ocean and trying to solve Customer, Product, or Financial data for all processes and decisions across the whole organization — too big of an effort, destined to fail before it starts.

Some organizations of course do place an accounting value for things such as acquired customer lists under “intangible assets” on their balance sheets. But the value of those customer lists are most commonly defined based on the value of the projected cash flow and revenue based on complex models with plenty of caveats and assumptions that act as necessary filler in quarterly and annual securities filings. So of course the customer records on the list are not what is being valued — it’s the future anticipated revenue net the future anticipated account maintenance costs that define the value.

 And that’s where we should all be focusing our energies: on the business-value-generating processes consuming the data, not the data itself.

 Would love to hear any thoughts on use cases I may have missed where calculating a value on data itself is actually delivering value!

Comments

big data boiling in the ocean

Rob,

Afraid there isn't much here I can agree with. Data contains and reflects increasingly everything on the planet, including all assets--human, financial, fixed, fluid -- tangible and increasingly intangible. In fact transferring very real capital from your bank account to mine is only a matter of data, and so I suspect that you would want to place a value on it. Otherwise, let's talk ASAP! In fact I find in my own research in attempting to overcome some of the most stubborn problems in the enterprise, including info overload, poor innovation, and systemic failure-- that a commonality is failing to value data fairly often acts as a strong disincentive to the knowledge workers who the system depends on. That said, not all data should have value--only the original owner can make that judgement, and the would be buyer of course, or regulators in some cases I suppose.

I also have trouble with the boil the ocean metaphor, which is used too often these days to justify all kinds of protectionist policies in the enterprise. You can't have it both ways in the enterprise-- either you have data silos or you don't, and I argue that increasingly the world cannot afford them, albeit in highly secure formats in most situations. I can't begin to share how many of the world's most serious crises and biggest missed opportunities could have been reversed with properly structured data across the enterprise. In the U.S. Army for example-- a publicly available case, they are just now combining four sep. systems that were simply not interoperable. With the evolution of technology over the past decade, as we are beginning to see with technologies like Watson, HANA, and in our Kyield and others-- boiling the ocean is both increasingly doable and preferable to fatal toxicity in essential ecosystems.

The question on both issues is of course how it's done, which is anything but trivial and the ultimate differentiator moving forward.

Mark Montgomery
Founder
Kyield

I didn't say data wasn't valuable...

Of course data is valuable, it's the reason i've been happily employed for 20 years. I said it's not worth the effort trying to assign a monetary value to data records themselves. In your example of my bank account data, it's not my bank account information (account number, pin, balances, etc) that have the value - it's what you can do with the data that would be damaging. This account information could allow allow you to perform business processes like cash withdrawals and wire transfers, for example. These are the business processes that define the value - and risk - associated with this data. If these processes didn't exist, I would happily tweet all of my account information for all to see - because it would be worthless.

Regarding boiling the ocean, sure it's an overused cliche, but I've never seen one successful data governance, data quality or master data management initiative that didn't learn how to focus on critical few. I have not heard any of the Big Data evangelists address how data governance and data quality will be effectively defined and managed.

If data is valuable, then it should be assigned valuation

The subject line says it all -- a good topic to draw commentary, but no need to discuss. If something has value, and of course data does, then it must be valued and that value must be exchanged. That it isn't in much of our economy is to a large degree why it isn't functioning well.

One problem I see with your comment on boiling the ocean is that it's very difficult to be a big data evangelist and also be an objective solution architect--evangelists lack credibility for all the right reasons, which is what it requires to overcome these great challenges. The answers are out there for those willing to look. My paper unleash the innovation within is not a bad place to begin--tens of thousands of large orgs have-- somewhat dated now, but the research is rock solid, based on 15 years of ground breaking work and testing. It is what it is, not what anyone would necessarily want it to be-- the answers I found are not necessarily aligned with the interests of a small independent lab within a global economy that has a strong bias for incumbents, for example.

I remember doing Business

I remember doing Business Process Re-engineering back in the early nineties, and at that time it seemed it was all about the process and not the data. In recent years, I've seen the opposite trend, where the focus has been on the data solely.

I think they are really interwoven, where the effectiveness and efficiency of the process rely in large part on having trusted data, and conversely I agree that data without a business context lacks business value.

I'm working with a number of customers who have taken a top-down approach, starting with their business drivers such as cost reductions, then looking for ways to improve their global business processes such as procure-to-pay, and only then developed a data strategy that aligned with meeting the needs of the process. Moreover, these customers developed a data governance process also in line with governing the data across systems and Lines of Business that the process crossed. This is where I'm seeing the greatest business ROI.

Thanks Oliver

Yes, I've seen this top-down meets bottom-up approach works well too and think it's a great way to kickoff and structure a data program. Thanks for the comment.

All in!

Focus on ROI of your business and then the data quality for achieve that. Many customers ask: We want 100% data quality. And my answer: That can be a difficult task, but more importantly you need to focus on the data quality that will impact your business ROI. Data Quality without a positive business impact is a lost cause.

Couldn't agree more!

I loved your comment that customers ask for 100% data quality - because I agree that's an indicator that these customers don't yet understand the data quality competency and helps to gauge their maturity. I always joke that I can eliminate all duplicate customers by deduping all into a single record. There may be a risk of false positive of course, but all customers will be deduped! ;-)
Thanks for sharing your comments

I'm glad you wrote this,

I'm glad you wrote this, Rob.
I agree that putting monetary value on data is often hard and unproductive. There're two perspectives on data. One, data is like gold bars. Assets with intrinsic value that you can put in a vault. The other, data is like employees, who has value only if the employee does something for the company (often as a part of a business process). I think most forms of data fall under the later category. This was in my latest blog.
But I do want to make two contrarian points:
1. Is it a useful exercise to understand the financial impact if a data field is bad? Say, what would happen if a customer's billing address is wrong? So, you're not valuing data based on some intrinsic value, but based on the cost to the business if there's something wrong with it. This way, you can put some value on good data. This is useful when prioritizing what data quality problems to fix first. Of course, you can do this only with some types of data.
2. If a dataset is bought and sold, then you can definitely put a monetary value on it. Physician data for pharma. Instrument reference data for financial services, etc. The market puts a value on it. If you take an expansive view of data, then music, movies, books, etc. all have clear market value.
My bottom line is, not all data is created equal. The label "data" is way too broad.

Nice examples

Thanks Winston, and I agree with your point 2 regarding "market value of data". Of course the market forces will likely set that value based on how the data will be applied, but that's a given - that is an example where organizations (aka sellers or acquirors of data) has a business rationale for placing a face value on data.
Regarding your point 1, I agree as well but don't view that as putting an instrinsic value on data. Defining financial impact of bad data (hence value of good data) is in context of consuming processes - which does demonstrate how to value data in the right way (in my opinion)!

Kudos to the "not all data is created equal" line.

Thanks, Chen

Point #1 one is especially poignant when companies rely on Mailing. Things like Invoices, marketing materials, Media or other mail can be Returned to Sender when sent to Vacant, Moved, Changed, Unformatted for the specific country, or just plain Undeliverable Addresses. For example just missing the Suite number in a high rise will invalidate the address' deliver-ability. At First Class rates a single large mailing can cost thousands of dollars in Returned Mail alone if not properly validated for accuracy each and every time a mailing is performed. Not to mention duplicated data ie: sending the mail pieces twice, or Householded sending them to multiple people at the same address ie: father and son.

In fact this type of mailing and contact data can be accurately scored and and exact amount calculated of it's worth. This has nothing to do with the value of what was mailed just the cost of the mailing itself.

Begin with the decision in mind

Great post Rob
In my work with clients I focus on decisions - what are the decision points in business processes where data can be applied, what historical data exists that might guide the decision, what could we predict about the future based on our historical data that might point us to better decisions, what does "better decision" mean in terms of our overall corporate objectives and measures etc.
When you do this you quickly see that some data has value in one way to one decision and in a completely different way to another. And that while IT departments are good at estimating the COSTS of managing data they are not so good when it comes to estimating the VALUE of that data to decision-making or, for that matter, the value of those decisions.
Knowing what your decisions are, understanding what metrics/KPIs/objectives are impacted by which decisions and being clear what data (current and historical) is valuable to making a good decision is far more useful than a monetary value for data.
James

Thanks James - I like the

Thanks James - I like the focus on decision points. Not all data consumtion is decision-centric of course - data often supports interactions and experiences (although arguably those inflection points are micro decisions as well perhaps) - but I do think business stakeholders understand the business impact of decisions and can better partner with IT to drive KPIs and business value metrics to place a value on the impact good (or bad) data has on the business.
Thanks again.

Hi James, Great point about

Hi James,

Great point about the value of the decision flowing to determining the value of the data. We have a little LinkedIn think tank started on the topic. Pls join: http://www.linkedin.com/groups?home=&gid=4069646. Looking forward to your further thoughts. Also see www.centerforinfonomics.org for resources/blog.

Cheers,
Doug

The value of data is reflected in the value of decisions made

Rob, I’m with you on avoiding the value problem when you are talking about trying to assess the future value of data.

But the picture changes somewhat when you use the word cost. The mere existence of an attribute has a high cost – the cost of adding it to the database, including it in analysis for every project that ever uses the database, maintaining it, building it into transactions, etc and then the external cost of acquiring and validating the values in the attribute over time. In this sense, the data at least has the value represented by its historical cost.

When it comes to deriving value from the data, I agree with James that the primary driver is decisioning. It is fundamental that value in an organization originates in decisions made by the organization (this article discusses why: http://bit.ly/7S9RSx) – if you don’t have the data, decisions can’t be made (by definition), and if decisions can’t be made, the organization cannot create value. So there is also an ‘opportunity cost’ associated with non-existent or bad data.

‘The data in and of itself has no value’ – but I think it has a high cost, both historical and future, and both real and contingent.

Regards,

Mark

Nice job flipping the conversation from value to cost

I agree that there are many costs inherent with data management as you discussed - and of course much of the value conversation we've all been discussing in this blog conversation is arguably focused on 'opportunity cost' of not doing something with the data. I really liked Winston's parallel that data as an asset is similar to employee as an asset. And in your example, employees have inherent costs associated with them well before they deliver any value to the organization (recruiting, onboarding, benefits, etc). But employees at firms that manage performance well must clearly demonstrate some link to business value or they'll be managed out. The same should be true for data that just has administrative costs or capture and mainteance, but no recognized business value.
Thanks for your comments!

If I had a million customer records . . .

. . . would I have a million dollars?

Excellent thought and debate provoking blog post, Rob.

The strategy around how the business value of data is quantified and qualified and how we make the pitch to get executive management to invest in data has to emphasize that we are talking about a strategic business initiative, so I agree with you that (talking about) data in and of itself has no value.

All too often the problem when we talk about data is exactly that: we talk about data.

Borrowing the words of Jill Dyché, we need to bring the conversation back to a "business" conversation, and not just another "data" conversation. A true data strategy is about starting with a business strategy and then mapping data to it.

Best Regards,

Jim

Check out tons of other great comments...

My Forrester blog is also syndicated on Information-Management.com and there are some fantastic comments on that site as well on this great debate. You can check those comments out here: http://www.information-management.com/blogs/data_management_business_int...

Hi Rob, A good post and I

Hi Rob,

A good post and I agree that sometimes these information valuation exercises can get way too theoretical. Depending on your business model, however, I think it can be quite useful.

I work for an organisation that makes investments in early stage carbon capture and storage projects. In exchange they share knowledge with us that we disseminate to the broader community to accelerate CCS projects (e.g. http://www.globalccsinstitute.com/resources/projects/tenaska-trailblazer...). We put a value on our "information investments" by mapping them to the key decisions required to drive projects forward from which we then derive dollars to help guide investments.

Also, couldn't you make the case that research organisations like Forrester put significant value on data and information? That is, after all, a big part of what you sell.

Thanks for starting the debate - its a good one!

Sean

Rob, I completely agree. Also

Rob, I completely agree. Also Jim Harris proposes in his whitepaper the angle of business insight, very interesting. See
http://www.dashboardinsight.com/articles/business-verticals/the-role-of-....

information as an intangible asset?

Hi Marcel, A question: If an accountants and organizations can attribute definitive financial value to brand, copyrights and patents, then why not to information? An asset is something that produces "probable future value." Shouldn't information qualify as a class of intangible asset?

-Doug
www.centerforinfonomics.org
Center for Infonomics group on LinkedIn

Value of Data Records

The value of data storage is the availability for support of decisions and campaigns compared to lack of availability. Obviously the more exclusive and recent, the more vauable the competitive value obtained by its usage. Value of data is just like the inventory of any other high grade raw material.

The reverse is also true...

Hi Rob,

Your post highlights the historic systems divide between data and the business activities (processes/decisions) that create and consume them.

I agree that its hard to put a value on discrete datum or collections of data, but it's clearly an asset with intangible value otherwise why would we collect and store it, it's an assumed raw material (e.g. no oxygen, no human respiration process).

You could argue that processes and decisions are, if not without value, significantly less valuable without the data that form their enterprise business context (not just their execution context). Informed process/decisions requires a different way of thinking as static process models promote standardization/consistency, not interaction-specific context.

We've done some work in this area to close the loop between process and data. If you get a chance, please see http://slidesha.re/g7Dprj and http://scr.bi/hcOyFh.

Thanks for raising the issue.
Dave

Terribly flawed logic about assets

Rob, Sorry but your logic about info not being an asset because "data in and of itself has no value! The only value data/information has to offer...is in the context of the business processes, decisions, customer experiences, and competitive differentiators it can enable," is drastically flawed.

Applying your reasoning to recognized balance sheet assets like physical plant, raw materials, brand, patents, copyrights would leave them all valueless as well.

An asset according to US and international accounting standards boards is something that generates probabilistic future economic value. It's easy to see how the application of a record of data can contribute to generating revenue. Companies do it every day. Quantifying an intangible asset's economic value can be done a variety of ways: 1) determining its contribution to an income stream, 2) computing its cost (to obtain or if lost), 3) its value on an open market. There's no reason these can't be used on a unit or portfolio of information as well. I'd argue (and demonstrate because I've done it) that it's easier done for info than valuing a copyright or brand.

The real question is why is the accounting profession 50 years behind?

-Doug
www.centerforinfonomics.org

Have to disagree somewhat

Thanks for your comment Doug. I don't disagree, and am no expert in GAAP and other accounting practices. But the reason I believe information to be somewhat different is that the other assets you mentioned "physical plant, raw materials, brand, patents, copyrights" all have an intrinsic market value in and of themselves in that they can be valued based on what others are willing to pay for it. But in valuing data, aside from the market cost of, say, selling or renting your customer data to another entity, the true ROI value of data is significantly higher based on what it could do to impact top and bottom line within your organization.

Thanks for joining the conversation!

Thanks for the quick reply

Thanks for the quick reply Rob. Wouldn't you agree that brand has less of an intrinsic market value than data? There's no open marketplace for brand, yet accountants go through a valuation exercise to quantify it on the balance sheet.

And patents, copyrights and all manner of other assets are valued *before* they're ever sold...based on their estimated contribution to the bottom line. Why can't or shouldn't organizations do this with their information assets? Doing so would 1) encourage them to manage info as a full lifecycle asset, 2) help them validate strategies and expenses re info management & access, and 3) add to corporate premiums which are drastically out of whack because info isn't accounted for.

Look at a company like one of the credit bureaus. They're main asset is data. That's all they buy, manufacture and sell. But it's not represented anywhere on the balance sheet. Ridiculous isn't it? And because of this lack of accounting for information organizations are entirely unable to insure it.

Unique valuation for boatloads of data?

Hi Doug
A large company may have anywhere from a handful to even a few hundred brands. A large R&D firm may even have a few thousand patents and/or copyrights. It's not easy, but it's definitely 'doable' for accountants to put unique valuations on each of these. And I imagine one of the reasons I believe this valuation on brands and patents,etc is done on the balance sheet is because it's feasible that the company can sell the rights to the brand or patent without impacting the overall business. (again, i'm not an accounting expert and am sure there are many other reasons as well). I can't imagine a scenario where a company will sell it's entire customer list yet remain in business.

But even small enterprises may have in the order of magnitude hundreds of thousands to millions of customers and anywhere from tens to hundreds of thousands of unique products/SKUs. Each data record has a unique value to the busienss, so how can these companies be expected to put any relevant and useful valuation on each piece of data? And if they could - once again, where will that get them? Unless they are actually trying to sell this data (like D&B, Experian, Reuters, Acxiom, etc do as content providers) where market demand can help set the price, what's the value of putting a value on data that's primarily used to support customer engagement, front-office and back-office processes, etc?

As my initial post implied, I'm not necessarily saying it's impossible to put a value on data. I'm just wondering why anyone would bother with that exercise when by far the largest business impact data will ever have is on the business processes and decisions it enables. That was the primary point of my post.

Hello Rob,A very interesting

Hello Rob, I posted this comment before on "information-management.com" but would like to see if there is any movement in the discussion. One reason is that I indeed see that some organizations are starting to apply their existing Asset Management process to Information and data. I'm really interested in th "why". One case that comes to mind is a large net provider of energy. The data they are trying to manage is about the operation of their business, measuring the power grid and reacting to certain fluctuations. The main business issue related to this data is risk. So risk is a big factor in the calculation of the value for this data and information.

My comments to your Blog:
A very interesting viewpoint on a hot topic. Because I don't know you personally very well, I'm not sure how much irony is involved. Hence this comment larded with a few questions.
I'm assuming the reason to calculate monetary value of things comes from the need to communicate on those things in a way we can get a good feeling on the relative contribution to business goals as well as cost, and for reasons of accountability and reporting to share/stakeholder.
On the "Information Management" site one commenter addressed the similarity of data with (other) assets of a business, or organization. Do you see data as an asset of the organization?
If so, would Asset Management be a good department to also manage the data? Would you agree the same rules should be applied to all asset types when it comes to defining the value for the organization? And I'm not an accountant, but I think assets are valued by either putting an amount on a balance sheet and depreciating the once established value (of purchase) on a timely basis to be able to, in the end, determine the total value of the organization. Some people want to know ;-).
I understand the main point you're making is that, however valuable data is, it's value is not related to the rows of data in the database (or e-mails) but to the usefulness to, and readiness for, the processes that add value to the organization. Then I would say that the process is equally important as is the data that is consumed by it or is used to create- and / or add value to it. Bank clerks wouldn't be able to do their jobs if they didn't know how (process), no more than if they didn't know with what (data).
Some processes exist just to add value to the data, you'll find them in the marketing and R&D departments. Some processes do not directly add value by nature, e.g. administration and accounting. Does that mean that the data involved in these processes have less (monetary) value?
In a reply on this site I read: "it's not worth the effort trying to assign a monetary value to data records themselves." I have a feeling I agree with you but on what unit of information should we then put a monetary value? And if we decide to not put a monetary value on it at all, what type of value should we put on it in order to have meaningful communication on the subject?
We could decide not to see data as a corporate asset but as something like the combined knowledge and skills of people (human resources) or like processes. Business processes are based on those resources, skills and knowledge and are just as important to the business as is the data. Why choose the business process that adds value as a leading concept and not "the data that adds value, given the right process"?
When I think of it, in trying to establish the value of 'contribuants' to business value, the characteristics of contribuant 'human resource' come closest to those of 'data' (agreeing with Winston Chen on Forrester Blog). For both it's difficult to establish amount of contribution to the strategic and tactical goals of the business and usually only the cost of both is found in the financial reports, not the value.
I read: "the value of data should live in the consuming processes". Then for 'the records' of accounting and administration, how can we achieve a value and a unit of value that we can all relate to?
The last sentence of the blog reads: "Would love to hear any thoughts on use cases I may have missed where calculating a value on data itself is actually delivering value!" I can't think of one. But I'm wondering, are you equally interested in cases where the subject of calculation is not data but another asset type or resource?
I think in short my comment would be: A very sharp and good argument to -not- put a monetary value on data, but I'm missing the alternatives and the ratio behind them, assuming we need people, processes and information to deliver value, we need to communicate about the relative value of these three and accountability is important.

Info has value before it's used

Rob, You write "a customer record doesn’t have value unless you can sell, market, or service that customer." I wonder if you believe your batteries have energy before they're used to power something? Generating benefit and having value are very different things.

And as for your last argument about selling customer lists, companies do this all the time. Well, mostly they license them. For what price? A price that implies, er, their market value. Forget about customers lists tho....what about transaction data? When you give a grocer your loyalty card, you're trading information about you and your buying habits for free food. They've placed a specific value on that data; a value greater than their cost of the free groceries you get.

Data is both straightforward and increasingly important to value quantitatively for organizations smart about optimizing information-related investments. The fact that it's not on the balance sheet is immaterial and a matter for lagging accounting standards.

-Doug Laney, VP Research, Gartner