louisrosenfeld.com logotype

Home > Bloug Archive

Jan 28, 2002: The ROI of Information Architecture

Thank to everyone who discussed the value of information architecture this past week; awesome! In under a week, we're already up to 17 comments. #17 comes from somewhere up in the remotest part of northern England. That's where Paul Nattress (I admit, I love that name), like many of us, wonders if all this information architecture goodness can be converted into actual numbers (as in dollars, euros, shekels):

Lou - can we have a resource of arguments in favour of spending money on IA please?!

Ask and ye shall receive.

The ROI of information architecture; is there really such a thing? Certainly it would make our lives much easier. But can you really measure the value of information architecture?

Keith Instone and I moderated "Measuring Information Architecture" at last year's SIGCHI, and we tried to harangue our panelists into an answer. They responded admirably, and I recommend you have a look at their brief Powerpoint presentations.

Better yet, if you have a killer idea on justifying investments in ROI, or if you just think it's a fruitless pursuit, please share your ideas here.

email this entry

Comment: Dell (Jan 30, 2002)

I enjoyed reading the panelists's positions and wish I had been able to hear the discussion in person.

To me, justifying the cost of IA better not be a fruitless effort or else we might as well volunteer our services and quit worrying about it. No company is going to pay for something (or someone) they don't believe is valuable and usually "seeing is believing". I tend to agree mostly with Gary Marchionini's approach, with a little lean toward Marti Hearst's position. Maybe all we can do is justify the pieces of IA that lend themselves to it, and the rest is nice to have, from the outside world's point of view. It's like trying to justify the value of therapy when someone else is paying for it (sorry to use that comparison again). An employer might pay for your treatment for depression, but only when you show progress towards getting out of bed and coming back to work, not just when you forgive your mother for missing your 3rd grade piano recital, or saying that you "feel better".

Here are my not-so-killer ideas of justifying investments in IA, at least for a web site:

What is the impact to the business if users can't find critical information, or perform critical tasks? As Lou puts it- where's the *pain*? To me, pain or fear of pain is the key. If the business isn't suffering or isn't afraid it will suffer, then it's hard for check-writers to "see" the value of an art/science such as IA.

Pain can be manifested in:
-Lost time/productivity (thinking Intranets)
-Bad impressions on customers, resulting in fewer sales
-Fewer widget sales
-Fewer members registering or enrolling
-More marketing materials mailed out to explain/sell/bring awareness to the things users should be finding on the web site
-More offline support required
I'm sure people can think of more to add...

If there's no apparent pain then we have to slip good IA in where we can, disguised as something else like 'interface design' or 'usability' that is already being paid for. But when lack-of-information pain arises, we can be ready with our bag of tricks to show how good IA can alleviate it.

Comment: Lou (Jan 31, 2002)

Great comments from Dell.

But it's interesting to note that while the initial entry on the value of information architecture (http://louisrosenfeld.com/home/bloug_archive/000067.html) has at least 19 comments, this entry, which tries to get at the numbers behind that value, has only two comments (including this one).

Is this an unscientific validation of the opinion that there is no strong ROI case to be made for information architecture?

(By the way: ROI = Return On Investment.)

Comment: Paul Nattress (Jan 31, 2002)

It may be the case that IA sometimes has to sneak in with usability or content editing in order to make a good case for ROI. Rather than split the three areas up, would a collective ROI proposal be a more simple, clear way to get a green light from the managing director?

All three areas can work together well, in fact, I'd say you can't have content editing and IA working effectively unless there is close collaboration.

If there's a gap in the argument for investing in IA then maybe we should play on the fact that a lot of company directors know (or care) little about the nitty-gritty of all the different diciplines involved in the production of content for their website. Think "these guys write the stuff and those guys over there make the pages" and that's the attitude I'm talking about.

Unless of course, someone can add (and quantify) a comment that says:

Spend X on IA and you'll get X amount of extra revenue.

Comment: Lou (Jan 31, 2002)

I agree with Paul. Ultimately, IA and content management are two sides of the same coin anyway. To grossly overgeneralize, IA is a snapshot of an information system, while CM is about the flow of information in, around, and out of that system. Or, put another way, IA is the spatial while CM is the temporal.

I still don't think anyone can answer Paul's last point though.

Comment: Peter Van Dijck (Jan 31, 2002)

The money spent on IA mattters only in how the system that is being built matters. ROI justifications should be for that system, not for IA. (note: this post is only relevant if you're doing IA as part of building a technological system.)

Defending spending on IA shouldn't be about coming up with an ROI for IA, just like you don't want to bother coming up with ROI for the cleaning lady, or for QA. Defending IA is defending a quality product.

What I say is:
- It is difficult to build a technology product that works really well (medium risk project).
- You need it to work really well to get that ROI that is the reason of doing it.
- To make it work really well, we need a good development process, including IA, testing, usability, the works....

If you isolate one element of the development process (say, IA) and defend that separately, it looks as if you are defending an optional widget. Something orangy they don't really need.

I say: "If you compromise on testing, you're increasing the risk factor of the project. Since this project has a medium level of risk already (because it's technology based, say) (you have done risk analysis right?), you probably don't want that. If we are having a budget problem let's talk about features and phasing, instead of compromising the quality of usefulness of the product. We don't want this project to turn into a white elephant." Similar argument for IA.

Comment: Peter (Jan 31, 2002)

Re Previous post: I guess I'm agreeing with the posts above. Should've read them before jumping in!

Comment: Dell (Jan 31, 2002)

I guess I was saying you *can* justify the cost of IA if you can show how lack-thereof will hurt the business. If we're defining IA as "findability", as in a search engine, where information/content is placed, labeling, navigation, etc. then I think you can make the case that the lack of good IA is more costly than bad IA. Maybe that's backing into the argument but in my experience, it works.

So Paul could say, "If you spend $X on improving your web site's IA you could gain $X in sales, or reduce your offline support by X people". Sorry, don't know how to make a pound sign on my keyboard :)

To quanitfy such a statement, you would have to evaluate 'X' on a case by case basis and make a pretty good guess. What we really need is a bunch of research to back up our guesses, but that's a whole other topic for discussion.

Comment: Dell (Feb 1, 2002)

Antecdotal story: This morning I was approached by a co-worker with a "Chief" in his/her title who held a "User Experience" evaluation form from a major national research firm which was to be conducted on our web site. The "C" asked, "What do you think of this eval? How will we do?" Of the 25 items, I'd say about half (12-14) had to do with IA directly, that is, the user's ability to perform tasks or to find information.

I was surprised at how many items were clearly IA related. There's so much inherent value in good IA-if we can just figure out how to quantify it, I believe businesses will pay for it.

Comment: Paul Nattress (Feb 2, 2002)

I've just been down to London for some training on a content management system. On the train, I had the opportunity to read Content Critical by Gerry McGovern and Rob Norton.

They mention information architecture as the cornerstone of presenting content on the web. Good stuff. What made me smile was the fact that they begin the book by detailing the ROI of creating content.

I thoroughly recommend that anyone who deals with web content should read this book - the arguments in favour of high quality content are worth the price of the book itself.

(I hate to take advantage of the Bloug to promote a product but this book is relevant to our discussion so I hope it's ok.)

Amazon.com link - http://www.amazon.com/exec/obidos/ASIN/027365604X/qid=1012650334/sr=1-1/ref=sr_1_11_1/002-4401370-8748022

Amazon.co.uk link - http://www.amazon.co.uk/exec/obidos/ASIN/027365604X/qid=1012650865/sr=1-1/ref=sr_sp_re/202-3691436-6704603

(Amazon need shorter links for their products! How about www.amazon.com/027365604 (with the number being the ISBN/ASIN number)?)

I agree with Peter - we're "selling" a quality product when we put forward our case for IA. Remember though, that our most technologically advanced vehicles, spacecraft and fighter jets, are made by the lowest bidder and not the bidder who can supply the best quality. In business, everything comes down to cash. If the quality plug doesn't do the job then we have to start talking about hard currency.

Comment: Jess (Feb 3, 2002)

Aside: Amazon URLs can be safely truncated like so

ROI is not always about money. What kind of return are we looking for? Political ROI (feather in the cap of project sponsor?), "branding" ROI with increased mindshare or warm fuzzy attitudes about the company? Or bottom-line profit/loss accountability?

So here's my ROI steps (though not terribly helpful for _selling_ IA right now)

1) Find out what's important for the business (business goals analysis), both formally & informally (eg the VP may not say that this project is an internal powerplay)

2) Set metrics for success - how will we measure if a project was successful? Assign dollars to outcomes, even if not directly financial. (eg return visits to the site are worth X)

3) Establish a baseline on those metrics.

4) Do the project. Measure changes in the baseline.

This won't isolate IA (thesaurus, labels, taxonomy, search) unless the project is something like "improve search on our site". But that's not a bad thing...better to sell a package deal that promotes User Experience as a business driver, and IA as one set of tools that contribute to good UX.



Comment: Chris Ford (Feb 4, 2002)

I'd tend to agree with Jess - at my place of work, we tend to talk in terms of hard and soft ROI, where 'hard' is the cash mentioned and 'soft' might take into account increases to market share, brand extension and / or increased customer retention, for example.

I also agree that working from the bottom up works well for us - our Discovery period takes into account :

Business Objectives and Imperatives

Competitor analysis
Audience analysis
Brand analysis

Funtionality strategy
Content strategy
Brand strategy

Technology assessment
Information architecture
Creative assesment

which gives us a clear understanding of both the aims and goals of any new product development and the success metrics by which we will be judged.

I'm currently trying to move our billing method from Time & Materials to a value-based charge : ignoring that the fact that this is a fairly large cultural change for any organision, we have had some fairly heated discussion over how best to progress. FWIW, the sucess metrics approach has met with most support internally.

I'll be interested to see how this discussion progresses.

Chris Ford

Comment: sean patrick coon (Feb 7, 2002)

I tend to agree with Peter in his stance on IA. Leave a la carte sales to a waiter.

But in a development process debate with a client, centered around ROI, Information Architecture will most likely become the article of focus in the clients eye, so we must be able to defend it.

To expand on some of Jess and Chris' thoughts on branding... it seems that we too often try to defend the value of an IA from within the forest itself -- as strict deliverables or responsibilities. Why not try from above and take the view of the overall result and work our way to the specifics?

I feel that an extremely important element of a proficient (equate *valuable*) IA has been left out of the equation -- the ability to help derive brand elevation.

Where there is no arguement that IA's are primarily responsible for information clustering, heirarchy, retrieval, recall, navigation and design (all to varying degrees depending on the type of project), brand elevation is the synthesis of all IA sub-disciplines.

It is the *intangible* aspect of the role that can clarify the value of an IA. Sell this to a client first, and then work your way to the specifics of each job.

An example of brand elevation in the reverse...
Imagine Amazon as an e-commerce solution *with* it's present direction, but *without*:

(a) the seemless architectural solutions to such interwoven functionality, products and marketing schemes that it presently serves

(b) the interface design needed for the average consumer to interact within such a complex environment and

(c) the narrative flow which exudes from the nomenclature, indexing and copywriting.

The core awareness of the brand would probably equate to an "average/less than average experience." Not only would Amazon have a much smaller share of the market, but it wouldn't scream "ahead of the curve" or "revolutionary concept" or "user-centric" in the minds of consumers either. It would probably equate to a functional experience, but not a pleasant one to say the least. These are essential traits to the growth of the brand.

The brand itself would suffer from the absence of the expert coalescence of each IA sub-discipline (clustering, heirarchy, retrieval, recall, navigation, design, internal collaboration, etc.). If this can be stated as true, then the underlying truth for developing a solid information architecture is ultimately found through the development of a complete *brand percept* -- a true representation of the *clients* perception of the brand.

The client's inclination is for the user to:

(a) be able to use their product and

(b) have a positive "user experience."

But at the end of each quarter, money is made and lost -- possibly more -- by how the user *perceives* the brand. That "soft" ROI is true whether it's an off-the-shelf application or an intranet project.

Time is money.

Good topic!

Comment: Lou (Feb 7, 2002)

Sean, good comments, but one question: does brand elevation pertain to intranets? I think that most of your comments do, but I'm not sure I can get on board with the statement that "brand elevation is the synthesis of all IA sub-disciplines."


Comment: sean patrick coon (Feb 7, 2002)

Lou, would it be fair to say that, in terms of an intranet project, (1) tangible success metrics for employee interaction *and* (2) the project sponsor's political positioning, are primary internal success factors? Because if that can be stated as true, then:

(a) top notch Information Architecture (information clustering, hierarchical development, retrieval, recall, navigation, design, brand application, etc)


(b) high end collaboration, quality results from remaining development disciplines, and seemless client side/third party involvement


(c) success of project -- (1) meeting the success metrics for employee interaction *and* (2) the raising of the sponsor's internal political positioning.

Does not (c) ultimately create a varying degree of brand elevation?

Brand elevation could be derived from the direct result of better employee interaction with a knowledge center -- easier to find/use information --> time saved is money saved --> higher quality of work --> company prospers --> client's percept of their brand becomes more salient.

And conversely, such brand elevation makes the project sponsor a star. Yes, I think you could say that it pertains to intranets. It's somewhat *softer* than software or web site development, but applicable.

But I would agree with you that the "synthesis of all IA sub-disciplines (a)" *cannot* be explicitly labeled brand elevation. A major element? Yes. Sole determinator of it's success? No.

We walk this high wire act with another IA distinguisher... the vaunted *user experience*

We (across the boards, not just IA's) often state that IA's *own* the user experience -- leaving the remainder of the development team (whom can be heuristically evaluated as well) to the side.

There is a similar correlation here: Our work contributes greatly to the user experience, just as our work partially determines the level of brand elevation. Why not sell it as such?

Two bullets are better than one.

Comment: Lou (Feb 7, 2002)

Sean, you make great points: brand definitely has value internally. And that internal value indirectly has a positive impact on external ("traditional") brand. Thanks for the clarification!

Add a Comment:



URL (optional, but must include http://)

Required: Name, email, and comment.
Want to mention a linked URL? Include http:// before the address.
Want to include bold or italics? Sorry; just use *asterisks* instead.

DAYENU ); } else { // so comments are closed on this entry... print(<<< I_SAID_DAYENU
Comments are now closed for this entry.

Comment spam has forced me to close comment functionality for older entries. However, if you have something vital to add concerning this entry (or its associated comments), please email your sage insights to me (lou [at] louisrosenfeld dot com). I'll make sure your comments are added to the conversation. Sorry for the inconvenience.