Making the Business Case for Taxonomy

Seth Maislin, Senior Consulting Taxonomist | seth.maislin@earley.com | Earley & Associates, Inc.

Research librarians know well the value of their work: improving retrieval of documents and other assets, encouraging business units and roles to share ideas and coordinate processes, integrating multiple sources of information, not losing assets, not making errors in intelligence, pushing for greater efficiencies, not letting your competition get the better of you. And yet there is just no way to say the word “taxonomy” and inspire passion in a CEO. Or convincingly provide a cost-benefit estimate.

After all the taxidermy jokes, there still remain two major obstacles to making a business case for taxonomy and controlled vocabularies (CVs):

• Translate trivial-sounding, content-level processes into the language of business strategy.
• Translate proposed taxonomy initiatives into realistic dollar amounts—and afterwards, prove that you were right.

Librarians and taxonomists need an approach for speaking with their managers about the value-add of taxonomy. We need to speak using the same business and strategic language that managers use to make decisions.

Linking Taxonomy and Organizational Strategy

Taxonomy efforts happen at a words level: preferred and nonpreferred terms, asset naming, conceptual representations, and content relationships. Organizational strategy, however, gets practiced at a much higher plane than semantics; organizational strategists concern themselves with revenue growth, competitive edge, and brand recognition. Between the two are additional management levels: business processes, and business unit objectives (see Figure 1).

Figure 1. Four levels of organizational thinking and activity.

Taxonomy lives and breathes at the level of content processes, where the content repositories and archives reside. This is the world of the taxonomy worker, who concerns himself with search and keywording and IA; this is also the world of IT, which focuses on databases, interoperability, and reliability. This is a world of words and lines and records.

Content exists to support business processes: customer support, web sales, customer acquisition, document aggregation, and basic analytics. Here live team leaders, who are focused on all things easily measured: time expenditures, customers, transactions, clicks and views.

Business processes aggregate into business unit objectives, which are all about outcomes that are too complex to measure easily: satisfaction, branding, opportunities, policy. Managers may be less interested in money than in the possibilities that having more money can create.

Finally, at the highest level, company leaders concern themselves with the organizational strategy, which is usually about growth. The development of new offerings and the investment into new markets must operate in tandem, so that the company is investing its resources as profitably as possible.

So how we do translate a seemingly mundane discussion about terminology, CVs, hierarchies, and content models into one about corporate growth? The key is in our ability to anticipate the business gains at each level up the pyramid, and then to articulate those gains.

Consider a research librarian who researches lawn mowers as part of her job, which is to help her company observe human behavior. She recognizes numerous kinds of documents just about lawn mowers, with only half coming from lawn mower companies: product operation documents, spec sheets, pricing sheets, marketing materials, consumer opinions, video blogs, environmental impact statements, market performance analyses, mower hacks, accident reports, videos, political cartoons, and correlation studies with everything from rainfall to household income. She believes identifying and categorizing all of these lawn mower documents will be useful to the business, leading ultimately to huge payoffs.

Not surprisingly, the CEO thinks she’s a little bit wacko, or at least naïve to business practicalities. Seriously, keyword search should be good enough. It’s not like we’re in the lawn mower business. But the librarian in this story has a deeper understanding than the CEO suspects:

1. Content level. New mower-related documents are appearing daily, from numerous sources and feeds. Some of these documents aren’t even about mowers as much as about golf courses, noise pollution, and portable fuel cans. The librarian wants to institute a tagging policy into the workflow, building a CV and putting the right keywords in place to validate the CEO’s “keyword search” idea. Thanks to this new approach, researchers will sift data faster and identify critical and viral ideas sooner, including some she's likely missing today. Bogus opinions and manufacturer bias are swiftly discarded—or emphasized, if that’s what someone wants.
2. Process level. Thanks to some quality tagging, librarians can provide up-to-the-minute answers (about lawn mowers) in less time and with greater accuracy. Writers can trust the results; presenters can trust the analysis they’ve asked for. Delays are gone, costs drop, and the wasteful back-and-forth game has been replaced with intelligent dialog and an opportunity for deeper analysis. The backlog is reduced. Knowing about lawn mowers morphs into intelligence gathering about gardening, about summer, about motors, all of which depend (at least in part) on knowing about lawn mowers.
3. Business level. The business unit starts to recognize a consumer upswing: clients wait less, customers pay more, and there are more customers out there (including internal users) than before. Empowerment leads to stronger engagement, so more people are visiting the website and existing companies further open their doors to better content. Third-party ratings improve, as does loyalty. Content providers are starting to adopt the librarian’s tagging approach; new standards are in their infancy. Vendors and partners are upping the stakes on collaborative initiatives, sharing more (and more often, now that there’s no backlog), and incentive programs are being born. And when it matters, the company can guarantee compliance, with actual guarantees.
4. Organizational level. In addition to increases in sales and market share, the company’s well-organized portfolio (of intelligence) and effective self-awareness opens the door to acquisition and new vendor partnerships. The company has become a stronger influence in the marketplace.

This may sound too good to be true, but it’s definitely not false. Remember that organizing content about lawn mowers was just a piece of the puzzle, a single tangible example for a pilot program. The logic is sound; taxonomy is a proven tool for bettering sales, customer loyalty, efficiency, compliance, and market share, all while reducing costs. No, what’s missing from this analysis are actual numbers. Using a vocabulary for document types (about lawn mowers!) to improve the likelihood of acquisition? What’s the reality?

Concretizing the Value of Taxonomy to Managers

Some decision-makers think in terms of ROI, or return on investment. ROI is a simple ratio between the cost of an initiative and the savings or revenues that result from that initiative. For example, if implementing a CV costs $10k but saves you $25k, then your ROI is 2.5. When there are clear cause-and-effect outcomes to a new project, ROI is a fantastic and even illustrative way to quantify value. Unfortunately, many people see ROI estimates as good in theory but requiring leaps of faith. For this reason, ROI really only works when it’s not an estimate, such as after a project is complete. Yet, librarians need a tool that works before the start of a project.

One solution is to use a pilot project, something inexpensive and small in scope. Find your “lawn mowers,” the low-hanging fruit that’s contained, tangible, and easy. Get a baseline value before the pilot, measure results after the pilot, and then calculate a honest-to-goodness real-life ROI. For example, how does query response time, measured in librarian-minutes, change once your taxonomy in place? Or perform a search audit and measure the quality of the top 20 search results (i.e., accuracy and relevance) before and after taxonomy application? When your customers are internal to the company, you can count certain kinds of transactions. An increase in requests or other transactions can be an indirect way to measure satisfaction or confidence. Measure knowledge reuse, perhaps by measuring download frequency, checkout length, or number of items printed. You can measure satisfaction now and later by asking your customers to provide scores and also explain their thinking behind them. Put all of these numbers into a spreadsheet; draw pretty charts. Get your managers to scratch their heads and think, “You did all that just by changing the words?” Search and SEO projects are ideal in this regard, because you can focus on just a few terms, and measurements of accuracy, efficiency, comprehensiveness, and rank are easily obtained. If you are clever in your keyword choices, you should have no trouble demonstrating what’s possible if someone would authorize some spending.

A completely different approach is estimating cost of doing business, or CBD. This is a way to recognize that certain expenses and investments are no-brainer preconditions to conducting business. Do you really want to be part of a company that makes your customers angry? Are you comfortable letting your competitors provide a better user experience? Do you like losing business? Do you want to be distrusted by the very people who should be engaging your services? In a nutshell, growth is essentially impossible without a proper foundation, and if your company runs on information (and whose don’t these days?), that information needs to be solid. CBD means taking a good look at the “Before” condition and adding up its costs: delays, inefficiencies, risk, mistakes. A single compliance failure can cause irreparable damage to both your finances and your reputation. So can a believed lie. If all you need is a little more oversight of your incoming metadata, why wouldn’t you do it? It’s like checking your email for viruses.

Which brings us again to the idea of translating your ideas to the top. This is a skill that requires practice; many librarians, information architects, and taxonomists can speak intelligently about schemas and synonym rings and user experience but still need help communicating to upper management. Thankfully, it’s not that hard.

After all, if anyone can trace an idea up a hierarchy, it’s people who build hierarchies.