Insights

(Why) do I need a Taxonomy?

Technical Director - Digital
Valtech

January 22, 2018

"...the process of designing, refining, and user-testing their proposed taxonomy is far from finished, and will it ever be?

Recently I’ve been involved with a project kick-off where an early topic of discussion was taxonomy and ontology. In this case, the client already has made some headway. Being in the medical services delivery arena, they have a subject matter area where the taxonomy ground has been covered to an extent since the beginnings of the practise of medicine (doctors love classifying things). Nonetheless, the process of designing, refining, and user-testing their proposed taxonomy is far from finished, and will it ever be?

As we go through the process of designing, applying, and using taxonomy metadata applied to web content, we inevitably discover the value it carries (or the confusion it causes) – and so we learn more and we begin to seek to encode further meaning into the relationships within our classification system. Thus, we enter the realm of ontology.

This client fully gets the value of these activities, they’re bought-in and investing genuine effort in the process. You could almost say they are enjoying the academic nature of the exercise as well as knowing that it has tangible value in matching up content they have with user needs to deliver better digital experiences.

This got me thinking, that in a way it’s understandable that some customers don’t seem very interested in taxonomy. For some it’s hard to see past that first impression, that it’s a dry, academic exercise with no practical, real world application. You can almost see the question being framed in their minds “how does this contribute to bottom (or even top) line?”.

I’ve also been reading, the handy little volume from Tony Byrne and Jarrod Gringas “ The Right Way to Select Technology: The Real Story on Finding the Best Fit”, published by Rosenfeld Media. In the very first section – on building business case – they explore the different ways of quantifying business value to technology acquisitions. They explain that the classic ROI is just one way, and that others exist, including the cost of doing business.

Joining these thoughts together, I realised that I don’t believe digital activities are any different to digital technology purchases. There’s always a cost to doing things, in the case of activities (like building a taxonomy) this is a man-hours cost. If you’re spending money, know why you’re doing it; so what’s the business case, in the simplest possible terms, for building a taxonomy?

Caveat – you may well not need to build one, it’s quite possible that a perfectly good classification system for your business’ content already exists, so don’t reinvent the wheel, but also don’t miss opportunities to gain competitive advantage by having the best experience possible.

Simply put – this is first and foremost about visitors finding your content/product. When I say visitors, I include robot visitors especially search engine crawlers, but think beyond that too, all the way to buzzword heaven – voice UI. Everyone trawling your site, with eyeballs and a browser, with assistive technology, a search engine crawler, a voice UI bot, all of them need one thing: to understand your content. What is it about, and what is related to it?

We know non-human visitors are still and always at least pretty interested in metadata, but it’s becoming increasingly important to us humans too. Product recommendation engines in ecommerce, and frankly every faceted/filtered search rely on this too. Think about the value good facets and filters add to your searches, and the frustration of bad ones.

Some of our first online attempts at classifying our content were simple and obvious – our site nav was how we organised and presented what was what and what went with what. As often as not, in the bad old days, our site nav looked like a one-for-one port of the business org structure: “here is department A, here is department B,…”. We’ve thankfully moved on from that, so why would we assume we can invest any less effort, or reap any less value, from further classifying our content with taxonomy?

Our goal is enhanced understanding for each and every visitor. Understanding what content we have, how to find what they want with the least friction possible, and how to discover more stuff that is likely to be relevant to them.

If that hasn’t convinced you, here’s another (still) hot topic: personalisation.

We’re all on board with understanding our visitors right? Knowing who they are, what they are interested in, etc.. Well, if you know who your visitor is, how are you going to give them the right content if you haven’t classified that too?

So, finally, if you’re struggling with, e.g. getting the best out of your CMS, and dynamic personalisation, or engagement automation, or machine learning content recommendation engines is why you are scratching your head, then it’s not all about the software, and we can help.

You don’t just need great tools, you don’t just need a personalisation capability, you need to thoroughly understand your content, and you need to tag it up with organised metadata (quite possibly a taxonomy) so that every visitor (human and machine) can understand it too.