Ten years ago in a pitch situation, I had to settle for educated guesses on what the data might look like. However, if you manage your own marketing website, you have something more firm than educated guesses; you have data. Let’s take a look at what kinds of data work the best and how you can use them.
- Site paths: literal tracks in the snow
Any basic web analytics package, including Google analytics, will record the most common site paths that visitors take on your website. For clues, take a look at the most common paths from the home page and see where they go three or four levels down. If a substantial number of visitors end up on a page not directly linked to the home page, consider surfacing that page a bit. If visitors have gone to the trouble to find something deep inside your site, you can only imagine how many others wanted that same information and gave up.
Conversely, look for the links from the home page that draw the least traffic. While some links demand a home page presence out of custom or necessity (contact us, investors, etc.), others may simply merit a step down. Deciding to take links off the home page depends to a certain degree on brand decisions. However, since less usually means more in web design, take the initiative and start cutting.
- Keywords: what visitors expect
Organic search keywords provide a free version of brand association research. Back in ye old WATS (Wide Area Telephone Service) days, brand marketers fielded expensive phone surveys that often asked consumers what terms came to mind when they thought of a brand. Now, people do this unconsciously when they look for your brand via Google.
Naturally, brand and product names will often come up high in keyword sources. At the very least, marketers should make sure that their home pages represent the most commonly-searched brand names (including sub-brands) and product categories. Visitors expect to find what they want quickly.
However, marketers should also look at leading terms that do not fall into brand or product terms. For instance, do keywords that describe customer types come up (e.g. “women’s jeans” or “small business software”)? In that case, the IA should reflect the most common customer types to make sure that people can self-identify and find what they want. Of course, if “adult” is a keyword you see a lot, perhaps that’s more of an indication that your site needs to tone down its photography.
One thing worth mentioning is that visitors may have a different conception of how your offerings map out than you do. While marketers may organize products in a specific way because of internal issues (business units, intended uses, etc.), it pays to listen to how customers view the breakdown of products.
- Competitive reviews: Maybe someone else built a better mousetrap
When in doubt, consider how your competitors organize their IA. Assuming that your brand has a competitive set of more than two or three companies, it stands to reason that customers and prospects, as a whole, will spend more time on their sites than on yours. As a result, it may make sense to bring your IA more into line with theirs.
At this point, steam has emerged from the brand folks reading this post. “You can’t just copy your competitors,” they are saying. “Your brand needs to stand for something distinctive.” Brands must be distinctive, yes. However, sites exist for the most part so that customers and prospects can get information. Arranging your site at odds with what visitors have come to expect based on the category hampers their ability to get information.
Thus, marketers should look over their shoulders to see what their competitors’ IA look like. Too much difference may lead to too few visitors.
Detailing the IA of a site below the top two or three levels requires even more attention to data. However, these three sources should give marketers enough tools to sketch out the basics. Of course, if you’ve got some clever uses of data to drive IA, please share in the comments below.