With the emergence of Facebook, Twitter, RSS, and blogs; the development of the iPhone, iTunes, the Kindle, and the pending iPad; and the continued utility of email, which has only been enhanced by smartphones — well, there’s a question haunting the status quo of Web development for publishers:
Do you really need all that Web site?
We’ve been enamored with the big site for more than a decade now — a shovelware monstrosity that has mutated to include e-commerce, supplementary data, rapid online publication, and complex search engines.
Designers slave over these sites, and entire departments try to analyze the data and make sense of it. Domain-level sites produce a lot of related activity because they’re the centers of our online identities, the foundations upon which we’ll build, the areas of wise investment and strategic renewal.
But is that where your brand generates the most value online? Is that where you have your most dynamic community?
Is your .org, .com, or .edu going to remain the center of your future digital identity?
If you’re certain that your most profitable, highest impact, and smartest move over the next 3-5 years is to do more at your main domain URL site, then it probably makes sense for you to build an appropriately extravagant presence there. But what if you’ve been blinded to alternatives? What if you’re greatest value is to be found on Twitter? On Facebook? Through email? On devices?
And what is the opportunity cost if you’re overspending on a set of dressed up directories that your customers really don’t care much about?
Basically, what is the purpose of that Web site you’re running?
It’s an interesting question that’s posed nicely in a post by Mitch Joel on his TwistImage blog, which is itself an example of the issue at hand.
Mitch’s company has a complex site, but his overall digital presence is mainly supported by the TwistImage blog. For cents on the dollar, he cultivates a loyal Twitter following, blogs actively, and lets the viral work its magic.
Contrasted to the search-centric mentality we’re accustomed to, the concept of the “feed” has renewed the power of browsing. At the same time, the sheer abundance of information has ensured that alerting is and will be more important to information awareness. Meanwhile, devices like the iPhone, the Kindle, and the iPad have become or might become central to users.
Scholarly publishers have site-centric approaches for defensible and rational reasons — institutions buy access to domain sites; technology providers in our space are geared toward site-centric approaches (from HighWire and its ilk to CrossRef and its approach); and only a minority of sites have been able to scale beyond the site-centric approach to see what other options might exist.
Yet, when you list out the forces at play (browsing, feeds, alerting, devices) in creating an audience, it seem that in some ways, scholarly publishers could get trapped in the site-centric model.
We reflect site-centric thinking when we do usability testing, for instance. I’ll bet that most of your usability testing has been about the site, and not about the usability of the complement of information options you use or could use. Did you ask if the email you send is usable and ties nicely with the site? Did you ask if landing on your site from Google made sense? Or were you just testing the usability of your site? If so, that’s site-centric, and potentially part of the trap that keeps us in the rut.
Because of habits of mind like this, we’ve probably over-engineered our site offerings. And with online still severely undervalued as a communication medium, these lavish expenditures may not earn back.
For instance, I’ve always wondered why, instead of creating a full-featured site, I shouldn’t just make a great email system and a good search engine, with minimal HTML and a bunch of PDFs on the back end. After all, if customers want the convenience of an online journal, this approach would satisfy all editorial and customer requirements, while slashing my costs.
Is there really a premium for “pretty” among scholarly Web sites?
Email is the lifeblood of our sites, driving the majority of traffic for us. Google does most of the rest. Twitter and Facebook are increasingly important in some fields. PubMed and other specialized search repositories are also major sources of awareness and traffic.
Why do I need a big Web site to benefit from all this?
Let’s say I have a decent site. If I were to add a good Twitter presence and maybe make the home page more like a blog (maybe even run it off WordPress) while delivering PDFs to authorized users, my online presence might improve substantially, and at a fraction of the cost of what I’m paying the site-centric providers now.
I could still sell the site to institutions. I would probably have the same level of traffic (with a blog main page, I might actually be more visible in Google), and I’d be less tied to any particular provider of technology products.
Are we spending tens or hundreds of thousands of dollars on over-engineered, site-centric solutions? Is the “Age of the Big Website Build” drawing to a close?
23 Thoughts on "The Big Web Site Build: Are We Approaching the End of an Era?"
If one is taking a forward-looking approach, is a focus on the PDF format really a wise choice? The static nature of pdf files makes them useless for a lot of the really interesting new types of data, particularly movies generated in imaging labs. We’re also seeing all sorts of interesting new enhancements, from PLoS’ article level metrics to recommendation engines being built around the html versions of papers, not around the static disconnected PDF files. PDF takes the user offline and removes any hope of interaction through new media technologies.
Furthermore, given how poorly PDFs render on something like an iPhone, wouldn’t it at least be better to think about file formats like EPUB, which are going to be much better supported on e-book reading devices (Nature seems to think that way).
I’m not arguing against your overall point here, and I do think it’s important to consider a variety of strategies. What I’m questioning is whether PDF should be the focus of those strategies. Any thoughts on whether its use has peaked and is now in decline, or do you think it’s likely to be the foundation of scholarly journal publishing for a long time into the future?
I agree. My point about PDF is more about both a long-standing thought I’ve had (which was from the days when PDFs were the only choice) and also reflects a preference I still think exists among scholars, mainly that they want sites to be useful archives of articles. E-Pub and other formats will help move to devices, but PDF preserves formatting and engenders trust. Its utility in scholarship may not erode very quickly.
It is an interesting question. Right now, PDF still rules, so any design would have to take it into account. But if you predict a near-future where we see more and more access from mobile devices (with very small amounts of storage) and ubiquitous internet access, does the usage pattern change? Are there advantages to accessing an online version of the paper, one that has constant updates added or lives as part of an interactive ecosystem?
Or should we take the success of iTunes over subscription services as a lesson. Do people want a permanent copy of their purchases that remains in their possession, rather than accessing something in the cloud?
Both! Or, rather, most publishers should probably offer both, or be _able_ to offer both. Some readers will prefer to own, others will prefer to access (and get the continual updating). Some content is never going to be updated, other content cries out for updating. Focus on the content!
Kent, perhaps you can distinguish between searching on the site and structuring a site for which searches on other sites (or services) bring users to the target site. I believe you are critiquing the former, but there may be some confusion that you intend the latter.
Thanks, Kent, for shining another spotlight on another elephant in the room.
My advice: don’t start with the site, start with the content. I would never advocate that publishers pull the plug on their websites. But I think the main message is one of proportion, and focus. If publishers shifted their emphasis to making their _content_ more discoverable and more usable — by all who may be interested in it, in all the ways they might find it, and in all the modes in which they’d like to use it (and _pay for_ it)–I think they’d see a big payback.
This gets at the PDF issue. Sure, of course, make those PDFs available. But put some effort into how you can make them (and the content they contain) more discoverable and usable. Make that content (the _content_, not just the _PDFs_) discoverable and useful in other ways as well. EPUB is part of the solution, but it isn’t the whole solution (especially for scholarly — and especially especially STM — publishers). Get that content into various deliverable forms (or, better yet, develop the ability to _generate_ the optimal form when you need it — just in time, not just in case), and put some useful semantics on it, and let it go to work for you!
— Bill Kasdorf
While I agree with your overall point that the days of the behemoth Web site may be drawing to a close, I don’t think that scholarly journal sites are a good example of overly engineered sites. Most journals sites are fairly simple and adhere to the Louis Sullivan rule of “form follows function.” Journal sites are mostly tables of contents with articles in PDF and HTML formats (and there are, as David touches on above, good reasons for maintaining both formats at this point in time). I think few journal publishers suffer from the delusion that their site is the center of the universe for their readers. A quick look at usage statistics will disabuse anyone of that notion. Publishers know that users are coming to their content from search engines and from indexing sites like PubMed and have built their sites to interface well with these discovery tools.
Here’s a highly relevant article to questions of design and user behavior: How do researchers use online journals?
What about the future for the behemoth site that scholarly publishers use for ecommerce? We sell mostly scholarly books, but some journal subscriptions on our site. Our current, our very old site is using 2002 technology. Despite this, we still sell a very decent amount. We are about to launch a new site. We expect that with much greater marketing capability and greatly increased electronic content offerings, our sales will grow quite nicely for some time to come and therefore the site will more than pay for itself. We may not catch up to our Amazon sales figures, but we’ll close the gap some. Most importantly, we’ll have some more customers directly coming to us.
E-commerce infrastructure and “behemoth” sites might be viewed as different things, right? What if you fragmented out your online offerings, but left the e-commerce infrastructure in place to support them? Would your sales increase? Would the perceived relevancy of each microsite increase? Could you drive down costs and reduce time to execution? I’m confused by your statement about wanting to “catch up to [your] Amazon sales figures”. Do you sell more through Amazon? Then, I’m sure you’ve asked whether you should even have your own storefront. What would you save if you didn’t? What if you just pointed people to Amazon through the microsites you might build on blogging platforms? What if Twitter drove traffic to Amazon?
I wrote the post mainly to get people to stop and think. I don’t know what the answer is, but I’m certainly going to be taking a harder look at options instead of just doing the rinse-and-repeat of online site development. There are too many great options that cost less and do more, and devices are melding with content and sales in new and exciting ways.
I did a research project last year that turned up the interesting fact that when publishers made a concerted effort to sell books directly from their Web sites, not only did they have some degree of success (in my small sample, direct sales came in at about 2% of total volume), but they also found that their sales at Amazon rose. It appears that a strong online presence lifts all boats.
Yes, which in my mind actually suggests that a broader online presence equates to marketing in some ways, even passively, given Google and other factors. Marketing creates awareness, which drives sales. Makes sense. Thanks for mentioning this. Yes, a strong online presence lifts all boats, which begs the question, “How best to get a strong online presence?” I think some rules (well, they’re more like guidelines) might have changed.