‘What we can’t measure we can’t manage’. It’s a deadeningly familiar mantra – and one that tells only a partial truth: if the entire focus of your management style is on stuff you can put a number to, then I’d suggest that there’s something seriously missing in your management toolkit.
However, metrics are incredibly important, especially in the case of digital which, it seems almost tautological to point out, is a highly data-driven affair. In fact, how can you realistically expect to manage an effective online business unless you have quantifiable metrics to tell you not only whether things are working or not (which is usually fairly obvious) but also how they are working?
Certainly you can look at your sales figures, but raw sales figures on their own are a trailing indicator. If you’re selling a subscription-based product (as the bulk of publishing businesses still are) by the time your sales figures tell you the subscription has not been renewed, it is usually too late to do anything about it.
So you also need to be looking at, for instance, usage. Usage is more of a leading indicator. It can tell you how you’re doing right now, and how you’re going to be doing in the future – whether you buyer is getting value out of their product, and so how likely they are to renew.
To me this is pure common sense. But usage is not the only game in town – far from it. Here are five ways that analytics can help drive value in a publisher business.
- Drive up usage
- Identify latent demand and new markets
- Improve conversions
- Guide publishing policy and strategy
- Gain business intelligence, market intelligence
1. Drive up usage
There are many useful pieces of information that you can derive from looking at how users interact with content and how they get to it in the first place, to help you drive up usage. I’m going to mention just two of the more obvious examples.
Firstly, what kind of devices are they using, and how can we improve their experience on those devices? It is often assumed in our industry that mobile usage will be low, and therefore not worth bothering with. Our experience at HighWire has been that analytics often prove this received opinion wrong.
We sometimes have this strange belief in our industry that academic users are somehow different – but why would students or researchers or professionals have different usability needs from a website produced by a publisher compared to those websites produced by Google, Facebook or Amazon? It makes no sense. They are busy people, they need information, they need to get on with their lives. The idea that suddenly the rules that pertain in the B2C market – which is to say, for the vast mass of ordinary people – are suspended when they go to a publisher’s website is a fallacy.
And sure enough we have recently reached sustained usage figures of 28% or more for smartphones and tablets on some of our clients’ sites. Furthermore, where we have redeveloped sites to be more mobile-friendly, we have seen significant upticks in mobile usage, proof that there was latent demand previously being unfulfilled.
In a way this is a question of learning the lessons of the B2C web, and when you look at the trends in that space, all roads lead to mobile. Look how much effort goes into making websites like Google or Twitter work not just well, but exceptionally well, on mobile devices. Monitoring device usage in your analytics, and contextualizing them within the wider space of general web behaviour, can help you build an experience for your users that will ultimately drive up usage.
Secondly, I want to take the example of discovery, and routes to content. Here it is important to understand not just what you see, but what you don’t see.
You would expect there to be a number of different sources from which people arrive at your content. Obviously, Google will figure massively – but there are other ways and routes too – including library link resolvers, abstracting and indexing databases, and citations from across all scholarly content, including Wikipedia. You should be looking to analytics to validate that you are using those third party routes into your content effectively. Are they proportionate, are they balanced, are there gaps that you should be looking to improve in order, again, to drive up usage?
2. Identify latent demand and new markets
It is important to emphasise the point that, with analytics, what you can’t see is often as important as what you can. Once users get to your site, for instance, it is worth looking at who is trying, but failing, to get access to the paid-for content: who’s bouncing off the paywall?
Here, we have the advantage in our space that in many cases we can identify where those users are coming from. Identifying institutional users by their IP addresses is one of the standard ways in which access is managed (though not the only way: our SAMS system knows many others as well). This makes it possible to generate some actionable business intelligence from these bounces.
You might want to send off a sales person to talk to that collection of institutions and point out that their users want something they are not able to get at the moment – so why not take out a subscription?
Then again, by looking at geographical or other factors it is possible to identify market demand that might come from unexpected quarters, opening up the possibility of new markets for content.
3. Improve conversions
On any e-commerce site a critical metric is conversions – how efficient the site is proving at turning browsers into buyers. Though the dominant business model in our industry is subscription-based content, where the end-user is not the buyer, increasingly a proportion of users are becoming buyers. In that part of the market, analytics has a fairly straightforward role to play in applying some B2C best practice to achieving that one-time purchase or end goal.
However, analytics also has a role to play in the case of subscriptions. So what does conversion mean in that context (I hear you ask)?
Suppose a library has purchased content on behalf of its users, through an annually renewable subscription. By the time the users go to access the content, the purchase has already been made – however the value of the purchase is not realized until the user accesses the content. If attempts to find the content fail, if the content is so badly optimised for tablet that the user can’t bear to read it on her iPad, if the site is unusable on smartphone and the user abandons his session before he gets to the content – then the value of the purchase has not been realized. So OK, it’s too late: the purchase has already been made. However, when it comes time to renew …
There is therefore an argument for considering success in accessing the content as a conversion, and optimising appropriately. The success that users have in picking content, or downloading articles or epubs, or reading the pieces of content that they find, should be seen as a key metric. The purchase might already have happened, but these ‘conversions’ are what will justify the next purchase.
4. Guide publishing policy and strategy
Analytics can also, of course, show which specific items of content the users are consuming. In a large collection it is well worth analyzing which bits get used and which don’t – with an eye, perhaps, to deciding which subject areas and disciplines to develop further. In terms of developing a forward publishing list, such analysis of existing content usage can provide insights and guidance to drive policy about the content that is to be curated and published.
5. Gain business intelligence, market intelligence
So far, most of the data we have talked about is fairly specific to a particular publisher, and we have mentioned only in general terms the idea of putting this specific data against a wider context. There is another dimension of analytics however, to do with understanding the market and setting a baseline against which to contextualise the data gathered on an individual business.
This is a developing area, and one with great future potential.
A market-leading access management system such as SAMS produces a great deal of data that can give a view across the big matrix of content providers and users. Drawing on aggregated and anonymised data from SAMS (with due attention, naturally, to data protection regulations) it is possible for us to look at usage of content on a global scale.
We can do this because SAMS looks at all the institutions that use scholarly content, and a reasonable fraction of the content that they’re using. Not only can we say who is using what, and how much they are using, but we are able to build up profiles of organisations based on the content that they use –across subject categories, product types, etc. You would expect a teaching institution focusing on a specialised area of professional practice, for example, to have a particular kind of profile. You would expect a typical university, however, to show something different. But now it is possible to back up these assumptions with analytics –and if necessary, qualify them. In the round, this can lead to powerful data on the shape of the market.
This is therefore useful business intelligence. It lets you compare institutions to each other, to understand not just how they differ, but also where there are gaps, and therefore where potential exists for sales uplift. For instance, say you are in the business of licensing technical content for a specialized area: it might be that a particular institution in your target market is accessing a lot of that type of content, but only a very small amount of it from you. That’s a powerful piece of insight.
This type of analysis can also help you see the shape of the market as a whole, and provide insights for product development. It could help you see where there is going to be space for more products in a particular category, but also where products in another category are occupying a niche into which you might not want to go any further.
These are just a few examples drawn from the type of areas that we cover in our analytics practice at HighWire, but I hope give a flavour of what is possible.
People sometimes say to me that they are too busy delivering to focus on analytics, and feel that the benefit they might derive from doing so would be merely incremental. The wealth of information that can be gained from analytics is so extensive that it is true, some of its insights will be peripheral to the needs of the business. But an awful lot more is absolutely core to business success – and as we move forward into an ever more data-driven age, you are certainly missing a trick if you don’t avail yourself in some way of the value that advanced analytics can provide.
Latest news and blog articles
Full-text HTML of preprints now available on medRxiv