Understanding the analytics iceberg

Publishers are awash with useful data about authors, customers, usage, and plenty more, and it has never been easier to grab a set of numbers, put together a few charts in Excel, and create some interesting reports.

Suppliers have been doing this kind of thing for years of course: with email marketing providers charting open rates and click throughs, web analytics tools graphing website traffic, content platforms tracking article download trends, and so on – in each case analysing the particular chunk of customer data those systems have readily available.

Individual publishing staff often do something similar in-house: manually putting together dashboards in response to management demand for headline reports, often with many long hours of effort in collating data and presenting it visually.

However, the results of all those types of efforts don’t entirely hit the spot. Presenting the numbers you have to hand with an attractive chart is just the tip of the analytics iceberg – it’s just the ‘quick win’ providing only limited insight or longevity – because a really effective analytics programme requires a whole mountain of extra work beneath the surface to create reports of lasting value which are truly comprehensive, trusted, and repeatable. Let’s briefly look at each of those 3 aims to understand why there’s so much more hard work to be done beneath the surface:

1. Comprehensive. Scholarly publishing is quite a complex business, and any single measure in isolation isn’t going to tell the full story. For customer analytics to be comprehensive, at the very least they need to synthesise the "big 3" elements of author submissions, subscription sales, and article usage. With any one of those elements missing, your analytics will have a blind spot regarding the creation, selling or consumption of content, all of which are critical health indicators for every publisher.

2. Trusted. Doubts over accuracy can often undermine the value of customer metrics, so a carefully de-duplicated single customer view is an essential foundation for any analysis – without it there’s a high risk of double counting customers. There’s also a lot of work required to ensure that revenue counts match the figures from accounts, and to properly allocate fees from complex multi-site or consortia deals. Data quality is also a major challenge when consolidating variant names and codes for different products, packages, countries and regions. If all those issues can be overcome, then you can start to create reports which people really trust.

3. Repeatable. Ensuring analytics are repeatable doesn’t just mean doing it again, it means being able re-create the same measures in exactly the same way, month in month out. Repeatability is the biggest problem with many hand-crafted in-house dashboards: it can often become too difficult or too time-consuming to reproduce those exact same reports consistently over time. And of course without consistent tracking, there’s no visibility of recent trends, and no feedback on whether steps taken are having the desired effect.

So, if ever you’ve thought that customer analytics are all about presenting numbers via charts and dashboards, then bear in mind the ‘analytics iceberg’. There’s a lot of hard and specialist work to be done beneath the surface to produce reports which will be comprehensive and reliable enough to provide real long-term insight and value.