Pinterest-related analytics is a growing business. PinReach is up for sale with a $10,000 price tag, Pinpuff recently received backing from a start-up incubator and GroupM just signed a group-wide deal with Curalate. Pinterest itself is clearly missing a trick. Semphonic’s Gary Angel says that with the right analytics strategy the site could monetise itself as a source of real-time user profile data.
Pin down your audience
Pinterest is one the most talked about social media platforms of the moment, with commentators evaluating its place in the social landscape and its potential value as a business. And yet it has no clear monetisation strategy.
For the vast majority of social media sites, monetisation comes in two forms – selling targeted advertising or selling basic customer information lists. But this misses arguably the greatest opportunity. Surely, with all the personal data that popular social media sites can gather on their individual members, their greatest potential is to help other sites perform more effective targeted advertising by providing real-time profile data?
“With all the data that popular social media sites gather on their individual members, their greatest potential is to help other sites perform more effective targeted advertising by providing real-time profile data”
The demand is certainly there. More and more sites are focusing significant investment in improving their ability to optimise anonymous traffic. Unidentified online visitors cannot typically be targeted effectively, but with the use of data-driven services that show what other sites the individual has visited or interests the individual has, real-time segmentation and personalisation becomes an option, even for anonymous users.
At the most sophisticated of today’s marketing organisations, the basic infrastructure of comprehensive measurement of the customer journey is almost certainly in place. However tracking the full customer journey and understanding it are two entirely different challenges. Gathering the immense volume of digital data streams that constitutes a customer journey is certainly a lot easier through using an array of powerful and cheap data analysis systems, but how can stream behaviours like a website visit, or pins to a virtual pin board, be effectively understood and attached to a customer?
No simple task
Traditional visitor segmentation still has an important role to play in the process of determining a customer relationship, however when it comes to digital data aggregation, two completely different types of visitors might have identical web behaviour. This is when a second tier of visit-type segmentation needs to be introduced to help aggregate digital data in a stream of web behavioural events, capturing the essential meaning of tracked behaviour. However this process is no simple task.
The creation of visit-based segmentation is a new and relatively unexplored art. Unlike traditional customer segmentation, where the variables are typically matched one-to-one with the customer (demographics and interest data, for example), digital visit segmentations involve many-to-one relationships and are entirely behavioural. Typically, it requires an analyst to figure out what the implications of the behaviour are in terms of broader interest profiles. Though difficult, this type of segmentation is quite powerful. Where the behaviours are directly applicable to the interest set (which would frequently be the case for Pinterest), it can be significantly more powerful than traditional segmentation.
In traditional segmentation, the analyst has to infer attitudes from demographic aggregates. Traditional visitor segmentations typically start with the business relationship and the facts already known about the customer, such as their demographics, psychographics and interests. In behavioural segmentation, however, the analyst infers intent and interest based on what the visitor looked at (and did). Typically, intent is derived both from key indicators and from the totality of behaviour in a session. Segments are usually built from a set of hierarchical rules that allow the analyst to categorise every visit to a website in terms of why the visitor was there. Interest segmentations are based on the specific types of things a visitor viewed and how those types relate to similar products.
The beautiful thing about visit-based analysis for both intent and interest is that a very large number of rather difficult to interpret events (like pins on a Pinterest board) are, when aggregated intelligently, often quite descriptive of a visitor.
A missed opportunity
Traditional segmentation tells you who your visitors were, while visit-type segmentation tells you what they were trying to accomplish. You can’t meaningfully understand web behaviour or web metrics without both of those. This type of visit-intent and interest-based segmentation is fundamental to getting meaning out of data.
Digital marketing is still fairly new – and the methods to isolate and understand consumer segments just aren’t well known. A powerful visit-type segmentation doesn’t just expose the data in a performance-efficient manner, it helps marketers understand how the data fits together to form a meaningful picture of the customer’s experience and the resulting marketing opportunity.
Whether Pinterest on its own can provide sufficiently broad profiling data remains to be seen. But by and large, social media sites are missing out on huge opportunities to monetise the communities they are building.
Gary Angel is president and CTO of Semphonic, a digital measurement and web analytics firm