Looking at web traffic and learning from trends

So, today I’m thinking about  the data created by people who visit and how I can potentially use it to add value to the site functionality and perhaps spot trends for future use.

What do I mean by user created data?

Well, everytime a visitor lands on the site they create a footprint. The footprint will in most cases tell me where they landed, what they did and where they came from. In isolation, such info has limited value, but in larger numbers it can begin to tell stories that can then be used at a later date or just added to the general mix in yet to be figured out ways.

So each page visit is recorded in a reasonably anonymised way. I get to see device specifics and their IP address and what they did whilst here.

Unless they sign up and choose to share, I don’t know who  (nor do I care) they are individually  or why they are seeking to go from A to B (that would be a bit creepy) but it’s useful to know if some journeys are popular (have lots of users viewing them) as we might be able to look at other patterns around that space and make a few informed decisions. Year on year we might see that on some dates we see a lot of activity around particular routes and destinations.

This year for example, the Glastonbury festival is on the 24th June to the 28th June.

When I query the database for page views today that had the destination of ‘Glastonbury’ I see an uptick in user activity around those, meaning that people were seeking information on how to get to or from the festival.

Travel times, fuel costs for instance.

Here’s an example of a page for London to Glastonbury and another from someone looking to go from Manchester to the Glastonbury Festival.

There are a few others, but as I only put together the recording scripts yesterday, the pots are small, besides, you get the picture I’m sure.

In the days after the festival has finished, lets say post 28th June I’d imagine that traffic to these pages would be close to zero as interest wanes and people move on to other activities.

But yes, so what could I do with all this? Well, I could  look at the locality of Glastonbury and know that it is in Somerset, England, United Kingdom.

I might then create a ‘trend’ module that looked at such activity for display to site visitors from the same region or a radius within. I could call it “Trending Destinations” for instance which would show at the top to visitors from the UK “Glastonbury”.

I might then do a look up using a social API like twitter or interrogate Google trends for information around this destination and create some kind of mash up that informed users why this activity might be going on and add a little realtime context. I could also create a trendmap with dates and predict activity for the future,  or an events section that used retrospective data to create new stories.

Different people from different parts of the planet will see different trends. A bit like Twitter do with theirs. They are geo targetted and specific to where people have indicated in their accounts or from the geolocation data provided from their IP address or browser.

I’ll post more when I’m done, but watch this space.

If you have any great ideas for what else could be done in this sphere then do please share, if it’s super cool I’ll try and code it and use it.

Posted in development and tagged , , , , .

Leave a Reply

Your email address will not be published. Required fields are marked *