Say hello

Jellyfish would love to hear from you.
Give us a call, email or fill out the form and we'll be in touch.

0844 488 1001

hello@jellyfish.co.uk

Close
*Required field
Close

Thank you

We'll be in contact soon

Multi-channel attribution: There is no but in attribution

By | September 10th, 2013 | Department: Analytics | Categories: Video, Insights


Managing Director Matt Owen explains how multi-channel attribution can clarify consumer behaviour and facilitate optimisation for better results across multi-channel campaigns.

Video transcript

Hi, my name is Matt Owen, and I'm Managing Director of Jellyfish UK and the title of this video is "There's no but in attribution." So sorry about the slightly cheesy title, but in other words we should be exploring, understanding, or even better using attribution right now.

So why is that? That's because attribution will help us to get a better understanding of digital journeys, how our consumers get towards our brands online, and also how to optimise for better results across multi-channel campaigns.

So what are the three reasons? So number one is that we can explore attribution modelling right now using analytics packages. So here's a screenshot from Google Analytics, and you can see here that I've highlighted a couple of interactions where there's clearly a multi-channel path going on from organic search to direct conversion, and we can see some keyword level detail. So that's quite interesting from a reporting point of view, to understand first of all that this is actually occurring at conversion level.

There's more we can do with Google Analytics. So here's another screengrab, and this is using a totally standard default segment in Google Analytics that allows us to look at any interaction where the last click is direct and then filter that via channel in this case paid search. And we can see that paid search is actually contributing to what you might think are direct conversions, and that would be pretty interesting if you're running a paid search campaign or paying for those clicks, for example, to understand that sort of interaction and how paid search is assisting what you may think are direct conversions.

So analytics can do quite a lot for us with this respect. But actually, when we look at the detail of paid search management, we kind of need a different sort of a tool to help us with this, something a little bit more dynamic. For example, using the Infinity product, we can see here a range of attribution models that we can apply to any paid search campaign. So we can move away from what you might think of as a standard last click model to various other models, like weighted first click, equally weighted, first click excluding direct, and so on. So in practice what this means is it allows us to view a different kind of data and a different view of insights into the campaign.

So let's take an example. We can see here that the keyword brand design has made a contribution to a conversion goal here, and it appears to have made a 1.8 contribution. That might sound a bit counterintuitive, but essentially what we're saying is that it's made a proportional contribution to a final goal conversion. So maybe in a five-click conversion path, one of those interactions was on that keyword, and hence it's contributed 0.2 to that final conversion.

So away from the science part for a moment, that's really significant from a PPC management point of view because we need to understand for discovery keywords that they have a value and that we're not ignoring them in the optimisation process, and that's what would have happened if we were working on a last-click model.

The third reason why attribution modelling is really important is around ZMOT, that's zero moment of truth. Do google that term. There's lots of interesting information around that. But essentially it's around the increasing complexity of consumer journeys and how many more sources of information are being used by consumers. So the number of information sources is doubling pretty much year on year. So that's a massive trend, and with that diversity of channels and platforms, attribution modelling is really, really important across online, offline, interactive TV, all these kinds of channels.

So to sum up then, attribution is going to help us massively to better understand digital journeys. There are many tools we can use now to drill down into the data at a platform level and to really try and get better results across the different channels as consumers move from paid search to organic search to email and so on.

So it's accessible, it's affordable, and it's doable. So what are you waiting for? Thanks very much.

Tags: ppc, Analytics, Optimisation, attribution, mult-channel optimisation