Conversion attribution – dead before it’s arrived

I’ve been a great believer in conversion attribution but recently I’ve been thinking that with the evolution of how we interact with the web, conversion attribution has failed before it’s really even had a chance to shine.

True attribution just isn’t possible

Currently it’s that simple, true conversion attribution just isn’t possible – it relies on a user using just one device to browse the web and convert. This is however far from the ‘norm’, we are using two, three or even four devices:

  • Work computer
  • Personal computer
  • Smartphone
  • Tablet

So if a customer uses two or three devices on the way to purchase there’s no way to track the path to conversion.

With the growth of smartphone sales, mobile being Google’s fastest growth area and mobile web package sign-ups soon to overtake that of broadband. The conversion attribution model either needs to evolve to work cross device or the results just won’t be accurate, which leaves the question of – why bother with attribution?

Who’s developing conversion attribution technology?

The majority of companies developing conversion attribution technology are paid search technology providers such as Kenshoo and Marin Software.

To me this seems odd, are digital marketers expected to truly believe in conversion attribution technology developed by companies who’s revenue is made up of a % of paid search spend that goes through their technology? There has to be bias towards paid channels, if not they’re placing their main revenue stream at risk, and lets face it their not going to do that.

This leaves us with few companies who don’t provide paid search technology, one of which being TagMan. I have much greater faith that companies such as TagMan who’s revenue doesn’t rely on paid search spend will in time offer true attribution. However this still brings me back to the multi-device issue. Technology such as TagMan isn’t cheap and multi-device attribution can’t be achieved so how can we justify the cost of such technology when the results are a long way from accurate.

Should we base results on part of the picture?

This to me outweighs the above issues; the whole idea around conversion attribution is to provide insight into which channels played a part in a conversion. But as soon as a second or third device comes in to play the attribution model fails. So if the data is wrong should you be relying on attribution technology to guide online marketing budget across each channel?

Think of it in another way, you could only be getting a quarter of the data. If you have say a £6 million online marketing budget would you be comfortable using data that’s nowhere near accurate to guide the spend per channel? No, I thought as much.

What next for conversion attribution?

I’d love to know because the multi-device issue is monumentally difficult to solve.

I really want attribution to be successful because it will allow smarter use of budgets and allow us to make informed decisions and increase ROI of campaigns, but until the multi-device hurdle is overcome it leaves us with a problem, especially as many will jump on the attribution band wagon without giving a thought to accuracy. It wouldn’t be an issue if we only had one device, but that’s just not today’s ‘norm’.

  • The reality is that we search and purchase from multiple devices
  • The reality is mobile usage and search is growing fast
  • The reality is mobile friendly sites are making it even easier to convert on a mobile device
  • Should you trust attribution data from paid search technology providers where revenue is generated from a percentage of spend?
  • Is the cost of the technology worth it, for only part of the picture?

There’s a long way to go for conversion attribution before we get to true attribution. I hope it’s achievable, but right now we’re a long way off.

One thought on “Conversion attribution – dead before it’s arrived

  1. Interesting article Dave. Like your point on bias from the likes of Kenshoo/Marin: obviously it’s in their interests to promote the role of PPC, but I think we have to trust them for now, to take advantage of the technology (in the same way we’ve used Google and Yahoo’s Analytics offerings free of charge over the last few years despite the potential bias/data exposure risk here).

Leave a Reply

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

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>