One day, you might ask ‘what the hell have path-to-conversion reports ever done for us?’ These reports now exist, are a massive leap forward and are already driving some advertisers to better online marketing plans.

However, given the pace of change and the ease with which we take such advances in our stride, here’s a pre-emptive strike against us ever taking full path-to-conversion reporting for granted.

To make the case, here are two insights, one for affiliates and one for display campaigns, drawn from path-to-conversion and attribution reporting that, without them would have been impossible to derive. In the future, the fact that we didn’t have such insight at our fingertips will seem astounding.


Looking at attribution reports for a particular client it becomes apparent that the affiliate program run by a major network is mainly featuring towards the end of the user journey, often being the last click.

This is partially because ‘direct-to-site’ and ‘SEO’, non-paid channels that might otherwise claim the last click are set to lower priority in the system, but it still suggests that affiliates are focusing on the last click rather than initiating the purchase.

This is obviously to be expected from a very performance-led channel, paid on sales, but this still prompts the recommendation that the client should review their affiliate base to ensure that it does not mainly consist of affiliates running a ‘goal-hanging’ strategy.

Recommendation? Request that the network concentrates on building up the base of content sites and also look into introducing a bonus or a special commission scheme for affiliates who assist in driving conversions, rather than just last click.

To do this would mean passing the affiliate ID into the container tag to plug it in to the path-to-conversion reports and hence be able to analyse on affiliate level. It might also be worth reviewing the messaging used in affiliates vs. display, and potentially using display messaging with a stronger call to action.

The data effectiveness report shows that the cost per conversion for this channel is attractive compared with other channels, but you still don’t want to be paying for sales where the affiliate had little real impact.


The same client’s data reveals that display campaigns have a very high level of assists (i.e. display appears in the path to conversion a lot). But, when verified against attributed conversions (when you take into account all other marketing events that also appeared in a path to conversion), the picture changes slightly.

It becomes apparent that, while display features in the journey very frequently, the actual impact is questionable.

Looking deeper at the raw path-to-conversion data, we can see that the display campaigns are continuously served to the same users, frequently exposing them to an ad more than five times, and, on occasions, as many as 40 times.

The effectiveness of this advertising is highly questionable and, by looking at the media effectiveness report, we can see that the cost per conversion is higher than any other channel.

Recommendation? A frequency cap on this activity should be introduced to reduce exposing a user to the same ad. To start with, a frequency cap of 2/3 per user per 24 hours, but the non-converting data in the data feed should be analyzed, to establish the optimum frequency cap for this activity.

When we are all taking such data and insight for granted in a few years' time, it is important we should remind ourselves that once upon a time we were all making do with much, much less.

Paul Cook

Published 9 December, 2010 by Paul Cook

Paul Cook, the founder of RedEye and TagMan, is a contributor to Econsultancy.  

28 more posts from this author

You might be interested in

Comments (2)


Matt Lillig

Based on my data on the value of the 'assist' in display advertising, I found that the assist totals varied from campaign to campaign. 

At Yahoo! we only attribute an assist to a display ad as long as of the following campaigns in the click path resulted in a conversion.  For example, if a consumer clicked on a display ad on day one and then clicked on a search ad on day five and converted, the display ad would receive an assist and the search ad would receive the conversion.  If no conversion took place with the search ad on day five, then an assist would not be contributed to the display ad.

In revewing the data, some display ads had lots of assists while others didn't.  It simply came down to the messaging and whether or not the ad resonated with the consumer (which brings in the value of targeted messaging).

More reasearch still needs to be done to see if a certain types of display ads draw more assists (for example, rich media video ads vs. static display ads) but for now it seems that proper messaging is still king.

over 7 years ago

Matt Isaacs

Matt Isaacs, Founding Partner & CEO at Essence

Hi Paul

Thanks for sharing this - great insights.  Although I'd hope most advertisers were using sensible frequency caps in their campaigns already (or that their agencies were advising them wisely on this in advance).

Like you, I feel the significance of the insight inherent in path to conversion data is often underestimated.  As we move to a world where more and more media is delivered via digital means - and with mass IPTV at last something that feels a real rather than a distant possibility - the long term potential to deliver a real understanding of an ideal media mix is enormous.  Alongside good econometric analysis, this could really become the bedrock of delivering true media and marketing understanding.

While not based on any quantitative data, I get the impression that while many advertisers have embraced the concept of path to conversion tracking most are still struggling to know what to do with the data.  The data is of course not the end but merely the means.  Real insight requires the right tools and a level of analytical and statistical rigour that is sadly still uncommon in the industry.

Bravo for raising the profile of this important issue - let's hope it encourages more folk to invest the time and effort required to make the most of the opportunities.



P.S. Readers just might be interested in our case study of our first explorations in path to conversion/attribution for TalkTalk from way back in 2007 - which was even before DoubleClick and Microsoft were able to capture the data let alone do any analysis!

Full disclosure - we don't work with TalkTalk any more (nor am I implying we do).  But we continue to do lots of work in this arena in with our existing clients like Betfair and eHarmony.

over 7 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.