With the rise of cross device tracking, it is perhaps timely to consider the true length of customer journeys and whether as a result, cookie lengths should be revisited. 

Reduced cookie periods

Traditionally it has been an industry standard to set cookie lengths to 30 days, with this being the time frame an affiliate can be considered to have played a part in the customer journey and therefore be rewarded for a conversion.

Premised on being a conversion channel, the click to sale time lag through affiliate marketing has always been minimal.

Typically sales would convert within the same day and the majority of these within the first hour. This would vary from sector to sector and by the type of affiliate, but all of the data has traditionally pointed to a quick interaction to conversion time window.

As a result of this, many advertisers have looked to reduce their cookie periods.

After all, if the majority of sales convert in the same day is there really a need for long cookie lengths?

One argument is surely that if most sales convert in a day, a longer cookie period has little impact upon advertisers, but a longer cookie period is more appealing to affiliates.

If two advertisers have a similar product offering, commission levels and conversion rates but one has a longer cookie length, affiliates would be more inclined to promote the advertiser that gives them a longer time period to be rewarded for their influence in the customer’s journey. If this has minimal impact on the bottom line, it makes sense to offer a longer cookie period.

Has the length of customer journeys been misunderstood?

However, what if the length of customer journeys has been misunderstood? Are they actually longer than we had originally anticipated? What is the real length of a customer journey from initial interest to an actual conversion?

With the launch of cross device tracking we have been able to delve into our data to get an understanding of how customer journeys look across a number of devices.

Previously we only saw part of the picture – namely the length of the customer journey from final click-through to conversion on a single device. Now we are able to gain a better understanding of customer journeys and the impact of adjusting cookie periods.

Looking at the data from one of our advertisers in the retail sector we can see the impact of cross device journeys on cookie lengths.

The chart below looks at the share of sales within a certain time period for cross device and non cross device sales.

74% of non cross device sales convert within a day compared to just 14% of cross device sales.

As the timeframe increases, we see a significantly higher share of cross device sales vs non cross device sales which underlines that there is much more affiliate influence within the customer journey than we first thought, however a lot of this upper-funnel influence actually crosses multiple devices.

Looking in more detail at the distribution of sales lag we see 63% of non cross device sales convert within the same day whereas only 8% of cross device sales convert in the same day.

Furthermore, only 50% of cross device sales convert within ten days of the initial click.

This is evidence that customer journeys are typically longer than the industry first realised and cross device tracking is the missing piece of the jigsaw to truly understanding customer journeys.


Breaking the data down further we are able to understand what this looks like across different affiliate types.

Firstly we look at a cashback partner. We know that their model is primed to convert so we would expect to see a significant volume of sales occur within a short time frame.

This is true of non cross device sales where we see 87% of sales convert within the same day but we see only 11% of cross device sales convert in the same day.

Just under 50% of cross device transactions occur more than 9 days after the initial click. This is perhaps surprising considering the nature of cashback sites but highlights that cashback users are researching their purchases on the site before later completing the purchase on another device.

We would expect content sites to have a greater click to sale time lag. They are typically involved in the early stages of a customer journey and are very much an influencing channel.

The chart below shows the distribution of sales lag for this affiliate type is much longer than we see with cashback, which is what we would expect.

This also demonstrates their importance as a cross device influencer with 50% of cross device sales converting more than 10 days after the initial click.

It is also possible to see the devices that are initiating cross device sales. In a typical single-device environment we see strong conversion rates through desktop (5.72%) and tablets (4.9%) we see smartphones (2.82%) lagging some way behind.

This data is slightly misleading as it does not take into account transactions that are cross device. While smartphones are not converting at such a rate we know that traffic through smartphones has grown at a considerable rate over the past few years.

26% of all traffic across the network originated through a smartphone last month vs. 18% through a tablet. So we know consumers are browsing through mobile but not necessarily converting.

The chart below looks at the devices initiating sales and a number of cross device sales are originating form a smartphone.

This highlights their importance as an initiating device – an insight that we had been missing prior to cross device tracking.

Matt Swan

Published 14 April, 2015 by Matt Swan

Matt Swan is Client Strategist at Affiliate Window and a contributor to Econsultancy.

25 more posts from this author

You might be interested in

Comments (4)

Pete Austin

Pete Austin, Founder and Author at Fresh Relevance

Agree with this. A long cookie timeout helps with multi-device selling.

I suspect that the figures are not quite as dramatic as shown here, because cross-device identification is difficult; customers who use two devices within quick succession are less likely to be recognized on both than those who return several times over a longer period; hence they less likely to be included in this data; and this makes average conversion times look longer. But I don't think this invalidates the conclusion.

over 3 years ago

Roland Latzel

Roland Latzel, Director of Marketing at MailStore Software GmbH

Interesting article and valid question. I am wondering as well what are today's current techniques to identify users in cross-device-usage, especially beyond being logged in e.g. in an online-shop.

over 3 years ago

Pete Austin

Pete Austin, Founder and Author at Fresh Relevance

@Roland: mainly cookies, supercookies and hardware-based ids. For example see:

Also you need logic to stitch these sessions together: decide whether what seem to be separate sessions by different users on several devices are best considered as one session by a single user on multiple devices.

over 3 years ago

Adam Davis

Adam Davis, Customer Strategy Manager at Softonic

Very interesting. What is the conclusion as to when cookies should expire?

over 3 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.