Google have just released a whitepaper – Beyond last click: Understanding your consumers’ online path to purchase – with some very interesting stats about the purchase path length and types across different verticals.
Rather than just regurgitate the entire paper here I have pulled out some of the highlights.
- The research journeys (search terms used, website visited etc across different days) were found to be quite long – lasting a month or more in some categories. Only 20% of conversions occur in the first 24 hours for purchases in the travel & apparel sections
- On average 1 in 3 conversions occur 30 days after initial research began
- 70% of purchasers use search at some point in their research journey
- The volume of research is not linked to average order value
Some Other Key Points
If you were reviewing the performance of a new campaign at the end of moth one you may be missing as many as 40% of your conversions that will come through later (for apparel & travel sectors). Always think about the delayed purchase path for your industry sector.
Purchase paths can be very complex – we need to ensure as full coverage across the entire search funnel as possible. ‘Research’ type keywords could be very valuable to you but you need to try and understand this through clever use of analytics. Last click attribution models may not show you the bigger picture.
- Look into multi click attribution models such as the updated version of Google Analytics’s Multi-Channel Funnels
- Give campaigns enough time to see all conversions coming through
- Think about the full search market when optimising – not just the highly commercial terms
- People visit sites several times during their research process – it pays to make your site memorable, and to have a first visit goal in mind (newsletter subscription, social media follows etc)
- Some sectors are more heavily search driven than others but they are all highly significant
- Generic search terms play a significant role in the search process. Some areas are more heavily brand driven (such as mobile phones) but still have significant generic search volumes