We have a client who is this close to breaking even on their ROI. They are a large ecommerce site and the profit margins are always narrow so it is incredibly important that that their campaigns be running smoothly.
Our client is active in their analytics account and often laments that we are running at a negative ROI. They are looking at this:
This is a Last Interaction attribution model of conversion tracking, meaning that that the sale is attributed to how the person came to the site the time they purchased without any consideration of if or how they came to the site prior to that conversion.
If we took the above chart at face value, the first response would be to make some serious changes to the paid search campaign and pump more into other avenues. The paid search drove only a few more sales that direct and organic traffic but at a much higher cost, so if we invest that $7,700 into organic initiatives (more blog posts, more content descriptions, make some videos, find some good partnerships for links, sponsor an event, etc.) then we could make loads more sales! Right?
Not necessarily.
What the last attribution model leaves out is that people don’t always purchase on their first visit. In fact, Inc.com recently reported that 98% of your visitors won't buy on their first visit.
For this client, first time visitors converted at 0.33% and return visitors converted at 1.06%. By only attributing the sale to the last point of contact we are not giving proper credit to the origin sources of the majority of the sales.
So what would the conversion attribution look like if we used a different attribution model, such as first point of contact?
Looking at the paid search we can see that the ROI is still low, but the return on ad spend (ROAS) is 10% higher! Not only that but our cost per acquisition is down almost $40, which makes a big difference in the bottom line.
If you look at the breakdown of first interaction conversions you will also notice that almost half of all conversions were initiated via paid search.
Paid search isn’t the only source that’s getting the short end of the stick with last interaction attribution, organic search is too.
Looking at attribution modeling really hits home that people use a variety of channels to find your site, and then may come back to your site through completely different. Looking at just organic, paid, or social traffic in isolation is not entirely representative of the complete buyer’s journey.
For this particular client, it is very common for the first interaction to be through a paid search ad. Then the user leaves (maybe their boss walks in the office and they have to close the browser to pretend that they’re working and not shopping) and then they come back, often in the same day, via a different source (often direct) to complete their purchase.
Online shoppers are fickle and educated, checking several sources for product comparisons, pricing, reviews, and more to ensure they’re making the best purchasing decision. Multiple visits before purchase completion is normal and to be expected. Make sure you account for it when you’re measuring the source of your sales!