Using a case study from one of our clients, we examine the halo effect of sending an email. We look at the uplift in revenue across other channels when an email is sent and consider the merits of different attribution models and how email marketers can best demonstrate the full impact of their work.
Learn how to measure the real impact of email in your organisation.
A lot of clients are struggling with measuring the real impact of email in their organisations. Demonstrating the full impact of your work can be difficult and very often email will be left undervalued. Proving that email's contribution to the bottom line is actually 30%+ of sales and not, say, 15% can be difficult. The most common attribution models will include the following elements:
- Types of action: Sent, Open, Click
- Timeframe: Days, Weeks, Months from action
- Order of touch: First, Last, Any
So which one will be best for you and will make sure that email gets its fair share of success?
Choosing the right one will depend on elements that are specific to your business, such as sales cycle for example. The most popular approach is to focus on tracking behaviour or conversion as a result of an email click. However, this can be tricky as we all know that in the real world not everyone will click through from your email and go to the website and purchase. In fact, they might not even open it. But there is strong evidence that some might see the subject line mentioning, for example a sale and make a mental note of it and purchase through another channel a few days or even weeks later. This is proven by the UK DMA's 2013 Consumer Tracking Report, which demonstrates the value of email's broad range of responses. 23% of respondents mentioned 'visiting the company's website via another route' as a likely response to an interesting mail and 25% cited 'visiting a shop or retail outlet'.
So, how do you measure and present the real value of email when there was no open or click?
Running some analysis of revenue from all channels excluding email is a good place to start. Take all your revenue for a given period of time, deduct email revenue from it and break it down by days when email was and wasn't sent out.
Here is an example from one of our clients. The orange bars show the average daily revenue in the month on days in which email was sent whereas the blue bars show the average daily revenue for the days on which no email was sent. Since none of the bars include 'email revenue' it's very clear that revenue not attributed to email is present, those days with email deliver higher revenue through non-email channels.
The next two charts look at the source of revenue based on last touch and first touch attribution for one particular month of the previous chart. This shows the breakdown of how much impact email has on each of the non-email channels.
Email has the biggest impact where no channel at all has been identified, after that, in percentage terms, is Social attributed sales, though of course the overall value of sales attributed to social are insginificant.
Alternatively, you can set up a hold-out group that will show you the true impact of email. Create a sample group of subscribers who will be receiving your emails as usual and after a while stop emailing them and compare the results. The main conclusion is that email drives sales in other channels. You don't need to open or click an email to buy. So look at the bigger picture and consider email more as a broadcaster, like TV or advertising where there is no particular action that can be attributed to. See how email contributes and supports all other channels. But you'll have to be careful as defining it too broadly might lead to overstating the revenue generated by email and defining it too narrowly might understate it.
Things to remember before you start working on your attribution model:
- Email drives offline response (it's a fact)
- Integrate your data and website analytics to get a full overview of email's impact on traffic and conversion
- Track and analyse online and offline traffic and conversions to see if there's any correlation with your emails sends