You can't grow your business without rock-solid data
Asking your customers is a simple but effective way to get started
There are many different sources of marketing data
Eventually, you will want to combine these sources into an 'attribution model'
If you don’t know where your customers are coming from, then you can’t grow your business. It’s as simple as that. Many startups cut corners on their attribution set-up, it can sometimes fall between the cracks between marketing, product and tech teams. We would strongly recommend giving this the time and energy it deserves, and making sure that you’re only hiring marketing professionals who feel the same.
That said, marketing analytics at a large enterprise company will look very different compared to a fledgling startup. There’s no need to over-engineer your solution in the early days of your marketing programme.
The simplest attribution method would be to ask every customer where they heard about your business. You could add this as a drop-down box in the customer journey, or simply ask the customer via email or phone. This would let you know which channels are performing best for your business, and is a good starting point.
However, this approach alone won’t give you the campaign-specific information you require to optimise your marketing budget across your various marketing campaigns.
To take your attribution analysis to the next level, let’s start by exploring the different data sources in more detail. These can be combined into an ‘attribution model’ at a later stage.
“UTM codes” refer to an extra bit of text added to the url of your site. It looks like this:
If source=facebook then you know that this user came from a Facebook campaign. On all the major ad platforms, you can add this as a ‘destination url’ i.e. it will never be displayed on the ad itself.
These codes are then picked up in a product analytics tool such as Mixpanel / Google Analytics, or you can store them in a database. You can then easily see which users came from which campaigns.
Google’s ‘Value Track’ methodology can also be used for automating the gathering of UTM data, and increasing the amount of data you capture.
Referring data is Google’s method of reporting visits that came to your site from sources outside of its search engine. This is useful for tracking hits from digital press sources from PR activity eg. techcrunch.com.
Ask your customer where they heard about your business, this is a simple but effective way of understanding what’s driving your business. It's also one of the main methods of measuring offline marketing campaign performance.
Making a model
It's natural to not have attribution data for all your customers. And for some customers you'll have multiple sources of data. In time, you'll want to export out your data and create a rules based model. This will combine multiple data sources to tell you where your customers are coming from.
There's no need to create a "multi-touch" attribution model at this stage, choose what suits your business best - first-click or last-click - and then stick with it. You can always use common sense, for example, to think through the impact of a first click attribution model on retargeting campaigns and adjust your method and conclusions accordingly.