Audiences can differ by vertical, market, account or product. There is no ‘one size fits all’, however there is one audience strategy concept that will fit with your account and your business – introducing MECE, also known as mutually exclusive, collectively exhaustive.
This principle is often used by management consulting firms, such as McKinsey, to describe a way of organising information. So why is this important when discussing audiences, and how can it work for you?
The principle of audiences
Imagine we are running a jewellery website, with the following categories on the site: necklaces, bracelets, rings and earrings.
Scenario one: Collectively Exhaustive (CE)
When reviewing their Google Audience data, we can see the following lists implemented:
- Visitors of category: rings
- Visitors of category: necklaces
Even if rings and necklace visitors were the most valuable, audiences should be collectively exhaustive. If your website has multiple categories with a low number of users, a new all-encompassing list could be set up:
- Visitors of category: misc. (-rings, -necklaces)
By doing this, we’re using all data points available and inputting it into the account. This allows more data signals for the likes of bid strategies. By having this additional lever in the account to upweight and downweight bids, we can prioritise bids and increase performance as a result.
A common misconception is that spend will increase by layering on audiences. This is not the case as by becoming collectively exhaustive, a new data layer is added into the account to make informed decisions on performance.
Scenario two: mutually exclusive (ME)
Let’s revisit the jewellery website. This time, we have updated their audiences with our suggestions, and have added a new audience:
- Visitors of category: rings
- Visitors of category: necklaces
- Visitors of category: misc. (-rings, -necklaces)
- All visitors
As we can see, the lists are no longer mutually exclusive.
Did you know that if you have a user who can fit into multiple audiences (each with the same bid adjustment), Google will not link that user to the most granular audience? Instead, they will be assigned to an audience list that they fit into, at random.
In this example, we would still see traffic coming through the category audiences, as these users will be assigned at random. However, it would not be the full eligible list of uses. i.e. users who have visited the rings area of the website may be classed as all visitors – data will now be skewed and uninformative.
To get around this, bid adjustments can be used to showcase the levels of importance, and to help ensure users are categorised to the right audiences.
- Visitors of category: rings – +20%
- Visitors of category: necklaces – +20%
- Visitors of category: misc. (-rings, -necklaces) – +10%
- All visitors: – +0%
How this can work for you
This segmentation can change depending on the product. Other examples include:
Segmenting based on user’s value:
- Gold membership
- Silver membership
- Bronze membership
Segmenting based on where they are in the user journey:
- Website visitors
- Basket abandoners
- Converters
Remember, they need to be mutually exclusive, collectively exhaustive. Contact us if you’d like to find out more about implementing this audience strategy for your own business.
