Setting KPI's in Blurry Situations
Setting key performance indicators (KPIs) is one of the most important tasks an agency can do. It’s not an easy one, though. An important conversation needs to happen to determine what to measure, how and why. And of course, the more detailed and accurate, the better.
In a perfect world
In a marketer’s perfect world, it would be possible to track everything with complete clarity. But we don’t live in a perfect world. There’s always going to be channel overlap and actions happening that can’t be measured. The question then arises:
If we can’t define KPIs perfectly, should we define them at all?
This used to worry me because I wanted our clients to have flawless KPIs all the time, no matter what. And if it wasn’t perfect, well, maybe it shouldn’t be a KPI. But then I realised:
A KPI is by definition an indicator.
An indicator does not have to be infallible; it’s designed to be discussed in context as a component of a greater whole.
With this in mind, setting KPIs became a must. Even if they’re not perfect. Especially if they’re not perfect!
So, if you’ve not yet set any KPIs for your online marketing efforts, here are some easy-to-start suggestions for anyone with Google Analytics:
- Users – the total number of users who interacted with your website
- Sources – contribution by source (e.g. organic, paid, etc.)
- Engagement – average time on site, page depth, pages viewed, etc.
- Views of key pages – with particular strategic importance
- Conversion goals – by type and volume (e.g. product sale, form fill enquiry, phone calls)
- Conversion goal value per conversion – because some conversions are worth more than others
All of these should be measured:
- Month on month
- Year on year
- By device
The key with getting KPIs right is to do the best job possible with what’s available, and to ensure discussions include as much background info, insight, and interpretation as possible.
Co-Founder and Director
Helping you navigate and leverage the world's most powerful digital marketing channels.