The data we discussed capturing in the first four habits doesn’t have much value unless you review, understand and ultimately ACT on it.
To understand how to gain actionable value from your data, you must understand the difference between a report and a model.
is a factual list of data—current or past.
is an assessment of the data with some qualifiers applied to help derive value
For example, imagine you are trying to answer the question “Should our company invest more in paid search?”
To begin with, you will need a report of all the people that interacted with your paid search campaigns. However, that information still leaves a lot of room for interpretation of value. Are those new leads or people we already know? Are they junk records or will they go on to to become customers?
Let’s assume our goal is to expand our new customer base and so we are seeking new leads. We can easily create a report that looks at new names from paid search. This means our question is more specific: “Should our company invest more in paid search to acquire new leads?”
In order to answer the question: “Should our company invest more in paid search to acquire new leads?” we'll need a few reports to start to answer that question:
From there, it should be pretty simple right? We want to calculate our ROI on paid search (or ROAS to use the paid search terminology) so we can simply add up the revenue that came from those specific customers and divide it by the cost of those paid ads.
You may not have realized it, but you were actually using an attribution model for that calculation. The example above uses the Lead Creation model. This model assumes that the channel that creates the record should get credit for all future revenue.
This seems simple enough, however it lacks the sophistication most marketers need to make smart decisions.
The Lead Creation model falls short when you consider how a person interacts with your company throughout the entire sales process. If you have a long sales cycle, this becomes even more important.Consider how different situations might impact your attribution report. For simplicity, let’s assume each deal only has one stakeholder. Again, this is unlikely in the real world, but it will help us explain attribution models.