Content builds a bridge and educates.

Your prospects and customers appreciate this kind of thing, and you hope to get leads and sales as a result.

Evaluating your content’s performance is a little trickier than it may seem, though.

marketing-attribution-modeling

We use HubSpot because it does an exceptional job of connecting the dots and allowing us full visibility into what’s working and how individual contacts are exposed to us.

Even with HubSpot in place, there must be conscious decisions made about how to attribute your content to subscribers, leads, and customers.

Marketing attribution modeling is the process of determining which and how many channels a customer was exposed to before their conversion event. The mess comes in when trying to assemble data from multiple touch points and channels into a cohesive report.

It’s also easy to report on the wrong things – especially when there are gaps in the data.

But sorting through the mess is worth it because attribution models can help you truly understand your conversion funnel and customer journey.

Marketing Attribution Modeling

The best place to start with marketing attribution modeling is to know the major types of marketing attribution models and when it’s appropriate to use them.

These are the attribution models found in Google Analytics, one of the most widely-used analytics platforms. Most of these models can also be found in HubSpot, but more on that later.

Last Interaction

The last touchpoint gets 100% of the sale.

Last Interaction is good for measuring which channels result directly in a sale – but all other touch points leading up to the sale are not considered.

A great use case is figuring out which channels are at the bottom of your sales funnel. Even if you think you know your funnel, this attribution model can challenge your assumptions and reveal what is truly driving conversions.

Last Non-Direct Click

This ignores any sales resulting from Direct traffic and instead attributes it to the channel that was clicked-through immediately before the conversion event.

This model devalues the Direct channel by design, which can make it easier to understand which channels result in direct visits but harder to assess the true value of your Direct traffic.

You can use this attribution model to see which channels that refer traffic to your website result in conversions. It’s great for knowing where to invest your time and money offsite.

For example, if a lot of organic traffic is getting credit here, you know to ramp up your SEO efforts. If a lot of paid traffic is getting credit here, you know to ramp up your ad spend.

Last AdWords Click

The last AdWords click receives 100% of attribution for the conversion event, meaning attribution only goes to the Paid channel.

Great for measuring the effectiveness and return of your AdWords campaigns even if customers don’t convert immediately after a paid click-through.

You can utilize this attribution model specifically for paid ad campaigns to see the true value of a paid click throughout your customer’s entire journey.

First Interaction

The first touchpoint gets all the credit – the exact opposite of Last Interaction.

Use this to determine which channel initially brings eventual customers to your site.

As the inverse of Last Interaction, First Interaction is a really great way to challenge your assumptions about top-of-funnel interactions with your customers, particularly to determine which first touch points are bringing the most leads to you.

Linear

Each touchpoint shares equal credit.

This is an effective way to measure general performance of all channels.

This is great for painting a broad picture on which channels are involved in any way on your customer’s journey, though Linear analysis can get messy and unwieldy really quickly.

Time Decay

The touch points closest to time of sale get more credit.

Great for measuring promotional events and campaign performance.

Position Based

40% of credit is assigned to both the first and last interaction, while the remaining 20% is distributed among all channels in-between.

Puts more value on first and last touch points to help you better understand top and bottom of funnel interactions.

You can utilize this attribution model to understand the correlation between top-of-funnel and bottom-of-funnel interactions on your customer’s journey, while reducing the noise of everything in between. This is a great compromise to develop more helpful data with less messiness than the Linear model.

Which attribution models are in HubSpot?

Marketing attribution modeling is built right into HubSpot with some twists.

Unlike Google Analytics, you first choose from three different interaction scoring types, then you choose your attribution model. HubSpot also has more of a page-by-page focus which you will see reflected in the model types.

Here are all the different types taken from the HubSpot website leading with the interaction scoring type followed by the different models.

By URL: URLs of pages that contributed to conversion
- All interactions: All pages viewed
- First touch: First page ever viewed
- Last touch: First page viewed on the most recent visit to your site
- Last interaction: The page a contact converted on
- First and last interaction: The first and last page viewed, with equal weight given to each page
- Simple decay: The last six pages viewed, with more weight given to more recent pages
By Referrer: URLs of referring pages that contributed to conversion
- All sessions: All external pages that referred a contact to your website
- First touch: The first external page that referred a contact to your website
- Last touch: The external page that referred a contact’s most recent visit
- Last interaction: The page that referred a contact to the page they converted on
- First and last touch: The first and last external page that referred contacts to your website
- Simple decay: The last six external pages viewed, with more weight given to more recent pages
By Source: Channels that brought a contact to your website
- All sessions: All sources
- First touch: The source of a contact’s first visit
- Last touch: The source of a contact’s most recent visit
- First and last touch: The source of a contact’s first and last visit
- Simple decay: The last six sources, with more weight given more recent sources

You can see how there is a lot of overlap with Google’s models. Google’s models would all fall into HubSpot’s By Source interaction scoring type.

Which model should I use?

Simply put, there is no ‘one true’ attribution model to use with marketing attribution modeling. Each serve different purposes.

It’s important to evaluate your marketing efforts as a whole and understand what attribution model makes sense on a campaign level.

If you have multiple marketing activities going on at a given time – and you probably do – then you will need to utilize multiple attribution reports to fully understand how your content creates conversions.

You could even use multiple marketing attribution models for the same campaign.

For instance, you could use a combination of time decay and position based reports to get a better picture of the customer’s journey.

Although marketing attribution modeling still requires a judgment call to choose the best model to assess your marketing performance, understanding the tools at your disposal is the first step.