In one of my previous roles, I was tasked with identifying demand gen sources that had the highest ROI. Attribution reports in HubSpot is what I turned to. I started diving into our data for closed won business. Back then, HubSpot would ask you to pick an attribution model between first and last touch to build reports off of. For our campaigns and deals, we picked up first touch attribution assuming that was the key touchpoint to start engaging with prospects.
After looking at the reports, I noticed that one particular customer with whom I was familiar with came into the system through a conference, stayed dormant for over a year but was reengaged through an ABM campaign that I ran. The model attributed the success to the conference through which the prospect was sourced, however it completely disregarded the 9 months of active engagement that the sales and marketing team invested in. I knew something was off.
While attribution models help organizations analyze and assign credit to different customer touchpoints to make better informed budgeting decisions or improve campaign performance, the choice of attribution models—whether first-touch, last-touch, or multi-touch—can have significant implications on decision-making.
In this post, I will talk about each of these models in depth to uncover a few flaws and provide you guidance on how you can go about selecting the best fit attribution model for your business.
There are broadly three types of attribution models that are currently in play in most organizations:
- First-touch Attribution: Like in the example above, this model gives 100% of the credit to the first interaction a customer has with your brand. While this approach helps to highlight top-of-the-funnel efforts like SEO or paid ads, it overlooks the customer’s journey and ignores the impact of mid-to-late-funnel interactions like case study downloads, multiple web page visits, email interactions etc.
Here’s another time I found this model falling short of telling the true story. A prospect found our brand through a LinkedIn post (we figured through UTM tracking), signed up for a webinar two weeks later through a paid promotion, and eventually converted into an opportunity after multiple sales follow-up emails. Our report gave 100% credit to the social post.
- Last Touch: This one gives full credit to the final interaction a prospect has with your brand before converting to a customer. This model can be equally problematic, as it ignores all earlier interactions that nurtured the lead.
In the same scenario explained above, if the final conversion occurred after sales emails, last-touch attribution would give all credit to the email follow ups, completely disregarding the previous interactions.
- Multi Touch: This model accounts for different touches a prospect has with your business along the buying journey. Linear, time decay, and data-driven are some forms of multi-touch attribution models that aim to spread out the credit across different touch points. If you’ve used Google Ads, you would be familiar with the data-driven attribution model that they use to calculate the actual contribution of each touch point.
Let’s revisit the earlier scenario: With multi-touch attribution, each of the interactions—LinkedIn post (awareness), webinar signup (consideration), website engagement (nurturing), and sales emails (conversion)—receives partial credit for driving the conversion.
For instance, a linear attribution model might divide the credit equally among all touchpoints, while a time decay model would assign more credit to the later touchpoints (e.g., sales emails) that were closer to the final conversion. On the other hand, a data-driven attribution model would use machine learning to determine the actual contribution of each touchpoint based on historical data.
Only a multi-touch attribution model supports all engagement activities throughout the buying journey that can provide you with a more accurate picture of what your prospects buying cycle looks like and where you should invest your marketing $$$. This is largely true for B2B SaaS enterprises. However, there are nuances depending on your business.
How to Select the Right Attribution Model for Your Business
Choosing the right attribution model depends on several factors that are unique to your business, such as your industry, sales cycle, marketing strategy, and data analytics capabilities. Here are a few questions to ask yourself that can help you determine the most appropriate model for your business:
1. How long are your sales cycles: Businesses with long, complex sales cycles (e.g., B2B, enterprise) typically benefit more from multi-touch attribution reporting because prospects interact with multiple touch points over a long period of time and there are typically more than one stakeholder involved. In contrast, businesses with shorter sales cycles (e.g., e-commerce) might find single-touch models more practical due to a relatively more straightforward path to conversion.
I was consulting for an online retail brand that was selling sunglasses. Unlike B2B, which I was familiar with, the customer journey here was pretty straightforward. We ran a few discount promotions across e-commerce stores and within hours of launch, we would see impact. In such a case it made sense to give all the credit to the first touch with the customer.
2. What are your marketing channels: If your marketing strategy involves multiple channels (email, social, paid ads, PR, affiliate, webinars, etc.), multi-touch attribution is likely the better choice. If you only have a few touchpoints and need to focus on specific entry points, first-touch or last-touch may be simpler to implement.
3. What are your marketing objectives:
- Brand Awareness: If your primary goal is brand awareness, first-touch attribution can help highlight which top-of-funnel activities are causing your new prospects to notice your brand.
- Conversions: For businesses focused on closing sales, last-touch or multi-touch attribution can better indicate what final actions drive purchases or conversions.IKEA or other furniture retailers could be an example. You go to their site when you are looking, if you don’t find a deal, you often sign up for their promotional newsletters, then when you see something at a bargain price, or are retargeted by an ad, you complete your purchase.
- Full-funnel Optimization: If you want to optimize the entire customer journey and not just one part of it, multi-touch attribution is the way to go as it will give you a complete picture of what’s working across all stages.
4. Understand Your Data Capabilities
Implementing multi-touch attribution requires a strong data infrastructure. If you have a really good marketing technology stack that includes a marketing automation platform like HubSpot or Pardot, a CRM software like Salesforce and an Analytics platform like Google Analytics, and Tableau, multi-touch attribution can provide much better insights.
However, if you’re still building your data capabilities, you may want to start with simpler attribution models and transition to multi-touch modeling as your infrastructure and your business understanding gets better.
5. Test and Iterate: Selecting the right model doesn’t have to be a one-time decision. Start by testing different models (first-touch, last-touch, multi-touch) and compare the results. By analyzing the performance under each model, you can determine which one delivers the most accurate insights for your business and audience. For businesses with shorter sales cycles like e-commerce or online retail, the length of these tests could be anywhere between 4-6 weeks, for B2B businesses however where these sales cycles are often months long, tests should be run quarterly or even bi-annually.
In my next post I will cover what an ideal multi touch attribution model should look like in a B2B SaaS environment, what are the different considerations and inclusions Let me know if this helps you at all.
Leave a Reply