Audience acquisition is becoming more complex as marketing is no longer linear. It is more like an interconnected mesh in which each activity has the potential to influence and contribute to the others.

This fact makes it increasingly difficult to optimise spending, and one of two things happens: You either cull aggressively those activities that you have difficulty measuring when needing to optimise spending or never cut anything for fear that it may be working.

We need to explore better attribution models to efficiently optimise spending for live events, which have the added complexity of sometimes not knowing the final conversion until after the event happens.

Attribution tracking already plays a fundamental role in digital marketing, enabling businesses to ascertain the efficacy of their marketing efforts by identifying the sources that drive traffic, engagement, and conversions. In an era where data-driven decisions are paramount, understanding a consumer's journey from the initial interaction to conversion is crucial—and how we apply these models to Live Events will have to change.

In this article, we explore attribution tracking and touch on the emergence of cookieless tracking, a trend to get ahead of.

It is crucial to understand one important context - The actual conversion for any live event is when the participant turns up and engages. So, accurate conversion tracking is almost always historical unless you apply predictive models.

When is a conversion, not a conversion?

It is essential to get our terminology right; for example, a conversion reported by your digital agency, which equates to someone seeing your online advertisement or clicking on it, is not a conversion - in our language, it is merely a CVI (Commercially Valuable Interaction). While important, we should treat it as a CVI only.

Attribution tracking refers to determining the marketing channels, campaigns, or touchpoints influencing a user to take a desired action, such as registering for an event or signing up for a newsletter. It provides insights into the customer journey, enabling audience acquisition teams to optimise their strategies, allocate budgets effectively, and enhance Return on Investment (RoI). However, in a complex marketing web, simply looking at what drove the last conversion is too simplistic.

Demystifying Attribution Tracking Definition and Importance

In this section, we review the main attribution models and discuss some vital considerations for making them work for Digitising Events at the end of this article. 

Single-Touch Attribution Models

First-Touch (or First-Click) Attribution - Assigns 100% of the conversion credit to the first touchpoint that a prospect interacts with, such as a click from an advertisement or a visit from a social media post.

Imagine you saw an ad for a product on Facebook, clicked on it, but bought it only after seeing a few more ads over the next week. First-touch attribution gives all the credit for your purchase to that first ad you saw on Facebook. We already see an obvious flaw, but there is validity in this approach as it gives you an idea of how much you spent to get that first engagement.

Benefit - It's straightforward. It helps in understanding which channels are effective in attracting customers initially.

Shortfall - It ignores all other interactions you had with ads after the first one, which might have influenced your decision to register.

Last-Touch (or Last-Click) Attribution - Gives all the conversion credit to the last interaction a registrant had before converting, such as clicking on a link in an email.

In the same scenario, last-touch attribution credits only the last ad or email you interacted with before registering. It also has a problem as it negates what drove the first interaction.

Benefit - Easy to understand and implement. It shows what finally convinced you to make a purchase.

Shortfall - Like first-touch, it overlooks all the other interactions you had along the way.

Multi-Touch Attribution Models

Linear Attribution: Distributes the conversion credit equally among all the touchpoints a registrant interacted with during their customer journey

Here, each ad, email or social post you interacted with gets an equal share of the credit for a registration.

Benefit - It acknowledges every interaction you have, providing a more holistic view.

Shortfall - It assumes each interaction was equally influential, which might not be the case.

Linear Attribution

Put differently, if you had a registration for a conference where a delegate paid, say, £2,000 to attend, and you were able to measure four interactions driven by your outreach, you would attribute £500 to each of those channels.

Position-Based Attribution Models: These models build on the linear model by allocating a certain percentage of credit to the first and last interactions, with the remaining credit distributed equally among other touchpoints. A typical distribution is 40% credit to both the first and last touch and the remaining 20% spread out among the middle touches.

This model gives more credit to the activity that drove the first and last touches and less to the ones in between; this makes sense as the most inertia is in acquisition and conversion.

Position Based Attribution

Benefit - It values the importance of first and last touches, which have higher inertia, while still considering the role of middle interactions in nudging the journey.

Shortfall - The preset percentage split might not accurately reflect the actual influence of each interaction.

Time Decay Attribution Models: These models give more credit to the touchpoints closer to the conversion, assuming that interactions closer to the conversion in time are more influential in the decision-making process. It's like when your boss calls a deal you have been working on and uses their position to close it.

The closer an interaction driven by a message is to your registration, the more credit it gets.

Time Decay Attribution

Benefit- Logically, the interactions closer to the registration might have a more significant influence.

Shortfall- It may undervalue the importance of initial interactions that sparked a registrant's interest.

This method can also generate false positives in situations where the decision-making cycle is longer. For example, in B2B events, delegates may need to go through several steps within their organisation before they can buy that delegate pass.

Algorithmic (or Data-Driven) Attribution Models

Utilises machine learning and statistical modelling to assign conversion credit based on each touchpoint's impact on the conversion.

It's like having an intelligent system that analyses all your interactions and determines how much credit each pne should get based on how much it contributed to a registration.

Benefits- It's data-driven and can provide a more accurate picture of what's working and what's not.

Shortfalls- It is complex and requires a significant amount of data to work effectively, as evidenced by our learning when creating The DiG 

Custom Attribution Models

first or last), while multi-touch considers all or most touchpoints in a customer's journey. Custom models are some combination of the two.

Each model provides a different perspective on the effectiveness of marketing channels and campaigns. The choice of attribution model can significantly affect the interpretation of marketing ROI and the subsequent allocation of marketing budgets.

Challenges in Attribution Tracking for Digitising Events

In addition to which model or combination of models you adopt, attribution tracking has other challenges, such as data silos, cross-device tracking difficulties, cross-domain tracking, e.g. iFrames and the complexity of multichannel interactions, which can obfuscate the clarity of the customer journey to your events or webinars.

However, what remains true is that embracing a more data-driven approach away from the single-touch models will make it easier to tune and optimise your acquisition budget.

Segue: The Cookie Crumbles

We need to mention the decline of cookies, as this can impact how we track and attribute conversions, mainly if you rely on third-party services; however, if you run registrations under your domain, you will minimise the impact as the key target is third-party cookies. With growing privacy concerns and regulatory changes, such as the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA), the utility of 3rd party cookies is diminishing.

The clear message is that if you rely heavily on third-party cookies, you will need to consider mitigating against this.

Wrapping it up: Attribution Tracking Models for Digitisised Events

Attribution tracking is more complex for Live Events because we only know the actual conversion once the participant attends and engages with the event - especially true for free-to-attend events. It impacts the proper judgement of conversion as:

1. The value of a participant significantly increases the more they engage at the event - particularly with sponsors.

2. There is a latency in judging the actual effectiveness of your acquisition channels.

3. Any participant who does not attend is not effectively a conversion.

We can mitigate for this reality by building machine learning to predict attendance and behaviour, to give advanced resolution on the effectiveness of channels.

Custom attribution models, though complex, are the most effective for live events. For example, LinkedIn and other channels use time-based attribution to justify their value, which is not really a conversion for the event owner—and in fact, with our work, we have seen that it does not correlate well to attendance, so money is wasted.

The challenge in the context of attribution tracking for Digitising Events is that all traditional digital attribution models assume near-to-real-time feedback of the true conversion, buying something or signing up for a trial - that is not the case for a live event and will never be.