The term ‘last click attribution’ describes a customer journey measurement system that assumes only the final action before a purchase contributed to a sale and deliberately ignores all the previous steps and influences that might have preceded it.
In around 15% of all online sales, there is only a single step in the customer journey. In these cases, the single click clearly affected the sale.
However, our research has found that 85% of customer journeys involve multiple channels and take place over longer periods. So, where should the credit lie, and why should it be the final touch point, as a last-touch attribution model would assume?

Another reason why last click attribution is misleading is that it ignores all the indirect channels that may have influenced the sale, such as TV or Press. That also includes any marketing activity concerned with brand development and not aimed at a short-term response.
However, as many marketers persist with using last click, we decided to undertake some analysis to find out just how misleading it was (and what a better alternative model is).
In this Article:
- How Does Last Click Attribution Work?
- Pros & Cons of Last Touch Attribution
- How Effective is the Last Click Attribution Model?
- Phase 1 – Weighting the Customer Journey Using Machine Learning
- Phase 2 – Comparing Values Attributed By Last Click vs Machine Learning
- Phase 3 – The Verdict
- What Model is Better Than Last Click Attribution?
- How to Get Started With Multi-Touch Marketing Attribution
- Conclusion: Just How Wrong is Last Click Attribution?
- FAQs
How Does Last Click Attribution Work?
The last click attribution model simply assigns 100% of the credit for a sale to the final touch point. In this way, the last channel through which a customer comes into contact with your brand is given all the credit for driving that sale.
This model works on the assumption that customers are most likely to convert when exposed to a particular marketing channel or tactic immediately before making a purchase. So, if a search ad or an email was clicked on and then led directly to a purchase, it would receive all the credit for that sale.
We’ve already alluded to the fact that customer journeys are more complex than this simple attribution model suggests, and frankly, last touch attribution doesn’t accurately reflect the reality of how customers interact with brands and make purchasing decisions.
Pros & Cons of Last Touch Attribution
| Pros | Cons |
| Requires minimal setup and data analysis compared to more complex attribution models. | Fails to account for the multiple touch points (e.g., discovery, consideration) that influence a consumer’s decision before the final click. |
| Doesn’t require advanced tools or complex tracking infrastructure, making it a cost-effective choice for small businesses. | Neglects the role of upper-funnel activities like social media engagement or awareness-building ads. |
| Integrated into many analytics tools (e.g., Google Analytics), making it easy to apply universally across marketing efforts. | Can result in overinvestment in channels driving last clicks and underinvestment in channels that contribute to the awareness or consideration stages. |
| Useful for campaigns where the goal is immediate conversions, such as flash sales or retargeting efforts. | Wrongly encourages a transactional approach to marketing rather than building long-term brand equity through a holistic strategy. |
Read more: Is There a GA4 Alternative? Yes, & It’s Here
How Effective is the Last Click Attribution Model?
To prove just how misleading last click attribution is, we’ve done some of our own investigating.
We took data from a large UK retailer who uses a wide range of direct marketing channels and examined all their customer journeys that led to sales in a three-month period.
Here’s how we carried out our research and what we found…
Phase 1 – Weighting the Customer Journey Using Machine Learning
The process we used to compare to last click attribution was to apply a weighting to every step in a customer journey based on machine learning (ML).
The ML approach looks at a number of factors describing each journey step, and the principal ones are to do with intervals before and after each step.
So, for example, if a customer is sent an email, opens it, but then does nothing about it for a week before visiting the client’s website, that email will be given a much lower weighting than if the customer had clicked through from the email to the website soon after they received it.
Using this approach, we can give a comparative weighting to every step in every journey.
In the case of this retailer, there were around 1m journeys in the quarter we examined to track and weight.
Here is an example of the weighting being applied to each step in a customer journey.

In fact, our algorithms have gone one step further and worked out the contribution made by each step in initialising, holding, or closing a sale.
So, here, direct mail gets most of the initialising score, and email has the biggest share of the closing score. But each step gets some share of the sale, as you will see from the column on the far right of the table.
Phase 2 – Comparing Values Attributed By Last Click vs Machine Learning
Now, we move on to comparing the last-click approach with machine learning that looks at all steps in the journey before a sale. You will see that very substantial differences emerge.
The table below shows the comparison between the value given to channels using last click, and the value attributed by our machine learning.


The differences are considerable. For instance, last click over-values PPC by 22% or, in this case, £1.1m, and under-values direct mail by 13% or £5.1m.
Phase 3 – The Verdict
So, in conclusion, last click is quite simply an extremely inaccurate way to attribute value to marketing campaigns and can lead to a serious misallocation of resources.
It should be avoided at all costs, otherwise marketers risk making decisions based on misleading data. Over- or under-investing in certain channels because last click attribution tells you to do so can ultimately harm your overall marketing efforts and results.
What Model is Better Than Last Click Attribution?
An approach looking at every journey step, and weighting them according to their role, is going to give a much more balanced result, and lead to a more optimised allocation of marketing media spend.
A multi-touch attribution model supported by machine learning is a markedly better alternative to last click attribution. It takes into account all touch points in the customer journey and assigns appropriate weights based on their contribution to driving conversions.
Furthermore, considering the entire customer journey gives a more holistic understanding of the effectiveness of marketing efforts and allows for better decision-making when it comes to budget allocation and campaign optimisation.
Marketing mixes shouldn’t be looked at on a granular, channel-isolated basis, but rather as a collective effort that works together to drive results.
How to Get Started With Multi-Touch Marketing Attribution
Multi-touch attribution can be a tricky nut to crack, which is why partnering with an attribution vendor who offers this solution is your ticket to success.
At UniFida, we help businesses accurately measure the effectiveness of their marketing channels and campaigns through our advanced attribution model – which takes into consideration all touch points and weights them accordingly.
We use our in-house machine learning algorithms to analyse each customer journey and assign appropriate values to each step, always trained on our client’s own data, rather than any generic model.
With our solution, you’ll have access to a full analysis of your customer journeys and the contribution made by each event, empowering you to make data-driven decisions for your future marketing efforts.
Ready to get started? Use the button below to send us an email or give us a call.
Conclusion: Just How Wrong is Last Click Attribution?
The bottom line is last click attribution is flawed and leads to inaccurate measurement of marketing efforts. It’s time for businesses to move on from this outdated model which risks hindering their marketing performance and results.
Leave behind the misconceptions and embrace a more advanced, holistic approach to attribution that considers all touch points and values them accordingly.
FAQs
Julian Berry is an accomplished marketing technology leader. Julian spent his early career working directly under renowned direct marketer Christian Brann. He then held senior marketing roles at NatWest and LTSB before establishing several successful consultancies. He founded UniFida in 2014, and pioneered multi-touch attribution platforms that help marketers measure and optimise marketing value across channels.


































