With decisions on budget, channel splits and platform prioritization being made on such a regular basis, data accuracy is critical to success. By using data-driven decisionmaking, your brand can take the guesswork out of optimizing your digital campaigns.
But, how can we all ensure that the data we’re relying so heavily on is accurate? The answer is twofold: attribution modeling and accurate conversion tracking. Let’s take a look at what we can do as brands and marketers to better understand the impact of our marketing, and how to leverage the results we’re achieving for the best possible ROI.
1. What Is an Attribution Model?
2. What Are the Benefits of Attribution Modeling?
3. What Kinds of Attribution Models Are There?
4. Which Attribution Model Should I Use?
5. How Should I Track Conversions From My Digital Marketing Activity?
Final Thoughts on Attribution Models & Conversion Tracking
Attribution models are a methodology used to understand which platforms and/or touch points contributed to a conversion. The conversion that’s tracked doesn’t need to be a sale: a conversion could be a newsletter sign-up, a contact form submission, a phone call, a store visit, or any other action that you want your customers to take online or offline.
At a more granular level, attribution modeling can be used on a platform-by-platform basis to understand which ads and assets are performing best. For example, on Google Ads, conversions can be measured at the campaign, ad group, ad, asset and even keyword level to understand which of your inputs are the most effective in generating conversions. This can help you to effectively optimize your digital strategy and campaigns with maximizing conversions in mind.
Across all digital marketing platforms and analytics tools, conversion reporting is critical. While metrics such as CTR, engagement rates and CPC can give an insight into marketing effectiveness and efficiencies, conversion volume, conversion rates and cost per conversion metrics are a prime indicator of what’s working well, and what isn’t working so well within your campaigns. But, without an understanding of attribution modeling, it’s almost impossible to leverage conversion-based insights.
With complete conversion tracking and accurate attribution modeling in mind, you’ll reap the following benefits:
Below is a breakdown of various attribution models, and how credit would be assigned in the following customer journey scenario:
Customer X’s journey;
The data-driven attribution model takes all known touch points across search, video campaigns and other marketing touchpoints and mathematically assigns accurate credit to each across your full account. Also known as the ‘Algorithmic’ attribution model on Adobe Analytics, data-driven attribution is unique to each advertiser and will take numerous conversion paths into account.
In Google’s own words: “By comparing the paths of customers who convert to the paths of customers who don't, the model identifies patterns among those ad interactions that lead to conversions. There may be certain steps along the way that have a higher probability of leading a customer to complete a conversion. The model then gives more credit to those valuable ad interactions on the customer's path.”
The first (ad) click or first interaction attribution model assigns full credit for the conversion to the first click between the customer and the brand. Using the customer journey example above, step one, the first Instagram ad, would receive 100% of credit for the conversion as it ‘introduced’ the customer to the brand initially.
The linear attribution model assigns equal credit to all touchpoints between the customer and brand before converting. Using the example above, all three touch points would receive equal credit for the conversion: 33% each.
The inverse of first click attribution, the last (ad) click or last interaction attribution model assigns the full credit for a conversion to the last (ad) click taken by the customer before converting. In the example above, the fourth touchpoint - the remarketing ad on Google - would receive full credit for the conversion.
The position-based attribution model is a little more complex. Prioritizing both the first and last touchpoint to calculate conversion credit, the position-based attribution model assigns credit in the following ways:
If there is one interaction between the first and last touchpoint, this will be assigned 20% of the credit. If there are four interactions between the first and last touchpoint, each one will be given 5% of the credit towards the conversion.
In our customer journey example above, the initial Instagram ad, and final display ad would receive 40% each, and the middle touchpoint of the Meta ad would receive 20% credit.
Finally, the time decay attribution model. This model assigns more credit to touchpoints that were closer to the conversion than others. In the example above, the final touchpoint may receive 50% of the credit for the conversion, whereas the middle interaction may receive 30%, and the first interaction would receive 20%.
Finally, custom attribution models. First of all, we wouldn’t suggest creating your own attribution model unless you had full developer support, as they’re very easy to get wrong! However, if you have a broad history of data and you know the triggers that often lead to conversion that don’t fit any of the default attribution models, you can create your own. This method also works well for omnichannel brands that have offline and online sales as well as utilizing data imports for conversion and sales information.
The settings to consider when creating your own attribution model are:
One long-lasting issue that digital marketers and business owners have had to overcome is choosing the right attribution model for their activity. Historically, platforms such as Google Ads and Meta (Facebook & Instagram) Ads have used a last-click attribution model. This attributes 100% of the ‘credit’ for a conversion to the last click that a customer took before converting. However, the last click attribution model isn’t necessarily reflective of all the touchpoints between a brand and a customer.
Based on a last-click attribution model, you could be tempted to allocate 100% of your digital marketing budget towards a lower-funnel form of marketing such as Search, missing out on crucial upper funnel and brand awareness activity through channels such as Display, Video or Social. Unfortunately, it’s been shown time and time again that, without topping up the top of the funnel with new prospects, the funnel dries up very quickly, and clients will see a lack of results overall after a short period of time. In this scenario, inaccurate or biased attribution modeling has led to a sharp shift in budget splits and an undesirable outcome.
Another word of caution is to bear in mind ‘lookback windows’. If the last click before conversion took place on 01-01-2024 and a conversion took place on 01-14-2024, a lookback window of 7 days would not assign credit to the click that happened on the 1st of January - as it fell outside of the lookback window. Lookback windows are also often set differently for views and clicks: by default, the lookback window for click activity is often 7 days, and up to 30 days for ads that were viewed but not engaged with. However, this will also vary sector-by-sector: brands who sell high ticket items such as houses, cars or boats will experience longer sales cycles by default than those who sell coats, beauty products or jewelry.
So, which attribution model should you use? If accuracy and a complete view of the customer journey is important to your business, the data-driven attribution model is a no-brainer.
However, before making the switch, if you’re using Google Analytics as your SSOT (Single Source of Truth) for data and conversion modeling, their Model comparison tool has the ability to show you ahead of time how changing your attribution model will affect the valuation of your marketing channels. You can use this tool to compare up to three different attribution models at a time to compare results and how they impact the ‘value’ of each of your marketing channels.
Whether you’re using platform-specific attribution modeling or an overarching analytics tool such as Google Analytics 4 or Adobe Analytics, you’ll need conversion tracking in place. As mentioned earlier in this article, conversions can take many forms, and you can track multiple conversions as part of your advertising efforts.
The most straightforward way of tracking conversions is with thank you pages. While this doesn’t work for tracking conversions such as phone calls or store visits, thank you pages can be used for form submissions and online store sales. Thank you pages or confirmation pages are usually only accessible after a successful conversion, making them a very accurate form of conversion tracking.
If your conversion isn’t as clear-cut, you can also use button clicks and interactions on your website to track conversions, too via tools such as Google Tag Manager. These can then be imported into Google Ads. Meta and TikTok Pixels have also come a long way in the last few years, and will allow you to track a wide range of interactions and conversions that take place on your website.
If you’re a frequent Google Ads user, you’ve probably heard of ‘Enhanced conversions.’ In Google’s own words, enhanced conversions is a feature that “can improve the accuracy of your conversion measurement and unlock more powerful bidding. It supplements your existing conversion tags by sending hashed first-party conversion data from your website to Google in a privacy-safe way.”
Essentially, the function allows you to both create and utilize your hashed first-party data to better measure conversion paths and actions.
‘Enhanced conversions for web’ allows you to better match separate conversion activities. For example, enhanced conversions allow you to attribute multiple conversions and events to a final sale: perhaps the user signed up for your newsletter, added 3 items to their cart, and then visited the website from a remarketing ad to complete the final conversion (sale). Enhanced conversions better maps these ‘lighter’ events to the final conversion, giving your campaigns more accurate signals to optimize towards when driving conversions.
‘Enhanced conversions for leads’ perform a similar function: allowing advertisers to measure both online and offline actions from users who submitted a lead through a website. For example, if a user submits a lead form, and then contacts the business separately or ‘offline’ (i.e. via call or email), enhanced conversions can use the hashed user data to attribute these actions back to the original lead form submitted via Google Ads activity.
In a nutshell, enhanced conversions gives you better visibility of the full range of interactions between a customer and a brand ahead of conversion. Instead of attributing all conversions to the final interaction that took place before a sale, enhanced conversions take into account multiple touchpoints using first-party, hashed data giving you a more complete view of the customer journey, conversion paths and delivering more signals for your campaigns to optimize towards.
While attribution modeling can sound like a minefield, it’s often something that goes unnoticed day-to-day. But, it’s a crucial factor to understand when optimizing campaigns and during data-driven decisionmaking. Without getting it right, you may end up making disastrous decisions for your brand.