Measure Customer Journeys with Marketing Attribution

By Miriam Kung | 30 Apr 2019

A long time ago, marketing leader John Wanamaker said, β€œHalf the money I spend on advertising is wasted; the trouble is I don't know which half.” πŸ€·β€β™‚οΈ Click to Tweet

Centuries later, marketers today still face the same measurement dilemma. Especially if you have many marketing channels and customers using multiple devices - you can imagine how complex it can get, with all the combination of ways someone can learn about you before becoming a customer.

So how do you figure out which marketing channels convert best and improve your ROI? ⬆️

In Part 1 on UTM tracking, we covered the basics of tagging your URLs with UTM tags to see where your website visitors come from. And in Part 2 on Conversion Tracking, we went over how to track conversions on your website to measure your ROI.

We’re now tying everything together with marketing attribution, to help you understand which marketing campaigns and channels perform best and how to optimize your budget πŸ’ͺ.

Below, I'll dive into:

  • What is marketing attribution
  • Common attribution models
  • Tips for choosing the best attribution model for your business
  • How PixelMe's new conversion tracking product in beta can help you understand your digital marketing ROI by measuring ad conversions across every platform

By the end of this, you'll hopefully see that using attribution means:

🀠 You (Marketer/Business): 1

πŸ€‘ Wasted Budgets: 0 🀩

What is attribution?

At its core, marketing attribution lets you determine which marketing channels and touchpoints resulted in conversions (e.g. signup or purchase).

This is done through attribution models, which are a set of rules that assign credit to each conversion along a customer’s journey.

Think of each marketing touchpoint as part of a play in a basketball game πŸ€, where each pass is a touchpoint ⛹️‍♂️. Attribution tells you the full story for each visitor:

  • How someone first got the ball (first touch)
    • E.g. A visitor finds your business through a Google Search Ad.
  • Each ball pass in the play (second touch, third touch, etc.)
    • E.g. A few days later they visit your blog. They do some research that week on similar products, and then visit your website directly.
  • And the slam dunk (last touch)
    • E.g. Finally, they get retargeted with a Facebook Ad, and sign up on your site.

Basketball touchpoints

Why do you need marketing attribution?

With marketing attribution, you'll know which campaigns and channels to - you guessed it... attribute your conversions and success to, and where to increase or decrease your digital marketing budgets.

Attribution budget decrease stats

Marketing attribution gives you the entire story so you can connect the dots from the first time someone interacted with your business, until they became a customer.

That way, you can start shifting your marketing budget into the channels that are working!

What are the different attribution models?

There's plenty of attribution models that exist today, with last touch (aka last click) being the most popular, followed closely by first touch (aka first click).

There's also been a steady increase in multi-touch attribution and custom models.

Attribution models stats

Coming back to the basketball game example, I'll describe 6 different types of attribution models based on the scenario below using the Golden State Warriors team.

Basketball attribution model

A few things to remember:

  • Every model gives a different percentage of credit to each touchpoint in your customer journey
  • The value of your entire customer journey is 100%
  • For this πŸ’£ Warriors team lineup, we have:
    • Player A: Stephen Curry
    • Player B: Demarcus Cousins
    • Player C: Andrew Bogut
    • Player D: Klay Thompson
    • Player E: Kevin Durant

Single touch attribution models

1. First touch

πŸ’‘ What it is: First touch attribution gives 100% of the credit to the first touchpoint that brought a visitor to your website. In this basketball play, Player A (Curry) would get all the credit (not the first time it's happened πŸ˜‚).

First touch attribution model

πŸ”Ž Use cases: Marketers who are focused on top of funnel metrics and goals around lead generation or brand awareness. This model will show you which channels first attract the attention of potential customers.

πŸ’Ž Pros: Easy to implement and track, since you're only looking at one part and assigning all the weight there.

πŸ™ˆ Cons: Since all credit is given to the first interaction, this over emphasizes the value of only one of your marketing channels. It only tells you one part of the story, and ignores all other channels that contributed to the conversion. In basketball, this would attribute the success of a game to only the point guard - in this case, Stephen Curry, who's good but not deserving of all the credit πŸ˜‰.

2. Last touch

πŸ’‘ What it is: Last touch attribution gives 100% of the credit to the last touchpoint that led a visitor to convert. Here, the Warriors would give Player E (Durant) all the credit for the play.

Last touch attribution model

πŸ”Ž Use cases: Marketers who want to understand which channels helped them seal the final deal in getting a customer to convert.

πŸ’Ž Pros:

  • Easy to setup since this is a simple and standard model for the majority of web analytics tools and ad platforms like Facebook and Google.
  • Great for understanding which channels directly led to conversions.

πŸ™ˆ Cons: Similar to first touch attribution, it oversimplifies your marketing efforts by only giving credit to the last interaction. This means if a visitor came directly to your website and converted, but had 2 other touchpoints beforehand (e.g. Facebook Ad and webinar), all other channels would be ignored. Again, it only tells you one part of your customer story.

Multi-touch attribution models

3. Linear

πŸ’‘What it is: Multi-touch linear attribution assigns credit evenly to every touchpoint along a customer's path to conversion. In this example, each of the 5 players get 20% of the credit.

Linear attribution model

πŸ”Ž Use cases: Marketers who use multiple channels and have more complex visitor paths. Instead of looking at one touchpoint only, it clearly paints your entire customer story.

πŸ’Ž Pros: Get more accurate insights into every step your customers took, from their first click to their last click that led them to convert. Getting the full story means you can better understand how each of your marketing efforts contribute to each stage of your customer journey.

πŸ™ˆ Cons: Should a click on a Facebook Ad get the same credit as someone who read a blog post? That's the main downside with linear models, since some touchpoints likely influence conversions more than others and deserve more credit.

4. U-shaped or Position-Based

πŸ’‘ What it is: U-shaped attribution gives 40% of the credit to both the first touchpoint and last touchpoint, and then divides the remaining 20% evenly amongst all the touchpoints in between. In this example, here's how credits would be assigned: 40% to Curry, 40% to Durant, and 6.7% each to Klay, Cousins, and Bogut.

U-shaped attribution model

πŸ”Ž Use cases: Marketers who want all the advantages of linear models, while giving more weight to key moments in a customer's journey.

πŸ’Ž Pros: U-shaped models help you see how different channels impact conversions, and give credit to more potentially influential touchpoints. Specifically, the first touchpoint, which led to discovery of your business, and the last touchpoint, which directly resulted in the conversion.

5. Time Decay

πŸ’‘ What it is: Time decay attribution gives more credit to each touchpoint that's closer to the conversion event. This means the first few touchpoints get the least credit, and the last touchpoint gets the most credit. In the example below, Curry would get the least credit, and Durant would get the most, with Thompson, Cousins, Bogut incrementally each getting more credit.

Time decay attribution model

πŸ”Ž Use cases: B2B businesses that have longer sales cycles, or companies who invest more efforts further down the conversion funnel.

πŸ’Ž Pros: This helps you focus on touchpoints that resulted in conversions, and gives more weight to the ones closer to the bottom of the funnel, which can impact conversion more.

πŸ™ˆ Cons: Similar to last touch models, though time decay gives credit to every touchpoint, it might not account for instances where the first impression and first few touchpoints of a customer's journey may still be highly influential.

6. Custom

πŸ’‘ What it is: Custom Attribution models, as you might guess, let you create your own custom model to fit your business and customer paths.

πŸ”Ž Use cases: Marketers who use multiple channels, have long sales cycles, have complex customer journeys that differ from the standard business, and/or need the flexibility to create their own models. Since the model is entirely customizable, you can assign custom credit as is relevant to each of your customer touchpoints.

πŸ’Ž Pros: Just like custom-fitted suits look and fit the best 🀡, having a custom model in theory is the ideal. It's the most effective way to measure your attribution, as it's designed to be optimized specifically to how your business is setup.

πŸ™ˆ Cons: Creating a custom model is quite complex. It requires a lot of expertise, data, assumptions, and ongoing testing and evaluating to ensure the model is the really the best one for you.

Which attribution models are best?

When it comes down to it, there's really no perfect model that can with 100% accuracy measure every single touchpoint.

But having some models in place, and testing various ones, will get you much closer to improving your marketing investments.

Here are some tips as you look into different models for your business:

πŸ—Ί Map out your customer journey: document all the touchpoints and channels a potential customer could interact with, so you can measure key customer interactions.

πŸ₯… Create clear KPIs: setting up clear goals will help guide you towards models that are best suited for your business. E.g. if you're heavily focused on conversions alone, last touch will be more relevant to monitor than first touch attribution.

πŸ‘† Track the average number of touchpoints before a customer converts: If you're a B2B marketer, Salesforce found it takes an average of 6-8 touchpoints to generate a sales lead alone. So for example, if you have a really long and complex sales cycle, using one of the multi-touch attribution models would be more beneficial to get a clear picture of the impact of every marketing channel vs. single touch models.

πŸ“ˆ Test and optimize: as with most projects, it's a good idea to try out various models based on your goals and customer journeys. In the end, you'll want to stick with models that give you clear and actionable insights. And since your business is always changing and growing, the type of models you use should naturally also change as you evolve.

Getting started with attribution

Now that you're more familiar with attribution and the different attribution models, wouldn't it be great if there was a tool that showed you touchpoints and conversions across each of your digital advertising channels, so you can easily measure your advertising ROI?

By popular demand from our customers, this is just the tool we've built in beta πŸŽ‰.

With our new conversion tracking product, you can:

πŸ“ Track visitors across your website, and see each touchpoint along their journey.

πŸ’° Combine UTMs and conversion tracking to see which traffic sources are most valuable.

πŸš€ Measure conversion events (e.g. signup, add to cart, subscribe) across ad campaigns.

πŸ™Œ Customize your dashboard to see first touch, last touch or multi-touch conversions across every channel.

By signing up for early access, you'll also help shape the product to be most valuable for your needs!

Want to try out our new conversion tracking tool? Just click on the blue chat icon below πŸ‘‡ to sign up for a free 7-day PixelMe trial or email our team at πŸ’Œ

logo twitter