B2B Marketing Blog

Everything You Need To Know About Marketing Attribution

By Jordan Con
Aug 31, 2016

More and more, CMOs and other marketing leaders are valuing marketers who are able to prove their impact. Whether that’s proving the ROI of the new marketing technology that the team invested in or proving that the search ads are generating revenue, businesses want to know that their investments are paying off.

They also know that from a technology standpoint, the data to prove it exists. Marketing attribution offers the ability to connect marketing efforts to revenue, enabling marketers to demonstrate their effectiveness with data.

In this article, we’ll walk through everything you need to know about marketing attribution -- how it works, from touchpoints and tracking through to modeling and reporting. We’ll also walk through tons of examples to demonstrate exactly what kind of data attribution is able to produce and see what it looks like. By the end, you'll know how marketing attribution works, as well as the models and reports that you need to prove your value.

everything-marketing-attribution.png

Touchpoints and Tracking

A touchpoint is an engagement -- online or offline -- that you have with a prospect. This could be an organic search that results in a click on one of your articles, a click on a LinkedIn ad, or an interaction at a trade show conference.

Touchpoints are the basis of marketing attribution.

A simple marketing attribution solution may just track a single touchpoint. At the other end of the spectrum, an advanced marketing attribution solution tracks every touchpoint and then applies a model to determine the amount of credit each touchpoint receives. We’ll get into this on the section on modeling.

Each touchpoint should also have several pieces of information: a marketing channel, a landing page URL, a touchpoint position, and potentially more.

Here is a simple customer journey that we will use as an example throughout the entire article.

I was searching for new marketing tools, when I was served an AdWords ad for marketing boats. (What’s a marketing boat, you ask? Good question.) The ad took me to a blog post that explained some of the benefits of marketing boats. Loving the idea, I shared the idea with the marketing team at our next meeting. After the meeting, Alexis and Andrew both searched for the marketing boat company by typing in the URL of the company. Lauren didn’t visit the site directly, but instead searched for it on Google, which linked her to the product page. She browsed the site and then downloaded an ebook on the benefits of marketing boats. At the next meeting, the team decided to pursue the idea. Clicking on another AdWords ad, Jordan decided to move forward and request a demo. Over the next two weeks, Alexis went to a boat show that was in town and Andrew read another blog article that he found on LinkedIn. Deciding that a marketing boat would significantly help the team, they brought in Dave, the department leader, to close the deal.

Here’s what that journey looks like in table format:

customer-journey-touchpoints.png

Let’s discuss some of the elements here. The marketing channel is the channel that drove the touchpoint. If a person searches on Google and then clicks on one of your articles, the marketing channel for that touchpoint would be Organic Search.

The landing page is the first webpage that the visitor views on that particular session. In the example, the landing page would be whatever article they clicked on from the Google SERP (e.g. http://marketingneedsboats.com/blog/why-you-need-marketing-boat).

The touchpoint position describes where in the customer journey the touchpoint took place. This could be the very first touch, the touch that led to the lead conversion, the opportunity conversion touch, or the closed won or lost touch. Alternatively, it could also just be a simple web visit, where no lead stage transition occurred.

The attribution model that you choose will have an impact on which touchpoint positions are a possibility. For example, if you’re using a U-shaped attribution model (which we will get to later), the only possible touchpoint positions are first touch, lead creation touch, and just a simple web visit. That’s because the U-shaped model doesn’t take into account any action after lead creation.

Additionally, if there was a form fill during the session, the touchpoint should also have that information. In addition to the landing page URL, it should have the URL of where the form was filled. We call this the Form URL. This can be the same page, but it doesn’t have to be. Here’s what that would look like:

customer-journey-touchpoints-form-url.png

Touchpoints and the descriptive data around it are the building blocks of marketing attribution. Tracking touchpoints is what helps marketers measure their audience’s response to their efforts and the descriptive data enables granular analysis. We’ll get to this analysis later, but first we must understand how marketing attribution gets its data.

Technology Integrations

When it comes to producing the best data, marketing attribution is only as useful as the data that goes into it. A good marketing attribution solution will handle a significant amount of data collection, likely through on-site JavaScript code. But it also must play well with the rest of the marketing stack.

Marketing channels -- how your brand and content gets distributed -- contain a lot of critical information. More than just what channel is driving traffic to your website, your marketing attribution should be able to determine which campaign, ad, or keyword drove the traffic, so that it can follow the visitors actions and attribute credit back to the specific marketing effort. While this typically requires careful and thorough use of UTM parameters, it is made much more convenient and easy by integrating with major ad networks.

Integrations can also add levels of data that are unobtainable through UTM parameters. An example of this is tracking view-through conversions. When using display ads, it’s possible that a prospect sees your display ad on a blog, continues reading the blog article, and then visits your website directly (types in your URL). View-through tracking is able to determine that you viewed the display ad and then, within a specific time frame, visited your website; therefore, the display ad should receive credit for that visit.

Additionally, a marketing attribution solution should complement your marketing automation software. Both lead nurturing and landing page actions are critical components of the customer journey, so it is also critical that the data be accurately included in attribution. Integrations ensure that accurate and complete data is transferred.

Finally, and most critically, marketing attribution must integrate with your CRM. The CRM holds all sales and customer data. Therefore, to accurately connect marketing to revenue (attribution, by definition), marketing attribution needs to integrate with the CRM.

The Definitive Guide To Pipeline Marketing  Everything you need to know to be a revenue-focused B2B marketer. Read more

Attribution Data Modeling

With a complete data set of the customer journey -- tracking all of the touchpoints and accurate data from marketing channels, marketing automation, and the CRM -- the next step is to represent the journey with numbers through modeling.

There is a wide spectrum of possible ways to model attribution data -- from simple single-touch models to complex custom models that require advanced knowledge in data science.

Some marketing attribution products don’t capture every touchpoint. They may just track the first touchpoint or the touchpoint right before a key conversion, like the lead conversion or the opportunity conversion. In this case, the attribution modeling is simple. They give 100% of the credit to that single touch. While it’s a good start and better than nothing, single-touch attribution models clearly misrepresents the customer journey that is far more complex than one touchpoint.

Using the customer journey example from earlier, this is what the data would look like using the various single-touch attribution models.

customer-journey-single-touch.png

Multi-touch attribution modeling is the next step. Now there are also tons of different ways to use multiple touchpoints to model your attribution data. The simplest is the linear model, where every touchpoint receives the exact same percentage of credit. There’s also the time decay model, where touchpoints closest to the conversion get more credit.

And then there are models that give different weights based on the prospect’s stage. In the U-shaped model, more credit is given to the very first touchpoint and the touchpoint that transitions the anonymous visitor into a lead (40% to the two key touchpoints, 20% split between the rest). Because this model stops at lead creation, it’s best used for creating lead reports, and not opportunity and revenue reports.

In the W-shaped model, more credit is given to the two touchpoints included in the U-shaped model plus the touchpoint that transitions the lead into an opportunity (30% to each of the three key touchpoints, 10% split between the rest). In both models, all the other touchpoints that don’t involve a stage transition receive less credit. Because the W-shaped model tracks opportunity conversion, it’s a great model for marketers who care about down-funnel metrics, like opportunities and revenue.

Here’s what the data from our example customer journey looks like using these various multi-touch models.

customer-journey-multi-touch-attribution.png

You can read more about every attribution model here.

Additionally, if you have a unique customer journey or just want to get more granular with how you distribute credit, it may be worthwhile to look into custom modeling. This requires deep analysis of your existing customers’ journeys and creating an algorithm to precisely weight the impact of each touchpoint. This could mean giving more weight to certain channels no matter where in the journey the touchpoint occurred, or giving extra credit to the MQL stage transition and the SQL stage transition rather than the lead stage transition.

Attribution Reporting

With an attribution model that you trust to accurately represent your customer journey, you can now start to build powerful marketing reports. These can be for the marketing department as a whole or for specific marketing functions.

At the department level, you can run reports that answer the questions like, “What’s the number of opportunities driven by each channel this month?” And if you’re using your multi-touch attribution model, you’re able to answer questions like these with fractional data that’s a far better representation of reality. For example, Paid Search may have influenced some part of 100 opportunities this month. Because you’re using a W-shaped model, you would know all the partial credits for those 100 opportunities. Summing up all of the partial credits, you would then be able to say that it was directly responsible for 22.4 opportunities.

The same type of report can be made for pipeline revenue and actual revenue.

marketing-attribution-channel-report.png

Specific marketing functions will also want to use attribution data to report on their performance. Let’s use the content marketing team as an example. Here at Bizible, we use attribution data to determine the success of our blog posts and content downloads, like ebooks and reports.

As you remember from earlier, each touchpoint should have a landing page URL associated with it. Because we have that data, we are able to see how many leads, opportunities, pipeline, and revenue was generated based on the individual blog landing page and where in the customer journey our prospects viewed the blog.

Here’s what a lead report could look like with touchpoints that had a blog landing page:

marketing-attribution-blog-report.png

Because this is a lead report, we’re using a U-shaped model that gives 40% credit to the first touch, 40% credit to the lead creation touch and splits 20% credit among the rest of the channels. So if a person viewed a blog post as their first touch and then later, maybe three or four touchpoints down the road, converted into a lead, the blog post would receive credit for creating 0.4 leads.

When analyzing this report, you could see which blog posts are driving the most leads, and how they are performing over time. For example, you would want to dig into what type of blog article “/blog/article6” was because it was effective at driving leads overall. How can you replicate what made that blog article so successful? You might also be interested in “/blog/article5” because it is driving consistent leads over time. Could this be good evergreen content that is succeeding on organic search?

This leads us to the ability to filter reports based on the touchpoint data. An insightful report is one that filters this data by marketing channel. This would help you understand what type of content is working via social media, organic search, or email.

Similar reports can be made for email marketers (which emails influenced or directly generated the most business) or paid media marketers (which campaigns, ads, etc. influenced or directly generated the most business), or any other function.

Marketing Attribution for ABM

In the example that we’ve been using throughout this article, all of the analysis that we’ve been able to demonstrate is made possible with marketing attribution that has ABM reporting capabilities.

That’s because, in case you didn’t notice, there were multiple people involved in the customer journey. Let’s take a look at the customer journey again.

customer-journey-touchpoints-1.png

With a marketing attribution solution that doesn’t take into account the account-based nature of this sale, the data would have looked a lot different. Rather than being account-based, it would be lead-based and that means treating people as individuals up until the opportunity stage. That means, until I (Jordan) requested a demo, the four of us would have had separate and individual journeys. Paid Search would be credited for a first touch (Jordan), Direct would be credited for two first touches (Alexis and Andrew), and Organic Search would get credit for a first touch and lead creation (Lauren).

With lead-based attribution, it’s not Bizible’s first touch; instead, it’s Jordan’s first touch and Alexis’ first touch and Andrew’s first touch, and Lauren’s first touch.

Account-based attribution is able to determine that we are all from the same company and belong to a single account. That way, the attribution data can tell a more accurate story of what is really going on here. For example, there’s only one true first touch, not four.

From the perspective of the sales team, this understanding helps, too. When they call into our company, they can see the whole picture and have more relevant conversations.

The Definitive Guide To ABM Measurement  Effective ABM requires effective ABM measurement. Learn more

Conclusion

Marketing attribution is a powerful piece of the marketing stack. It brings the data together into a single source of truth and helps marketers accurately understand which of their efforts are contributing to down-funnel metrics, like revenue.

When marketers are able to measure their performance in terms of opportunities, pipeline, revenue, and ROI, businesses succeed. That’s because marketing is more aligned with the sales team (speaking in terms of revenue), they’re able to optimize their efforts for revenue, and ultimately, they’re able to drive more growth.B2B Marketing Attribution 101  An intro guide to attribution for revenue-driven B2B marketers  Download Now

  New Call-to-action