3 Steps to Realizing Multi-Touch Attribution in Your Organization

As a digital marketer, you have more channels to acquire and engage with prospects than ever before and there are countless postings discussing how many marketing touches and responses it takes to generate MQLs, SQLs and new customers. Tracking all this activity in a systematic manner that yields results however requires a solid gameplan and potentially a lot of work either building or overhauling your campaign tracking infrastructure. So let’s start by reminding ourselves why striving for real multi-touch attribution reporting capabilities is worth the effort and can radically transform your marketing organization.

  • Optimize your marketing spend based on data rather than intuition
    Instead of playing the budget guessing game each year, you’ll be able to make informed decisions like whether to renew with a list purchase vendor, or invest more in trade shows because you’ll know which channels are better at leading to pipeline/revenue. And individual channel managers can also see how their specific programs are performing – did digital ad A perform better than digital ad B?
  • Demonstrate marketing is a revenue center, not a cost center
    It’s hard to believe this is still a battle being fought, but one of the metrics you’ll now have at your fingertips is the percent of total pipeline and revenue being influenced by marketing. This can be used in challenging times to preserve marketing budget or staff because you can now make a convincing case that cutting either will result in a future reduction in pipeline and revenue.

Step 1: Identify All Your Channels and What Constitutes a Response for Each

Email is the most straightforward channel to explain this concept and Marketo out of the box gives you the framework to get started. Every Marketo program must have a designated Channel. Each channel has Statuses that are either considered a Success or not. Success in this context is another way of saying it’s a response. In the case of email, you could decide that a click is a response, while an open is not.

Go through this exercise with all your channels, noting all the relevant statuses within each and which ones are considered successes, or responses. Then have your Marketo admin go into the Tags section and add/update values to match what you’ve mapped out.

Step 2: Attach All Responses to the Leads and Contacts in Your CRM Platform

As a Marketo customer you may be wondering why this is necessary since it’s the initial collection point for all prospect response data. And Marketo has their Opportunity Influence Analyzer right? This tool is great for studying the customer journey on individual opportunities but it cannot scale to answer the really important questions marketing teams are looking to answer, like what percent of total pipe/revenue is being influenced by marketing, or what channels are working best at various stages in the funnel.

The reason you want a CRM-centric reporting scheme is due to the fact that this is where your opportunity data resides, and this is the starting point for any type of attribution tracking scheme. Basically, it’s an exercise in working backwards – an opportunity is associated to an account, which has associated contacts, and those contacts have campaign responses attached to their records. Your reporting solution must be able to connect all these dots. In the final section of this post I’ll discuss reporting options, and many will only work if all your data resides in this single CRM database.

Using our email example from the first section, you’ll want to create corresponding CRM campaigns for each email “program” you want to track. In the screenshot example below, we’re sending a one-off email out of Marketo, a March newsletter, and we have one CRM campaign specifically to track all responses (clicks) on this newsletter. Alternately, you could have multiple emails in a single nurturing program that connect to a single “Email Nurturing XYZ” campaign in your CRM system.

As with your Marketo programs, your CRM campaigns need to have channel types and statuses that map to what you set up back in Marketo. In SFDC, the Type field on Campaigns works as a good location for the Channels set up in Marketo and you can set the Status values inside your CRM campaigns to match – Clicked Link in this case.

The goal is to pass all campaign responses you collect on a lead record in Marketo to the Campaign History section of the corresponding lead or contact record in the CRM system. If I’m a prospect who came to your site via a paid search ad click, downloaded a white paper and later clicked a link in an email you sent me, that’s 3 separate responses, and therefore we would want to see 3 campaign response entries on my record in the CRM system.

3) Select the proper reporting solution for your organization

Once this response pass-thru framework is in place, you have the basis to track multi-touch attribution. Since the data is all in your CRM platform, in a perfect world, the CRM system’s reporting tool would have the power to build all the reports you need but they typically short in several key functional areas outlined below.

Traditionally, the biggest roadblock for smaller organizations with these types of tracking ambitions was they lacked the budgets to acquire enterprise-level BI platforms and internal developer resources to build custom data warehouses and reporting dashboards. Having this level of customization and flexibility may still be the preferred choice for larger enterprises, but the good news is that within the past few years several turnkey solutions have entered the marketplace that put this type of reporting capability within reach, and can start providing value very quickly. They have the typical limitations of a turnkey solution – can only pull data from one system (SFDC), can’t build custom reports, etc. but they are designed with the reporting needs of marketers in mind.

Regardless of your organization’s size, there are 3 key requirements to keep in mind as you evaluate vendors or reporting solutions to meet your team’s needs.

  • It must be able to capture/display campaign responses for ALL contacts at an account?  In SFDC for example, there’s the Campaigns with Influenced Opportunities report that looks promising. Let’s say there are 50 contacts at an account. There’s one opportunity and there are 2 contacts directly tied to the opportunity at that account. SFDC can’t pull the other 48 contacts into your report. To truly understand what’s going on at the account, your solution must be able to pull in and analyze all the contacts, and ideally, “orphaned” leads as well.
  • It must be able to identify or include campaign activity from “orphaned” leads. When marketing passes new records into Salesforce.com for sales to qualify they are created as Leads. In order to be included in any type of Salesforce.com opportunity reports though, Leads must be converted into Contacts and associated to the proper Account. It’s not uncommon for sales teams to convert some, but not all of the leads responding to campaigns at accounts where opportunities develop, so if you’re relying on Salesforce.com reports you’re automatically excluding any records that are still Leads in your system from your analysis.

Vendors in this space that understand this reality have developed fuzzy matching algorithms to address this tracking challenge. Some will merely identify orphaned leads, but not include them in reports, meaning the burden is still on your organization to convert them into contacts. Other vendors are taking the additional step of including them in their reports. If having large numbers of orphaned leads is a persistent ongoing reality in your organization, this may be the most important requirement to focus on when evaluating solutions.

  • It must be able to add custom filters (business rules) to your reports. A great example is putting in time constraints. Let’s say your sales cycle is typically 6 months but you have campaign responses going back 10 years or more. You can’t reasonably say that a response from several years ago has any bearing on an opportunity created today. Your reporting solution needs this type of filtering capability to produce meaningful results.

Conclusion

The best advice is not to evaluate reporting solutions initially. Start with the building block principles covered in the first 2 steps to ensure you have solid data to report on. Doing the hard and often tedious setup work here will help internalize what you really want out of your reporting solution and significantly improve your chances of a successful outcome.