Experienced marketing technologists know that martech isn't as simple as finding an "easy button" you can press and automatically get more leads and higher conversion.

They know that if you feed your technology garbage, you'll get garbage out the other end. And they know that data quality is the foundation for it all. And attribution is a great example of why that's the case.

But for successful and scalable attribution, high-quality data is just the beginning. Let's look at three issues people often trip over when employing attribution technology.

1. Flexibility to Scale

Attribution is conceptually simple—just a bit of data tracing, counting, adding, and dividing. What makes attribution difficult is not the fancy models, but whether you can easily make it work with your systems, data, and processes.

Here are three technology constraints common to attribution solutions.

Salesforce or Bust

All attribution solutions work with Salesforce data. A few work with Marketo. But beyond that you're looking at writing custom Python code.

New martech solutions and vendors are still emerging rapidly, so you need the ability to ingest engagement data from a broad range of sources, including...

• CRM (beyond Salesforce)
• Marketing automation (beyond Marketo)
• Data warehouses and databases
• Sales engagement tools such as Salesloft and Outreach.io
• Sales insight tools such as Gong and Chorus.io
• Revenue insight tools such as People.ai and Clari
• Intent data such as G2 and Bombora
• Your own product's free trial and freemium activity
• Data from your channel partner

Incorrect Salesforce Setup

Even if most of your engagement data is in Salesforce, most attribution solutions assume your instance is set up generically and you're using objects such as Lead, Contact, Account, and Campaign Memberships in "standard" ways.

But if your Salesforce instance is more than five years old and you're a large company, it's never that simple. To make it work against such rigid technology constraints, you resort to doing things that make your Salesforce even more messy.

One large technology company we worked with wanted to incorporate sales meetings into its attribution model, but its tool worked only with the Campaign Membership object. So, it bulk-created a campaign for every day of the year and wrote code to match all the sales meeting data into those daily campaigns. When the project went live, team members had added over a thousand new daily campaigns to their instance.

Custom Object? What Custom Object?

Custom objects are a fact of life unless your Salesforce instance is less than three months old. They capture critical data that's specific to your company's go-to-market motion. Most attribution tools have limited, if any, ability to deal with custom objects.

One company used a custom object to capture data on its channel sales and partner influence data. Unfortunately, the attribution tool it deployed years ago didn't support custom objects, which required additional manual work and ultimately delivered the combined data into a data warehouse and used Tableau for reporting.

The company recently replaced the manual work with an integrated RevOps automation platform. Its next step: using that platform to replace the entire attribution solution.

2. Flexibility to Model

If there's any martech topic that evokes religion-like fervor, it's the attribution model. People will debate—until the cows come home—which attribution model is better and more accurate .