“The Bridge of Context” Between Data Science and Marketing/Sales

December 20, 2017 Noam Horenczyk

I’m not a data scientist, and I’m not a marketing or sales person.

I don’t know how to create neural nets, run an email campaign or close an opportunity.

What I do know is how to bridge the contextual gap between data science and Marketing/Sales. In bygone years this may not have mattered — after all, how often would Sales or Marketing people come into contact with a data scientist?

But today, with the explosion of marketing analytics platforms and Artificial Intelligence (AI) solutions for Sales and Marketing, these two very different groups are working together far more often than ever before. And that trend is only going to strengthen in the coming years, so it’s critical for modern businesses to develop a common language.

But first, two quick disclaimers:

1) I know that “Bridge of Context” is a phrase I made up, but hear me out and I’m sure you will understand what I mean by the end of this post.

2) I’ll be clustering B2B Marketing and Sales as a single group throughout the post. I know that’s not the way it is in reality (although in my defense, the lines are becoming increasingly blurred), but it makes it easier to prove my point.

Data scientists are from Mars, Marketing and Sales are from Venus

Everyone knows about the perennial gap between B2B Sales and Marketing. But if anything, the gap between data scientists and marketing/sales people is even bigger. These groups usually have different goals, motivations and styles — and they usually speak completely different languages. At Leadspace, I’m the one who stands between these groups: the amazing Leadspace data science team and our incredible B2B marketing and sales customers.

My role includes listening to the problems these marketing and sales people have, helping strategize and develop analytical products to solve these problems, package these products into meaningful and actionable solutions and finally to implement these solutions within our customers’ flows.

In this post I will present two of the challenges I face on a daily basis and the ways I try to overcome them. Hopefully this post will be insightful for:

1) Marketing and Sales people who works with data scientists and/or their outputs

2) Data scientists who serve Marketing/Sales people as the business stakeholders and/or end users

3) Anyone in a similar role to mine

Making data science accessible to B2B Marketing and Sales

Predictive models tend to be complex, but their outcomes do not have to be. In fact, they shouldn’t be. While a data scientist can talk for hours about confusion matrices, ROC curves, dimensionality reduction, etc., most Marketing/Sales people will lose attention and drift away pretty quickly. This is not necessarily because they are “less technical,” but simply because it’s not what they care about.

What Marketing and Sales people care about is the business implications — whether that’s on the macro level, such as “What does this model tell me about which companies/people I should target (in other words – what makes an ‘A’ account/lead)?”; or on a micro-level (“what is the propensity of a specific account/lead to buy and why?”).

The way we at Leadspace bridge this gap is by carefully thinking about the data and analytics we present to our customers. Instead of bombarding the customers with endless analytics, we emphasize the bottom line business implications and provide a flexible Tableau interface, supported by exclusively relevant data, for our customers to answer whatever question they have by themselves in a self-serve way. We’ve seen that this method provides just enough information for our customers to see the value of our modeling, without drowning in the endless sea of data science complexity.

This is where many predictive vendors fall short — although it’s not usually for providing too much information, but rather the opposite. Many vendors still use “black box” predictive models which give minimal insight into how they reached their scores.

Here’s how Leadspace strikes that balance between transparency and focus:

As our CEO Doug Bewsher has noted, this focused approach is critical to using any form of marketing analytics.

Making marketing/sales use cases accessible to data scientists

If you work with data scientists, you’ll know that almost every time you ask them a question, the answer will be “it depends.” (Or “we need more data!”- but that’s a whole different topic for another time…) As frustrating as that is, it’s actually true! Data science is so versatile that without understanding the business context you can be working on something completely irrelevant for weeks.

Still, the result is that it’s often extremely challenging to expose the data science team to customers on an ongoing basis. How many times would your customers hear “it depends” before they start pulling their hair out and running for the door? That’s why it’s crucial that the people who bridge the gap do it properly.

The biggest mistake you can make is to tell your data science team “this is what I want and how I want it,” without providing enough context. The result will invariably be a sub-par model and a frustrated data science team.

The way we at Leadspace bridge this gap is by having an internal call regarding each project, where I provide the data science team plenty of context, such as the specific use case; which metrics are most important to optimize; other scoring solutions the customer is using, and so on. Every piece of context helps them modify their modeling techniques, resulting in a better model.

From my experience, data scientists truly want to know the details of the problem they are trying to solve and have as much freedom to solve it on their own, rather than getting strict guidelines “from above” with no flexibility to use their own judgement, skills and creativity. And that’s one thing Sales and Marketing can agree on too — no one likes to be micromanaged!

Building this “bridge of context” takes time and patience, and requires focus from both sides to understand and communicate precisely what they want/need. But there’s no shortcut — black box solutions which skip this step simply don’t deliver anywhere near the same quality, as every business’s needs, objectives and challenges are very different.

On the other hand, Sales and Marketing teams who make the effort to bridge this gap will be best positioned to successfully utilize some of the most exciting emerging technological solutions.

Learn how you can build a winning marketing stack — download the free guide:

The Modern Marketer's Guide to Building an Effective MarTech Stack

Image credit: Pixabay | CC0 Public Domain

The post “The Bridge of Context” Between Data Science and Marketing/Sales appeared first on Leadspace.

Previous Article
3 Crucial Steps to Building a Winning Marketing Stack in 2018
3 Crucial Steps to Building a Winning Marketing Stack in 2018

This article was originally published on MarketingProfs, and is republished here with permission. When Lead...

Next Article
Leadspace Announces $21 Million Series C Funding Round to Fuel Further AI Innovation
Leadspace Announces $21 Million Series C Funding Round to Fuel Further AI Innovation

Today is an exciting day for Leadspace. Today we announce the completion of our Series C funding round, led...