Vennli, Make Your Content More Effective: http://vennli.com/
Is Your Marketing AI Project Doomed to Fail?
Blog Feature

By: Sandie Young

Print this Page

June 6th, 2019

Is Your Marketing AI Project Doomed to Fail?

Editor’s Note: This post is part of a series featuring speakers from the Marketing Artificial Intelligence Conference (MAICON). For more information, visit www.MAICON.ai.

Cal Al-DhubaibMany marketers are eager to get started with artificial intelligence (AI) projects—often only to struggle to see the return. 

If this is you, don’t give up. Marketing AI starts with experimentation, and experimentation thrives on trial and error. The key is learning from your failures.

In this post, the Chief Data Scientist at Pandata, Cal Al-Dhubaib (@caldhubaib), shares lessons learned from deploying hundreds of AI projects. Read the full Q&A to better structure your AI projects. And join us at MAICON in July to see Cal’s full session, Setting the Data Foundation to Succeed with AI.

Q: What are the most common reasons you’ve seen AI projects fail?

A: Artificial intelligence, and data science in general, are experimental in nature. The number one biggest reason we see AI projects fail is that they're treated as black boxes and organizations don't take the time to experiment with what works and what doesn't.

The next most common reason is lack of data literacy. Data literacy is a measure for how fluent an organization is with interpreting data and making data-driven decisions. A data-literate organization understands the life-cycle of data from collection to decision making, critical assumptions behind terms and KPI's, and the impact of data quality on results. It's really important to include education as part of any AI project to increase data literacy and ensure success.

Q: What can marketers learn from these pitfalls? How can they better structure their projects for success out of the gate?

A: There are two critical processes that need to run well for an organization to be successful with AI: experimentation and operationalization.

The processes are separate, but related. Experimentation translates business pain-points into data-driven solutions, and operationalization takes potential solutions and validates the business value and gets them into the hands of decision makers.

null

 

While these eight steps are all important, discovery sets the foundation for success. In our experience, successful projects start with the RISE process using these four questions:

Results: Who will use the final solution and what decisions will it affect? For example: The solution will support Sales Associates to prioritize which leads to call.

  1. Impact: Can you quantify the ROI and is it worth the time and effort? For example: This solution could decrease cost of customer acquisition by up to 10%.
  2. Success: How will you verify the ROI and know that you achieved success? For example: Cost of customer acquisition will decrease by 10% within one year of using the solution.
  3. Ethics: What regulations might affect your solution. For example: GDPR may limit what customer specific information we can use on prospects that are European citizens.

Q: What commonalities do successful AI projects share? What do they do right?

A: We see three common themes among successful

  • Buy-in. The organization is aligned on ROI and AI projects are prioritized according to organizational goals.
  • Trust. Solutions are delivered with appropriate amounts of education and stakeholders understand the underlying assumptions.
  • Consistent Performance. Performance and value creation is consistently tested and different users of the solution experience similar results.

Q: Who do marketers need on their AI implementation teams? What roles are critical?

A: While different organizations have slightly different titles, you will see some combination of the following roles:

  • Data Scientist – A data guru that can design and build artificial intelligence to solve business challenges.
  • Data Analyst – A creative problem-solver with quantitative, business, and data visualization skills.
  • Data Engineer – A software engineer who brings data solutions together, at scale.
  • The Stakeholder – A visionary who wants to thrive in their industry by using realistic solutions to data-driven challenges.

 

If the stakeholder, or the marketer in this case, is not a part of the implementation team, projects are doomed to fail.

Q: If a pilot project does fail, how can businesses gain buy-in for future AI trials?

A: Experimental disciplines, like AI, go hand-in-hand with being a learning organization. It's important at the end of any AI project that you document either a compelling return, or a compelling reason why the expected return wasn't achieved. The latter is sometimes even more valuable than a 'successful' project that reveals areas for improvement within the organization. As long as an organization continues to grow from AI projects, 'failures' are not permanent.

Q: What are you most excited about at MAICON 2019?

A: I'm excited to hear stories from marketers who are using AI in practice. With AI still a relatively young discipline, it's an exciting time to see how organizations are adapting their teams and processes to meet the challenge.

Q: What advice do you have for marketers just starting with AI?

A: Start small and get your organization excited. Nothing kills trust faster than taking on too much and having to explain poor ROI on a large investment.

Get free access to the Ultimate Beginner's Guide to AI in Marketing: https://www.marketingaiinstitute.com/beginners-guide-access

About Sandie Young

Sandie Young started at the agency during the summer of 2012, with experience in magazine journalism and a passion for content marketing. Sandie is a graduate of Ohio University, with a Bachelor of Science from the E.W. Scripps School of Journalism. Full bio.

  • Connect with Sandie Young