Is your business falling behind in the Marketing AI race?
Think this sounds like military jargon?
You’d be right because lots of marketing terminology comes from classics on war including:
- Sun Tzu’s The Art of War
- Machiavelli’s The Prince and
- Von Clausewitz’s On War.
Or…
I could have referenced Stanley Kubrick’s Cold War satire Dr. Strangelove by calling it:
“How I Stopped Worrying About Marketing AI And Learned To Love The Bot”
Measured in terms of patent numbers, artificial intelligence looks like a global arms race.
According to iPhylics January 2019, the top 3 AI players include:
- Microsoft owns 18,365 patents,
- IBM owns 15,046 patents, and
- Samsung owns 11,243 patents
From a strategic perspective, competitive mindset matters since:
The early AI adopters get a disproportionate advantage according to McKinsey and Gartner Research. And more importantly, this advantage increases over time.
But do you worry that you can’t justify the military sized investment that Marketing AI requires?
I’m not surprised. And you shouldn’t be either!
The average Marketing AI installation runs $50,000 to $250,000 according to @eMarketer. Click To Tweet
As a result, unlike content marketing or social media:
You can’t easily hide even a limited Marketing AI test in your marketing budget.
RECOMMENDED READING
So what do you do?
Follow these 5 steps to make a successful business case for starting to use marketing AI.
The Dirty Little Secret About Marketing AI
Whether you realize it or not, your company already uses Marketing AI.
How do I know?
Because regardless of size, you’re using either one or both of these AI driven functions:
- PPC Advertising including social media (namely, Facebook) and search (namely, Google) takes advantage of Marketing AI software. More importantly, they tap into the large datasets that these organizations have.
- Chatbots.
While 2018 Economist Intelligence Unit research revealed that businesses use a more diverse set of use cases such as:
- Predictive analytics
- Operations management
- Customer service
- Risk management
- Customer insight
- Customer experience (including personalization)
- Research and development
- Supply chain optimization
- Fraud detection
- Human resources
- Knowledge creation
- Pricing and promotion
- Social engagement
At a top level:
- None of these AI activities captures more than 1 in 4 businesses.
- Nor do all of these AI activities yield directly measurable results that translate to increased sales or decreased costs.
According to a 2018 PwC survey, businesses are either assessing and/or using Marketing AI
- 20% Plan to deploy AI enterprise-wise
- 27% Have already implemented AI in multiple areas
- 15% Plan to deploy in multiple areas
- 16% Implemented pilot projects in discrete areas
- 22% Investigating use of AI
From a technology perspective, US businesses use a variety of methods to implement artificial intelligence in their organizations according to Deloitte 2018 Research.
- 59% use enterprise software that included AI capabilities.
- 53% co-developed with an AI solution with partners.
- 49% used cloud-based AI and
- 49% used open-source development tools.
Therefore to remain competitive with your peers and larger competitors:
At least start to test how to use Marketing AI within your business
But ensure that:
When you use marketing AI, it makes sense within the context of your business and specific application goals.
5 Steps To Make A Successful Business Case For Marketing AI
To get on track for marketing AI success, let’s see what your North American peers are doing according to eMarketer:
- Learn from early adopters’ successes and failures. As with many technologies, while you may not want to be on the bleeding edge where the risk is high, you need to stay on the leading edge.
- Seek advice from third party experts and consultants. Supplement internal knowledge with outside help.
- Train current staff to bring their skills up-to-date, especially since new skills may not exist in the general labor market. Beyond building AI bench strength, give these employees incentives to build company knowledge and remain with your organization.
- Create a cross-functional team focused on AI opportunities and implementations.
- Induce new business models that take advantage of AI.
Marketing AI Business Case Step 1:
Choose A Specific Business Problem That AI Can Successfully Solve
To test Marketing AI for your company, choose a particular technology to fix a specific problem.
The key to Marketing AI Business Case Success:
- Define your specific do problem
- Find an existing AI solution that solves that problem and
- Can integrate with your existing technology.
At a minimum you’ll need to work with your technology and systems colleagues to fulfill all 3 of these requirements. Although in many instances, you may need to include more departments such as sales and/or customer service.
Beyond finding a specific use case, you need a related Marketing AI strategy and roadmap to ensure success.
The top 5 barriers to AI adoption according to McKinsey 2018 Research:
- Clear AI strategy missing (43%)
- Lack talent with AI skills (43%)
- Functional silos constrain end-to-end AI solutions (30%)
- Lack of leadership ownership and commitment (27%)
- Technological infrastructure doesn’t support AI (25%)
Marketing AI Business Case Step 2:
Overcome Lack of Senior Management Buy-in
Unlike other marketing initiatives like content marketing or social media, you can’t create a pilot program on the side to show proof of concept and financial value.
Why?
Because artificial intelligence requires significant financial and employee investment.
As a result marketing AI requires senior management buy-in.
Therefore:
You need to talk c-suite when you make the case for artificial intelligence investment.
To justify the 5 to 6 figure financial investment required and integration across multiple areas of your organization, vendor recommendations and selection don’t provide sufficient proof.
Instead, you need to do your due diligence
Check existing marketing AI test cases to:
- Determine whether the specific business challenges are similar to those you face.
- Ask how the AI technology worked with the company’s existing systems. Assess how these systems relate to the ones you have. If they differ significantly, the implementation may not work in your firm.
- Inquire about the AI implementation worked with the test business’s existing processes. What types of operational changes did they need to make? What types of unexpected problems occurred? How difficult where these issues to resolve?
- Speak to each vendor’s references directly. Don’t assume that a test company’s reputation is sufficient proof! Talk to people directly involved in implementing the AI technology.
- What went smoothly?
- Where did they encounter problems?
- How easy was it to get vendor support?
- Did they have to pay extra for vendor help?
- Did they need to change other technology and/or processes?
- What type of results did they get?
Marketing AI Business Case Step 3:
Requires Access To Useable, Quality Data
Artificial intelligence is only as good as the initial dataset that you use to train the machine. Also your program can’t handle a problem or event that doesn’t exist within the dataset.
Adding further complications, the average business had 15 sources of information in 2018 according to Salesforce.
“One of the big challenges in marketing is that the data you want is never in one place, it’s in 12 different places,” CEO of Chatkit, Mazdak Rezvani, said in an eMarketer interview. “You have to dig around to bring it together.”
For marketing AI business case success you need to assess you:
- External data
- Internal data
- Audience and/or customer privacy concerns and protections. Some organizations rely on the quality and handling of the “Big 3” such as Facebook, Google and Amazon.
Current AI data related challenges include (IDG 2018 Research via eMarketer):
- Data exists in different silos across the business (51%)
- Use of too many technologies creates complexity (37%)
- Difficult to access large sets of “clean” data quickly (35%)
Marketing AI Business Case Step 4:
Ensure that your Existing Headcount Can Support The Marketing AI Test Case
Contrary to popular belief, artificial intelligence creates jobs.
The problem:
Current employees don’t possess the necessary skills to handle this specialized work!
In fact, as noted in the IDG chart above, 3 of the top AI hurdles relate to talent including:
- People processing and/or preparing the data and people creating the models have trouble collaborating (35%)
- Lack of data engineering and scientists available to do the work (29%) Often, data cleansing and quality control are key functions and often require specialized skills. More importantly from a business perspective, they create human jobs! This is where the rise of the data scientist is necessary. For many businesses, these skills don’t currently exist with an organization.
- Collaboration between data scientists with varying levels of skills and technology knowledge can be difficult (25%).
More broadly, artificial intelligence will have the biggest impact on jobs involving repetitive activities and/or low digital skills. McKinsey predicts these positions will decrease by about one-fourth to 30%.
By contrast:
Jobs requiring non-repetitive skills and higher levels of digital skills will increase about one-fifth to 50%.
Further:
87% of marketers believe AI and machine learning are already important to business success. But the May 2019 DMA data reveals a significant gap between current skill levels and the skills businesses need to leverage AI.
Marketing AI Business Case Step 5:
Prove Marketing AI ROI
Given the size of the business investment in Marketing AI, you must be able to prove bottom line results to justify the costs.
As a result, the function for which you choose to test using Marketing AI must be of sufficient size to justify the cost.
Translation:
Select a marketing issue that involves prospects or conversions to yield positive returns.
To ensure Marketing AI business case success:
Be able to prove easy-to-measure results that your c-suite understands.
Among your options are:
- Increased sales (or qualified leads)
- Improved Net Promoter Score
- Shorter sales cycles
- Reduced costs
Need proof?
Of course you do! (I would!)
Drift Marketing AI Business Case Study
Drift’s use of their chatbots to support and increase sales cost effectively provides a great example of how you should make the financial case to your c-suite for using marketing AI.
Drift’s 4 key sales-driven factors include:
- Conversations Yielded New Lead Source. An additional 15% of leads were added o the top of Drift’s funnel.
- Conversations Became A Major Pipeline Source. Conversations drive half of Drift’s business.
- Conversation Leads Closed Faster. On average, leads closed within 3 days from initial conversation to Drift demonstration.
- Automation and AI dramatically improved signal-to-noise ratio. Bots handled almost half of all conversations on their own and supported most of the remainder.
Take note that Drifts measures included both increased qualified leads, shorter time to close and better use of sales people’s time.
RECOMMENDED READING:
Marketing AI Business Case Success Conclusion
To make the Marketing AI Business Case to your firm’s c-suite, you need to do your homework because it:
- Requires a significant financial investment
- Involves non-marketing experts such as data scientists and technologists
- Uses data that may be siloed in various departments
BUT don’t let these factors hold you back:
Because chances are that your competitors are already assessing, testing and/or using AI.
And you don’t want to get left behind since over time the AI leaders will wind up with a significantly larger advantage in terms of profits and market share.
So take a deep breath.
You can handle this.
Just follow these 5 steps and do your homework!
Happy Marketing,
Heidi Cohen
Editor’s Note: Hat tip to eMarketer for providing me with their Artificial Intelligence Reports.
Get Heidi Cohen’s Actionable Marketing Guide by email:
Want to check it out before you subscribe? Visit the AMG Newsletter Archive.