Pitching Artificial Intelligence to Business People
From silver bullet syndrome to silver linings
In this article I plan to share with you our recent experience pitching AI to business folk, and what lessons we learned along the way.
As a small firm of AI experts, we follow an awareness marketing approach. Rather than relying solely on one marketing channel, we attend conferences like the Toronto Machine Learning Summit. Although that’s a great way to meet rising talent and generate leads and press, the online game is important too.
We have been providing artificial intelligence consulting to enterprise clients for almost 3 years now, and a few months ago, we decided that a few short marketing videos would be a good way to share what we do with the world.
As it turns out, we had a lot to learn about how business people misunderstand the current capabilities of AI technology when watching video. Our content marketing approach works well on blogs and LinkedIn, but the video marketing stuff fell flat. Let’s talk about what went wrong, and what we did to fix it.
We set out to distill our capabilities into teaser clips (see below) to show decision makers what we do, and why it matters to them. Step 1 was establishing the concept: Showcasing various applications of AI, in order to hint at our expertise.
Here are the 6 key messages we wanted to send:
- Adapt or die: Those who don’t adopt AI will be blown away by competitors and newcomers who adopt AI
- Documents analyzed by AI: Automation of document processing solutions currently performed by humans. More generally, business process automation
- Our AI development capabilities: Just in general, positioning us and saying we do AI stuff for enterprise clients
- Unlock the value in your data: Monetize your data by creating AI systems that improve KPIs like revenue, quality, and costs
- Edge computing: Deploy your ML models into mobile apps, so that you don’t need massive GPU capacity to do inference
- What A.I. really looks like: Showing embedding vectors in tensorboard, because it’s cool
The vision seemed clear to us, and we rolled straight into production.
The Silver Bullet Syndrome — A Harsh Reality Check
Feeling quite proud of our new video content, we planned and released our videos into the wilds of LinkedIn, but much to our surprise, we got lots of views and very few leads. Bottom line, those leads didn’t move the dial on our sales of AI products and services.
Before jumping to the solution, let’s have a look at the content we created, to give you a sense of the problem with the teaser content. The following playlist contains the 10 teaser videos. Take a moment to watch these short clips, and think about the reaction of a CxO with no background in AI, or how it works. Here is a playlist of the teaser videos:
Once we understood that the campaign was going poorly, we took a step back from advertising to assess what went wrong.
What happened was that our “decision maker” customers in C-suite positions were making assumptions about the content that we didn’t expect. When we showed an example of AI generated art, our audience assumed that we have a product that does only AI art generation. Even more narrowly, they assumed we only generate Japanese anime art. We were trying to show expertise, but instead were being interpreted as a narrow band of product features and capabilities. The fact that our target customer assumed we were selling the application in the videos, rather than the capability required to make the application, was our bad.
PROBLEM: We were trying to show expertise, but instead were being interpreted as a narrow band of product features and capabilities.
The analogy I like to use after having this experience is that prospects were thinking that a video of knives cutting meat is telling them that we sell knives that ONLY cut meat. Whereas, we sell knives that cut veggies, fingers, expenses, and time. We concluded that in order to explain advanced AI capabilities, longer and deeper content is required. Watching the videos above, you can, in retrospect, see the difficulty of understanding the message in a short video on a complicated topic.
The Silver Lining—We Found and Fixed the Problem
So, what are some examples of content that does work?
Basically, real media opportunities with long format content and high production value. The following are 4 examples of content that clicked for us. Pun intended. I chose 4 different AI projects to make the point that pitching AI is more about the value messaging than it is the AI itself.
Long Form Content: Webinar on AI at VanillaSoft for Lemay.ai
This webinar at one of our clients was a friendly venue to discuss where I see the industry going, with a live audience in attendance. I pointed out last year that small companies are the worse off than larger ones because of the shortage of data scientists, and now that we have grown and the industry matured, I felt it was about time to re-discuss the evolution of the AI consulting business. We got good signup numbers (~150) that turned into leads, and then we also got follow-on contacts from people seeing the content online.
Explainer Video: Document understanding intro for Stallion.ai
Once we understood what went wrong with the short videos, we got a lot more detail oriented in the video production. This explainer video on document processing goes through the problem, solution, value proposition, and the generality of the solution. Rather than showing a fly-by demo and hoping that the viewer gets the message, the new video shows clearly the capability of the technology, and the reasons that value will be realized.
High level introduction: Internal audit AI for AuditMap.ai
In this video, we decided not to show the user interface because it keeps changing as we add new features for clients. We actually developed this AuditMap intro video before the 10 teaser clips above. We have been getting one main question after showing this video, which is “can I see the demo?” I have no problem with that issue, because we are keeping the good stuff for the ~1 hour demo. AuditMap is so complicated that we don’t want to show too much before actually talking with the customer (typically in banking) to address their questions in a live call.
Motivating the use case: Introduction to AI asset management for investifai.com
It took a lot of internal deliberation and client interaction to figure out what messaging works for investors. They are quite a different segment than our usual decision maker clients. We had to hammer home the key points very simply: AI is not emotional. Portfolio management is too complicated for humans to do without computers. Risk-adjusted returns are essential in this end-of-cycle point in time, where markets are getting scary, and plowing money into one asset class is not a smart bet.
So far the feedback on this video has been positive.
Conclusion
In this article I set out to give you a quick recap of our bumpy ride on the path to marketing our AI development capabilities and solutions. What we found is that profoundly longer content is required in order to explain these AI capabilities. Regardless of the underlying tech, the content needs to sell the viewer on why the solution is better, rather than why the solution is cooler.
If you liked this article then have a look at some of my most read past articles, like “How to Price an AI Project” and “How to Hire an AI Consultant.” I also have prepared articles on other issues faced by companies looking to adopt machine learning, like “Machine learning without cloud or APIs.”
Until next time!
-Daniel
Other articles you may enjoy:
