As I sat on my sun lounger on my holiday and got yet another insurance sales call, it struck me that there are so many factors that are still ignored in our AI models. For example, at that point, I wasn’t going to buy anything unless it was cold and refreshing. The number of variables required to predict my frame of mind at that point in time, match the product to my thinking (a cold beer!) and connect that with a willing seller are mind boggling, but would have guaranteed a sale. But it IS possible with the right AI system. Let me explain.
The key to selling is to establish a relevant dialogue and understanding with your clients. It’s first prize for all marketing teams, especially when there is such a limited time available to understand your customer during the sale. That’s why companies that use any type of personalisation in their marketing funnel deliver much better results (and also why humans are 4 times better at selling to other humans than computers are – that’s a story for another day).
As a company that creates personalisation software, understanding the customer mindset at the point of sale is a huge challenge and an exciting opportunity. What’s great about being good at it is that it helps both the purchaser and the seller to reduce sales friction (for example, none of our clients would have made that sales call to me on my pool lounger!).
We know that sellers want to predict and understand behavior in advance of the sales call, and their product marketers want to target the right customer with the right product. As I’ve always said, AI is a human-intensive process to set up. To make all of this work requires human intelligence, working with good data and systems to create predictive efficiency in the sales process. And all of this in real time.
There are so many places where this insight needs to be applied:
- In finding the right ad space to bid on to get the right customer at the right time with the right product
- Knowing how much the customer is worth and how much to bid for the ad to make it worthwhile
- Using AI to manage the sales through the funnel, from the marketing messages at the start to the follow up when interest is shown
For a new-age, tech-savvy company, hundreds of small calculations are running constantly to make the sale more relevant, targeted and interesting to the customer. Sometimes this effort drives what may seem a small uplift for the company, a 10% to 20% increase in sales. But that’s enough for them to capture the market and an industry from their less nimble competitors.
As AI becomes more mainstream, the options open to marketers will develop and change. Measurement and accountability will be more robust and there will be a demand for greater levels of return. Marketing efficiency will include real time insight and the test and measure solutions of the past will require far shorter timeframes. We’ve noticed that the systems we are using are changing to allow for this – streaming data is now required, with constant AI retraining happening in the background.
This machine learning has energized marketing and the channels used to generate leads and clients for both small medium and large clients. The challenges facing all organizations is how to find the skills, either in the business or externally, to implement and manage their solutions.
With all these advances in AI technology and reductions in system cost, there is one area where a bottleneck does exist – the human element. AI is a science, and the key to this science are the humans who code and run their models. It has created a new human field of study, data science, and new opportunities for jobs for future generations.
So whether you own a bakery, manufacture key rings, run a bank or sell insurance, AI is empowering business and removing sales friction. It is changing how we interact and engage with our customers, allowing us to provide the perfect service and product at the optimal moment, and communicate this with the most relevant message.
So as I sat on my pool lounger, I decided the next AI model I build will predict when I needed my next cold beer. It needs to look at the time of year and my geo-location, it must measure how long I’ve been out of the pool, what the current weather is and where the nearest waiter is. Success is whether at that particular point in time I’m looking for a beer. Let me know if you want to be part of the Holiday AI Lounger experiment – or HAL, for short.
Founder – Optaltix