Transforming business with AI
About ten years ago, when I had one of the world's largest retailers as a client, I visited their headquarters building for a meeting with some of their senior executives. Entering the building, in order to enter the main office area, one had to walk down a corridor where on either side you could see room after room of vendors pitching their products to the company's buyers.
Sitting at the bar at the hotel that night I met one such vendor who described his experience. He had driven with his sample products in his car into town in order to have his one hour face to face meeting with the retailer's buyer. For him the meeting was a make-or-break opportunity for his business. The volume, if he earned the business, might crush his small company but without the distribution this retailer could provide he would also fail.
I don't know how things turned out for him but I did come away from the experience convinced that, while a necessary part of the retailer's operations, this was a terrible process both for the buyers and the sellers. Couldn't anything be done to improve how this retailer screened potential products for their stores? It has been 10 years since then so perhaps things have changed a lot. Or maybe not.
But what I do believe is that this is the kind of business process that can be radically transformed using the current, or at least a coming soon, generation of artificial intelligence. The logic that buyers are applying to decide on what to purchase can be codified. Vendors could be pitching to an AI buyer that can evaluate the product and the vendor. All of this could be done with less anxiety and effort by all parties.
Let's take the process apart step by step and think about what transformation might look like.
Phase 1: Introducing a mock interview for sellers
As a first step the retailer might create an online AI agent which serves as a training/testing experience for the sellers. Just as in the physical presence of a buyer, the seller would pitch their product to the AI agent. But in this first phase the AI agent would simply provide (in a helpful and supportive way) an evaluation of the presentation with tips to the seller about how to improve but also guidance on what the buyer is likely to like or dislike about the product, value proposition, or terms. For example, perhaps the seller neglects to say anything about their ability to deliver certain quantities of the product in given time frames. The bot might provide a prompt saying "you'll need to inform the buyer that you can produce and deliver at least X quantity every Y period for this type of product."
领英推荐
The seller can then refine and improve the pitch, the product, and the offer before ever getting in front of a human buyer - making both buyer and seller's time more effective when they do meet.
Phase 2: Screening for meeting requests
As a second step the retailer could utilize a refined and improved version of the first phase AI agent to recommend approving or rejecting vendor requests for meetings. As before the vendor would pitch to the AI agent and the AI agent would provide feedback. But in addition the AI agent would prepare a summary of the interaction and a recommendation to the buyer as to whether or not the buyer should schedule a meeting. The decision would still remain in the hands of a human - responsibility for decisions should always be in human hands with the duty of care being higher as the consequences of the decision increase. But substantial amounts of everyone's time can be saved if vendors who are unlikely to move forward in the process can be screened out in this way.
Phase 3: Making buying recommendations
Eventually improvements in the AI agent can lead to full screening and recommendations for product purchases. An AI agent can patiently take as much time as is needed, and as many follow-up meetings as is needed, with a seller to work out details on pricing, production, and specific product configurations. A complete dossier on the opportunity would then be automatically prepared and delivered to the human buyer - again putting the ultimate decision into human hands. But by this time the relationship of buyer and seller, and their respective roles and activities have vastly changed. Buyers are now able to spend quality time with just those vendors that have reached the point of providing maximum value to the retailer. Vendors now have the chance to pitch multiple times, refining their offer until it makes sense for both parties to move forward or ceasing engagement if there isn't a business value to be found.
This isn't that far away. Take a look at some of the newest capabilities for analyzing video streams in Anthropic's Claude 3.5 or OpenAI's 4o and you can see where this is going. Massive reinvention of every business process is beginning now and will continue for the next decade. What are you reinventing?
Automation, AI, Analytics, Global Delivery * Manufacturing, Retail & CPG
3 个月I am with you Ted. Sometimes a new technology makes a totally new way of working possible all at once. More often - like in your piece - the new technology muscles its way into an existing business process step by step. Decomposing how AI fits into business process is the fundamental challenge for enterprise AI.
Executive Fellow @ Harvard Business School | D.B.A., GAI Insights Co-Founder
3 个月Ted Shelton I couldn't agree more. This very afternoon my colleague Paul Baier and I are doing a webinar GenAI for Busy Sales Professionals and we take them through exactly what you suggest above and some more things too. Last night on our learning lab a member of our community said his experience using AI in sales was transformative. In sales there are only three variables that matter: how much potential does the client have to buy your product or service (Q); how good your pitch is and how many times you give it (P) and how many times you give it to high quality prospects (N). Sales = Q*P*N Nothing, and I mean nothing else matters as much. AI can help in Q, P & N, and can help with the sales admin after you win the order!
AI Strategist | COO | Creating Business Value with AI | Top 1% Results
3 个月Ted, efficiency in vendor screening, better decisions with evaluations, and feedback from AI agents are going to happening in the next 2-3 years as AI becomes more reliable. It will upend existing procurement processes. I also think human oversight is needed. 80-20!
Florist | Volunteer Community Manager at Say, Pi | Google Local Guides Guiding Star on Google Maps | Kin Beta Tester
3 个月This is very amazing, Ted Shelton. ?? I am sure that this will definitely lead to an improvement. AI is really supportive in various areas. I'm looking forward to everything that comes in the field of AI. ??
Founder and President at Cunningham Collective, Best Selling Author
3 个月Nice piece! I think this type of process not only improves the buying exercise for both buyers and sellers, but in the end has the potential for making the products better suited to the market creating a continuous improvement loop for product market fit. Think of the implications for every kind of product or service. They are mind blowing.