We are still searching for AI agents that can significantly enhance our capabilities and knowledge. Basically the enhanced version of what we experiment when we are for the first time in front of a computer (remember when you were 6 to 10 and discovered what can be achieved?). As the technology evolves, we need to compartmentalize the nascent capabilities and reflect on what can be achieved at the company level for now. This primarily involves process optimization and quality management in the back-office. However, achieving this requires a solid foundation in company data management:
- Structuring data
- Standardizing formats
- Normalizing datasets across the organization
- Cleaning data for accuracy
- Enabling large-scale quality control
Without these steps, AI-driven improvements will be difficult to implement effectively.
Once this “small” and enriching task is done, you can work your way to building and leveraging your now efficient and reliable agentic AI ecosystem in order to accelerate sales and optimize your operations, starting with the demand. Obviously, this approach alone won’t suffice. Instead, we need to take the opposite approach—starting with small, targeted capabilities that demonstrate their power before scaling into a company-wide AI transformation. The first baby step would be to focus on supporting an educated decision making for your customers, that would then cascade to your company making more educated decisions on operations, tactics and ultimately strategy. 10 years worth of work, but if done step by step can be contracted within a 5 year plan, working with agile teams and an organization that is resilient enough to accept and implement initiatives that will be quite challenging in the beginning, which means your initial projects need to prove their value within 6 to 12 months as you’ll be under strict review from the company.
Accelerating sales to prove how powerful the agentic AI can be(come)
Empowering consumers may feel quite complicated as companies have always had a large product and service portfolio. Still, once you have retrieved the company sales data and identified the customer personas, that becomes easier. You will cut the chase and focus on building clear main customer profiles to understand how they consume whatever you can provide and highlight how are made the connections within your porfolio. This means you now know what is the purpose of the consumer’s share of wallet you are targeting and are able to define a more credible marketing and go-to-market capabilities and platforms, even being able to clarify your budgeting process and be at ease when you’re asked to work under the zero based budget method. One large problem solved—eliminating inefficiencies and sharpening customer insights to drive more targeted and effective strategies.
As you are walking the walk, you will find yourself really putting yourself in the shoes of your consumers and be able to connect with them. This needs to be embedded into your business rules to ensure the conversational agent aligns with your consumers. Additionally, you’ll be able to equip the NBO/NBA software with the necessary data to offer relevant alternatives and add-ons based on consumer behavior. You can here plug a CDP (the infamous Customer Data Platform) capabilities and “intelligence” so the A/B testing AI agent can directly test new user funnels to better understand at what time it is relevant to trigger a new interaction to recommend a product or invite to read some content that would build more bonding and brand love (that’s the job of the content generation agent in your agentic AI ecosystem).
As you are building your way to from a conversational agent to an agentic AI ecosystem, you can see that you would need to become more and more proficient in data management or Master Data Management, though I guess that wasn’t part of your expectations nor plans. Still, and I’m sorry to be the one to tell you this, that’s part of the journey. It’s now a data driven business ecosystem and everybody needs to get onboard. One way or another, the ones that are still rejecting the data and not at least understanding the outstanding value will be dinosaurs (I’m sure you’re already thinking about Bob and Joe in your office, don’t finger point because they are already working they data blackbelt or working on a strategic move, leaving you with all the work anyway).
Enabling faster sales, not working on building a more complex sales process
What you have just experienced is what will become your IT ecosystem in a few years from now on. That won’t become a reality until the datasets you have on hand are structured, normalized, and actionable. But as you are working on the critical piece, the sales, this will spread within your organization. That means the mindset and the culture will be rejecting what is incoming since instead of building more process and complexity, which is quite standard in large organizations, you’ll see yourself forced to simplify and let go of some parts of your portfolio and processes that are now proven with lower impact than expected, and most of all generating less value than costs (reminder, if value generated below [total cost of ownership + 30%], you should consider a deep dive into the benefits for the company and the organization).
Less is more. When reviewing your product and service portfolio, cutting complexity often leads to better efficiency, reduced costs, and greater agility in responding to market demands. That is what you can expect from what you would have done by now when reviewing the performance of your product & service portfolio and the processes that have been accumulated up till now. That will be a long journey to get rid of the overweight. But that will free space for new investments and capabilities for today and tomorrow since you have in your hands both what your consumers want and need, what they consume at your company, what is missing in your portfolio that they are looking for (just look into the conversations the consumers have with your conversational agent and the customer journeys you created with high drop rate, they are good indicators of opportunities that you have missed because of your partial or missing performance cockpit).
Overall, you’ll see the consumers are going straight to some of your portfolio items and if you provide them with the right guidance capabilities through your now relevant agentic AI ecosystem, you will actually see that you were generating gates and obstacles to sales and missing a lot of information to build the products and services that would actually be useful and purchased by your total addressable market. It will give you free space and capabilities to dive into a Blue Ocean strategy phase to address a newly identified consumer base instead of trying to sell products and services that are hindrances to your current customer base, with numbers that say that you need to push forward on non existing opportunities because of fantasies.
Actual use cases that will redefine your purchasing experience
What you are killing within your company is the consumer overload in (online) shopping. They want to be supported in clearly undestanding whether your company products and services are relevant. That’s the reason to be to comparison websites, ratings and recommendation engines, and you can’t beat them at some point, and with AI by their side, they will become way more relevant than your company in a shorter time span. You’ll be stuck in an unwinnable game if you don’t change how you interact with consumers—spending more time defending outdated practices instead of seizing new opportunities in a more dynamic market. (identified thanks to your agentic AI ecosystem, though it may also cripple your business for a while).
As you now can anticipate, your better understanding of your relationships within your product and service portfolio and your consumers will be feeding conversational agents for several use cases:
- Personalized shopping assistants: fed with NBO/ NBA capabilities, understanding of ever changing consumer behaviors and proximity analysis between customer profiles, the assistants will be able to recognize potential upsell and cross sell, or just work on “replenishment” phases, and thus prevent churn and commercial fatigue with irrelevant promotions that will only kill your margins
- Intelligent bundling: that is what you should be already doing with your revenue growth management strategies, yet since that also involves the business intelligence piece which is not fit for this kind of activities and nobody within the company will want to try and defy the BI tools for business purpose, you’ll be happy to see that once your sales data are cleansed by your agentic AI ecosystem, you’ll find yourself with really useful data for bundling, unbundling and value generating promotions at short and mid term (yes, that can happen)
- Adaptive learning: that’s basically the last step for AI ready organizations, when your agentic AI ecosystem is able to flag required updates in the processes and automatically design them and test them against a relevant segment of real life consumers for A/B testing purpose before providing you with the material for decision making.
Those are the main lines of use cases, you’ll be able to turn around them and identify others that are more adapted to your context and situation. And what you will find along the road are business cases that can be solved using existing sets of tools and capabilities, tagged with highly complex naming (dynamic pricing, MDM, CDP, strategic marketing, portfolio management, zero based budgeting, you name it). But this will be more intuitive as you have the conversational dimension involved, with the capabilities of those robots to structure unstructured data (a very scary name to say you’re organization doesn’t use the ever growing volume of data you are generating).
Key takeaways
For over a decade, companies have been trying to address the challenge of managing the volume of data they generate and attempting to squeeze whatever intelligence they can from it. It requires a massive amount of rigor and data management acumen, which is not massively available as of today thanks to the tech behemots acquiring it all. Agentic AI ecosystem, thanks to the GenAI capabilities, is providing a new gate to your data Everest and a new way of working it and generating value out of it, more fit for organizations as the IT and Data way isn’t for all.
With this new piece of technology, you are given the ability to discuss those topics in plain text and plain English. The competitive edge no longer lies solely in technology. The challenge has shifted from translating between techies and business teams to being entirely about business execution. Which means you’re in for a challenging organizational change. Again. Business as usual?
Next in line in this series will be how to actually deploy an agentic AI ecosystem. As you can undestand, this won’t be based on any return on experience as this is all brand new. This will be based on connecting the dots with what is actually the benefits of agentic AI, the ability to generate a systemic change in the data usage.
Leave a Reply