AI Series – Part 6: The Trifecta
There is a reason that Mr Benioff is introducing the Salesforce Trifecta as CRM + Data Cloud + AI. It’s these three elements that enable organizations to align their models which is what we are going to dive into now. Data flows into sustainable CRM foundations that align divisions and this allows actionable insight via Analytics and future insight via Forecasting and Predictive Models which provide the data for AI to flourish.
First, you don’t need to use Data Cloud, but it’s important to understand the concept of Data Cloud and its place in the order of the trifecta so you can build your system design appropriately. Let’s take an excerpt from SF Help:
Data Cloud allows companies to “Organize and unify data across Salesforce and other external data sources. After data has been ingested into Data Cloud, it can be used to drive personalization and engagement through the creation of audience segments. Additionally, through identity resolution you can achieve a single, actionable view of your customer.”
I’ve already mentioned the 3VofData (volume, variety, velocity) in my previous article on Modeling A, so now lets dive into Modeling part B.
Note to reader: I’m not going to be reviewing ‘how to guides’, or steps of configuring AI in your org as I’m approaching this as a tech agnostic process using Salesforce’s model for an example since there is so much good information to learn from in Trailheads, Youtube and the SF Community that applies to AI as a whole.
Let’s begin:
Brainstorm 10 Use Cases. 3-5 for each division. Each Use Case has to have the criteria of being measurable in milestones and contain current and future state metrics so that you can measure success rates and not just an ending metric which misses the gap analysis of where we are and where we need to go. Sales and Service are obvious starters and although Sales has more hard metrics to leverage, Service has rich grey areas of measure that impact multiple products, services and customer Trust – Ex. attrition rates, user input. These may be just as valuable as defined metrics and can open doors to hidden or new revenue. On a project some years ago, we enabled customers to find their own documents on their products and find answers to their own questions via an extensive knowledge base which went viral. Soon customers were discussing products amongst each other driving a viral social network of super product users and fans. I can tell you our client did not see this coming and their metrics went onward to places they did not foresee either. Think about these and other scenarios that you can measure for each use case.
Use Requirements gathering structure. I have an article just for this section here. With your 10 defined use cases, begin by providing the structure as you would a requirement. Who (who is it for, which users, system can also be considered as a who. Often started with “Ability for…”), What (what is it your doing), and Why (value statement for why the requirement needs to be executed. A metric is not a Why but it helps as a support guide for success. Often started with “so that…”). Some say that 85% of a project’s failure can be traced back to improper gathering of requirements & use case. I believe the metric is much higher. Here is a quick Ex. Ability for Sales Reps (in NA division), to view most searched products on company site (or aggregated off X sites), so that…SR can execute marketing campaigns aligned to the greatest customer interest to meet X metric for divisions sales quota. (*be as specific as possible in your requirements for vetting to your top 3-5). The so that… provides the why or value of the requirement.
Timeline: Now that you have fully defined the Use Cases as Requirements, choose 3-5 to begin executing on with easiest execution and the shortest timeline to gain momentum and trust. What you really need are momentum and clarity. Momentum is a magic word that propels a project and its team to continue with focus and drive. When you have it you think you are talented, lucky and going up. When you don’t you feel stuck and begin to slip into indecision. Clarity provides the target for your momentum. That is why proper requirements are so important. They provide the Clarity. Timelines that are executable, measurable and timely are your momentum. Get the first win and the rest will follow in its path.
Data Strategy: I have already covered this in many my previous articles so I am not including this here. Look to my blog for more information on this item.
Trust: This is a CoE and overall Company discussion of how to manage client data in an ethical and legal way. This involves being transparent to your customers & employees. I’ve seen numerous statements on sites stating how they are addressing customer data. I suggest your company do the same as it generates trust and brings discussions to the table that are better done in a collaborative setting vs having to call a lawyer. Just saying. “Transparency”, add that on to ‘Momentum and Clarity’ for the big words for your journeys success.
Here are some questions to get your CoE started: What data would impact our customers/clients privacy? What are the legal implications? How do we communicate with our customers on privacy issues? Message, consent forms, udpates, value of AI? Does this impact our employees? (What are the legal implications?)A big one I have brought up before which is really tough to target but needs to be addressed is that the data you are using is not biased. If it is – address that it is biased and figure out what to do about building ‘Trustful Data’.
Technology: Data, check. Use Cases, check. Trust, check. Now is the exciting time where the rubber hits the road and technology (platforms, apps, programming languages, etc) and infrastructure gaps now need to be identified. Run through an entire AI Use Case lifecycle and check on all technology points to make sure you have the right technology in place to close a full loop – that means start to finish and back to start. If you have gaps you also have options. I’ve spoken on platform choices like Dynamic Checklists from Youreka, vs custom vs platform offerings. Each choice can be utilized across multiple Use Cases later and also impact positively across the org so once the value is identified across the business, the cost is validated. Use visuals to show process flows to your teams. Boxes can be color coded via legend for Red, Yellow, Green to identify gaps. Identify in detail what determines each color. * Remember to Validate, Validate, Validate all the way through.
Team Knowledge. I often tell the story of a client the sales team won back in the day when delivery, service and sales did not speak with one another or share data. The project was amazing, career changing, a big pay day for Sales and Management, a real shining star of an opportunity. Everyone was excited until word got to the delivery team. They couldn’t do it. They didn’t have the capacity or the skill set. Everyone tried their best to hire consultants, full time hires, anyone that could get this off the ground. The rest we could figure out but it never happened. Time ran out. It was a loud thud. Why am I telling you this. Because the skill set and capacity you will need is one of those gaps that will impact your project and progress. You can’t pull Dianne or Dane over from their projects and have them ‘support where they can’. Not going to work. Hiring consultants is not sustainable. Get a dedicated team together and get committed to the task at hand or wait and work on the data and AI strategy for the future project. Hiring a primary lead from outside is a great starting point but as mentioned, it may not be sustainable.