Supply and Demand side of product building
This edition explores how AI can reshape products by enhancing user introspection and authentic demand.
Theory and Demand
When I was talking about LLM’s with my superboss, he mentioned a thought which feels relevant to sales and product building.
I am paraphrasing here but I hope you get the drift:
People reveal that they don’t introspect about their decisions and actions. So, products that expect them to change their workflows because in retrospect it makes sense have a tough hurdle to cross.
This connects well with his long standing position of products having an overarching theory especially in B2B SaaS.
I think great software products will have the following path. First, they are contrarian in a specific way. Second, their contrarian act (as manifested in the product) is insightful. It offers a perspective on “the job to be done” such that, if you agree with the contrarian view, the world is better or clearer in some meaningful way. Third, the world converges to this view via meaningful contact: it begins contrarian but ends as orthodox, and the transition happens naturally as the product’s perspective infects and converts the users. Good products win the argument with the world of customer expectations in which they began as the contrarian outsider. In product management we call this milestone “product market fit”.
Building a theory and presenting it as opinions to customers allows them to understand the impact of buying the product. At the end, a customer is looking to do a job with the product.
Cedric wrote the 4th post of the series on demand in which he summarises the book The Heart of Innovation(HOI). In the post he defined the process of identifying authentic demand.
Deliberate Innovation’s second idea is a razor that builds on this: a precisely crafted notion of what demand looks like. This idea is actually the more important one, because the ability to identify authentic demand tells you when you don’t have it.
The razor is this:
Authentic demand exists for a solution when someone is put in a situation and they cannot not buy (or use) the solution.
The authors call this the ‘not not’, and may be used in a sentence like so: “the job of the founder is to find a ‘not not’.”
The frame of not not can also help define Product Market Fit(PMF)
But the not not gives us a better way of defining PMF. Put simply, you have PMF when you’ve found authentic demand. As a reminder, authentic demand for a solution is defined as: when someone is put in a situation, they cannot not buy (or use) the solution.
The difference between both of these framings on PMF come from which lens they are viewing it.
The first focuses on the product’s perception by the customer and finding it insightful. The second focuses on customers response when presented with a situation and not not buying the product. You can view these two framings as approaching the definition of product market fit from Supply(operations) and Demand(market) side of business triad.
In both posts, the subsequent sections are used to define and make these concepts actionable. I strongly urge you to check them out.
Let me walk back to initial point of users not being reflective and thus resistant to change. LLMs can intervene and present an intuitive approach to enable users to reflect about their actions.
This all ties back to my own belief that in legacy domains, product development and change management are two sides of the same coin.
In sectors like logistics and agriculture, which are filled with complex workflows. The critical skill of the product person comes from doing step 1(business/domain) and step 2 (recognising the workflow) in product building workflow mentioned earlier. It helps in providing the right optionality before moving to the final step 3 (product development).
A good product changes the workflow and engagement of people in the workflow. The product-lead acts and operates as a change agent, advocating about the progress for the people involved while executing their own workflow.
This all might sound too mushy and theoretical. So, let me present you with a example case.
In Indian logistics, the process of digitisation of workflows doesn’t help the stakeholders in executing their tasks.
Let us suppose you want to generate a digital Bilti (Transport receipt). Doing it doesn’t make either the transporter or the trucker’s job easier. Yet digitising allows them to build a network and build a repository that could enable better transactions. As a product trying to do the later, one will start with generating bilti which the stakeholder views as having no value.
This is where theory building comes into picture. The “better” transactions need a theory. It should accurately represent how the stakeholder can get the job done in better manner.
With GenAI, we have a natural language way which allows us to digitise the bilti without building a table UI. It will also help disseminate the insights from the network through natural language .
In the case of the trucker suggest a different transporter in order to earn higher price when moving in the same route. And in the case of a transporter, presenting the previous execution of the trucker to determine what % from market price should be added or reduced when quoting price to trucker.
For both the use cases, the structure of a simple backend with transactions that stakeholders made in the past. GenAI just provides us a native technology that enables users to reflect and form opinions that coincide with your product’s better approach to transactions.
The demand is already there if you look at the top 10 use cases of GenAI. The top use case is being introspective as a human and understanding one’s behaviour.
The default situation of LLMs is being introspective and reflective. There are multiple modes like voice, image and text to make it feel natural for the stakeholders. Next the demand then presents in defining the not not situation which over here could be “what if you can reduce the number of phone calls you make by 90%”.
(on average a transporter makes 10-15 phone calls to book a single truck. What if it is possible in 2 calls)
The dominant discourse on product building is seeing a phase shift with AI and most of it revolves around taste and prompting to deliver better PRDs but the sense making of theory and demand which is what we talked in this post is still missing.
So, when building in legacy domains(where the front line has no incentive to change workflows ) building theory and framing situations that talk about the better becomes the skill of product building. LLM’s now make it easier since we can use them to make stakeholders reflect their behaviour.
Lastly, the way we derive situations depends on the capital side of business triad. These situation arise from the capital conditions of the domain. Is the industry seeing a capital inflow or outflow. Do you see customers looking to improve their competitive landscape or you see customers consolidating due to lack of runway and capital to extend their journey.
Round up
Commoncog Start Here
Two years of reading commoncog deserves a syllabus for anyone starting now. There is a body of work that would help anyone upskill in business.
Stop Building AI tools backward
Reading the post of Stop building AI tools backward by Hazel Weakly provided me with vocabulary I have been thinking when envisioning AI enhanced products.
Links that resonated
I was reading many posts this week and would highlight a couple of them from this week.
Unsettled
In the process, get clear on the goal. Learn that the traditional goals of supply chain focused on functional cost reduction sub-optimize the value of the firm. Embrace the principles of supply chain physics. Map your constraints by flow. Don’t just balance demand and supply volumes. Instead, understand and align demand and supply cycles to drive outcomes. Form and socialize your own hierarchy of metrics. For example, don’t focus on forecast error. Instead, analyze demand flow characteristics by demand stream to evaluate Forecast Value Added (FVA), forecastability, and bullwhip impact. Design your supply chain with a focus on the form and function of inventory. Here is the metrics framework that I am using at present in my outside-in classes.
On work processes and outcomes
The second model requires us to think more about the more common case when incidents don’t happen. After all, we measure our availability in 9s, which means the overwhelming majority of the time, bad outcomes aren’t happening. Hence, Hollnagel encourages us to spend more time examining the common case of things going right.
Sign off
Majority of this week I was busy going through the daily rigour and doing documentation work for filing a Visa. I tried to use GenAI to prepare me a checklist for filling Visa application and it sort of helped. This is the reason why the issue is delayed.
This week Adrianna finally received confirmation on her Green Card application. It took her seven years to get through to this stage and I am very happy to see her succeed. She posts her journey on Mastodon and has been an inspiration to me.
If I get the Visa then I will be traveling to my first tech conference. Spending time with engineers and presenting the work we do internally to the broader community. I am looking forward to this experience.
Singing off till next time,
Vivek, printing and shipping reams of paper to consular services.