Your Pricing Data Shows What
Customers Did. Kapnova Wants to Reveal What Will Make Them Buy Next
You
could look backwards for a lifetime without adding materially to your
decision-making ability.
EXPERT OPINION BY HOWARD
TULLMAN, GENERAL MANAGING PARTNER, G2T3V AND CHICAGO HIGH TECH
INVESTORS @TULLMAN
Oct 13,
2
One of the most ironic
aspects of the AI revolution, as the latest frontier models race to
lead us into an increasingly uncertain future, is that all our knowledge (and
every scrap of data used to train these engines) is based on the past, but what
we’re looking for guidance on from these tools and new technologies is making
all kinds of decisions that are 100 percent about the future.
Personalized pricing, which targets each of us as individuals
defined by our past purchasing predilections, is emerging across the board with
some modest resistance. It feels more every day like the present is becoming
the future faster than it is becoming the past. Every autocatalytic change
seems to accelerate the next change coming rapidly down the line.
At the same time, noted
scholars, MBA professors and professional investors have all repeatedly
cautioned us and advised prospective entrepreneurs and ambitious innovators to
largely forget about the past. Why. Because no one has become successful in the
past. In addition, it’s abundantly clear that we all have a tendency to
remember the past as far more glorious and pleasant than it was. Millions of us
long for those imagined days of old when the whole world seemed to move more
slowly. There’s the past and there’s the story we tell ourselves about it. In
fact, too many business owners, operators and managers today spend their time
looking forward to the past and, as a result, continue to fall further behind
their competition.
At best, we’re told that
we should be concerned about the past only as a prelude to the future. It’s
critical to remember your past mistakes and address them so you don’t repeat
them, but being a prisoner of the past or letting it guide your future is a bad
bet. Virtually every financial investment document written in the last century
has contained these prophetic words: “past results are no guarantee of future
performance” or something very similar which is boilerplate language actually
required in many cases by applicable provisions of the securities laws. And
yet, all the major LLMs actually have virtually nothing else to base their
operations on.
So, I was intrigued to
learn about a new company called Kapnova, which is an AI decision-intelligence business
focused on predictive pricing for consumer brands. The Kapnova team has built a
series of new tools, agents, simulations, and analytical systems based largely
on research coming from MIT that takes an additive approach, which doesn’t
ignore all of the past history and behaviors or the extensive correlations that
can be drawn from that data, but adds specific emphasis on what they call
causal inference—basically not simply recording and reporting on what happened,
but focusing on figuring out why certain actions were taken and what the
specific drivers and levers were which caused those choices and decisions.
Once you’ve identified
the key influences and inputs to specific consumer actions, you can build
models and tests to determine how incremental changes in the variables,
expenditures, and volumes of pricing, promotions, marketing efforts and inventory allocations can all act to
increase revenues and profits. It’s another level of inquiry and
differentiation that feels like a sniper scope approach rather than a shotgun.
Every LLM engine takes all the past tonnage and grinds it for correlations, but
we’ve been taught from Day One that correlation alone never adequately explains
causation.
Kapnova’s offerings and
solutions combine econometrics, causal science, and human analysis, enhanced by
the use of specialized agents. The agents apply the conclusions drawn from
testing prior actions and spends in constant reviews of marketplace conditions
and changes and then identify and highlight instances where strategic and
timely changes in different variables will likely optimize the results.
Opportunities are called out and scored for the client teams to make final
determinations, but the speed of the operations isn’t belated or
after-the-fact. It provides real time suggestions and directions.
How far can prices be
raised on a given product before demand falls enough to impact overall revenue?
Which ad spends are working to produce incremental sales and which discounts
and promotions are actually costing the client money when the customers would
have purchased regardless of the price? To be clear, you could be looking
backwards for a lifetime without adding materially to your decision-making
ability with respect to ongoing current changes.
We can’t go back and
change the past, or the actions and decisions that were made that may have led
to unfortunate results, no matter how much we may wish to do so, but it’s never
too late to change the future. Kapnova offers a look ahead at what the future
is very likely to hold for consumer product pricing. We’ll never get better if
we keep doing things the same way that we did in the past.
