Held & Todd
Group
We tell you whether AI actually solves your problem before you spend a dollar on it.
We tell you whether AI actually solves your problem before you spend a dollar on it.
They fail at the diagnosis. We do that part first.
Before you commit budget, we tell you three things: whether AI solves the problem in front of you, which problem is worth solving first, and what it would take to be confident in that answer.
We map where time, cost, and error actually accumulate in your operation, not where a vendor says they do.
Any firm can buy any AI product on the market tomorrow. What can't be bought is the time and internal expertise to tell a genuine operational fit from a well marketed demo.
That work is expensive precisely because it's invisible. It shows up as months of a manager's attention, a pilot that quietly stalls, and a team that's harder to convince the next time around.
Sometimes the honest answer is that it's a process problem, and no model will fix it. We'd rather tell you that early than sell you a project.
Real estate operators, property managers, and independently owned businesses without an internal technology team, who can't afford to absorb the cost of finding out an AI tool doesn't work.
Three public deployments, plotted by how well the problem was diagnosed before the tool was built. Select a point to read the case.
In 2021 Zillow shut down its AI powered home buying arm after its pricing model systematically overestimated home values, a loss of more than $500 million, and significant layoffs.
The model was sophisticated. The failure was strategic. It was scaled before anyone properly tested the assumptions underneath it, in a market moving faster than it could adapt.
Klarna's AI support assistant handled 2.3 million chats in its first month, cut average resolution time from roughly 11 minutes to 2, and drove a reported 25% drop in repeat contacts. Real gains.
But by 2025 the company acknowledged it had leaned too far on cost, saw quality slip, and began rehiring human agents. Strong technology, scoped past the point where it fit.
Morgan Stanley pointed AI at one well defined task: letting advisors search the firm's own research library in plain language.
Document retrieval efficiency moved from roughly 20% to 80%, a companion tool now saves advisors about half an hour per client meeting, and adoption passed 98% of advisor teams, because a human still reviews everything that reaches a client.
The technology across these three cases was broadly comparable. The outcomes were not. What separated them was whether the problem was diagnosed properly before the tool was built. That diagnostic step is the work we do.
Charlie leads the diagnostic framework. He studies finance at Indiana University's Kelley School of Business and completed coursework at the London School of Economics on AI strategy, management, and governance, the research base the firm's approach is built on. He previously helped start and ran a residential services business with five employees and more than a hundred clients, which is where the operating instinct comes from.
Alex leads the real estate side of the work, translating a diagnosis into the language and constraints of how property businesses actually run. He studies finance and real estate at Indiana University's Kelley School of Business, and works through the Real Estate Club on financial modeling and market research for commercial investment cases.
We're building our first engagements now. If you're weighing an AI decision and want a second read on it, we're glad to talk, no cost, no obligation.
Held & Todd Group is a new advisory firm. We'd rather be straight about that than pretend otherwise. In practice it means both founders work directly on every engagement, the book stays small, and we have every incentive to get the diagnosis right.