Finance brings one number, operations brings another, and the meeting stops. Both figures are usually calculated correctly, which is why the argument never resolves and why the tempting shortcut is the worst available option.
Read the postBuild versus buy compares two construction estimates, and construction is rarely where these projects succeed or fail. Four earlier questions tell you more, and they usually settle the build-versus-buy debate on their own.
Read the postPlatforms earn their cost at enterprise scale, where the same problem repeats across a dozen business units. A fifty-person business holds one instance of that problem and faces a very different calculation.
Read the postEvery unsanctioned AI tool inside your business marks a place where the official process was too slow or too painful. That information is expensive to gather deliberately, and your staff are producing it for free.
Read the postNvidia is reportedly buying the platform most open-weight AI models are built and distributed through. If your business uses open-source models, directly or through a tool that quietly does, that is a supply-chain event, not just tech news.
Read the postRate cards do not tell you what an AI provider actually costs. On one realistic workload, the model that looks 14% cheaper turns out to cost 11% more once the tokenizer is counted.
Read the postDeepSeek’s price increases, some as steep as 1000%, signal that AI providers are beginning to price like utilities. For small and mid-market businesses working on tight margins, the assumption that AI is cheap is quietly coming apart.
Read the postLonger research pieces on data and AI leadership in the mid-market. Each is sent by email, which is also how we confirm the address is a working one.
Why ungoverned AI use has become a measurable line item, and what the behavior behind it is telling you
Unsanctioned AI use now appears in a fifth of breaches and adds six figures to each. It is also the clearest map a business has of where its own work is failing.
Get the paper
How capacity limits and scattered ownership create risk that only becomes visible once it has cost something
Community banks are asked to compete on technology with a fraction of the budget. Three constraints explain most of what goes wrong, and none is a technology problem.
Get the paper
Why most AI initiatives stall before production, and how fractional leadership changes the equation
Roughly 60% of enterprise AI projects are abandoned for want of AI-ready data, and mid-market organizations feel it hardest. This paper sets out where initiatives stall, what the delay costs, and what a different leadership model changes.
Get the paper
Small and mid-market companies can't afford to sit on the sidelines while the landscape continues to change this rapidly.
Read on Medium
The promise of a productivity revolution feels increasingly hollow when the tools mostly produce drafts that look credible and lack substance.
Read on Medium
Shadow AI is the proliferation among workers of AI tools that are not approved by IT standards. It is a problem that is much harder to solve than it seems.
Read on Medium
As automation, AI, and technology take over more of customer service, companies need to remind themselves that the key is customer experience.
Read on Medium
A McKinsey report on advisor shortages arriving alongside an Ameriprise earnings report on hiring successes prompted a closer look at the profession’s staffing problem.
Read on Medium
The buzz around AI agents is deafening, but the true catalyst for adoption will be something more fundamental than processing power or sophisticated algorithms.
Read on Medium
Personalization is a promising application for generative AI, and enterprises still need to avoid the nail-meet-hammer approach that produces technically correct messages nobody wanted.
Read on Medium
Generative AI is coming for search in several different ways, and the opportunity to improve the retail shopper’s experience is a real one.
Read on Medium
Professionals early in their careers are increasingly keen to demonstrate that they can harness AI to generate business insight, which is not the same thing as being able to read the output.
Read on MediumHalf an hour is usually enough to establish whether fractional leadership suits where your business currently sits, and you will get a straight answer either way. There is no obligation attached, and no proposal unless you ask for one.
