The Future of Business is … Customer Centric Supply Chains!

Phil Fersht of HFS Research recently did a great LinkedIn post summarizing a fascinating conversation with Malcolm Frank that summarized a few key takeaways, including the following:

For 25 years, IT services optimized SG&A instead of transforming cost of goods sold. AI changes that. The real value now sits in agentic, vertical, customer-facing transformation, not back-office efficiency.

Customer-facing transformation is definitely where the value is in a global economy that is (borderline) recessionary, with joblessness and insecurity increasing by the day, and most people having less (and less) to spend on non-essentials and essentials alike. If you want their business, especially if your product or service is discretionary, it needs to be what they want. With constantly crushing weights on their shoulders, they need products that make them feel good, that make them feel like they are being listened to and catered too, that were created for consumer use (and not for the use by the atypical person in the lab who created something just for them), etc. The companies that deliver those will be the big winners, not the ones that still follow the old Ford Mantra (where you can have any colour you want as long as it’s black).

However, it’s not just creating the product that the customers want because IF you can’t deliver the goods at a price point the majority of your customers can afford and will pay in tight/recessionary economies, then you won’t sell any product at all!

We all need to remember that COGS was always a proxy, as it was easy for the accountants to measure, the same way we use revenue as a proxy for determining if a company is an appropriate target for our software and services. In Procurement, it’s not revenue — it’s how much spend is external, how much we can actually manage (retailers can have large leases that make up a significant portion of external spend which Procurement can’t do a thing about), and how many categories are big enough to give us leverage or real options when sourcing that can lead to savings, quality improvements, more resilience, etc.

This means that the future of business is about two things:

  1. tailoring to customers (because we’re long beyond you can have any colour you want as long as its black) to maximize the amount they will pay (to the point they can pay), which Phil astutely noted in his post and
  2. (dynamically) re-configuring the supply chains (as needed) to offer the products at profitable price points based on what the majority of the market will pay

So this would mean it’s simultaneously optimizing the product mix for customer adoption while ensuring the supply chains are ready to serve and re-optimizing them as needed.

As was noted, at the end of the day, back-office costs are pretty insignificant compared to supply chain costs and increased profits from increased price points that create a product that maximizes what a customer will pay (because the product is precisely what the customer wants, and not a product that is simply close enough that it might work for them).

If A SaaS Provider Offers You a 95% Discount …

Slam the door, lock it; close the shutters, bolt them; don’t answer the phones, and rip the cables out of the wall; turn on the frequency jamming, and throw the cell phones in the Faraday cage; close the gates to the parking lot, and man security 24 hours.

No matter what they tell you, a 95% discount from a vendor always means a combination of EXACTLY two things.

  1. the provider was trying to rip you off (because they thought they could due to their customer portfolio, surging popularity, or your lack of market SaaS pricing intelligence) and
  2. the provider is in financial difficulty

That’s it. The only unknown is the weighting between those two realities (and just how severe the financial difficulty is).

They’re NOT giving you a huge discount because they want your logo or case study.
They might want your logo and case study, but a solid provider with a solid solution who creates a good relationship can certainly get it without 95% discounts — most customers who get real ROI from a solution offered at a fair market price are happy to give you a case study for the free publicity.

They’re NOT giving you a huge discount to prove value in exchange for future purchases.
Everyone knows there’s no guarantee those will happen, even if you get the full promised value of the solution. You might have no use for their other solutions. You might never need any additional seats.

And any other reason they can come up with is also a lie.

Unless the company is run by a bunch of cons where their entire business ethos is charge as much as you can for as long as you can until the market realizes how much they are being ripped off (and then the cons skip town), the only reason a company will offer that level of discount is because they are desperate to get a sale on the books because, if they don’t, someone is losing their job in the best case or the company is going bankrupt in the worst case. Either way, that’s not a vendor you want to be putting your faith in. You want honest companies who price based on actual costs with a fair markup and who are financially stable — not dishonest companies who price based on how much they think they can scam you while being on the verge of bankruptcy.

And never kid yourself that it’s worth the risk because all the company needs is a few deals and a right-size on its pricing because a company losing money can’t stay in business — and any piece of enterprise software fairly priced at 1M will cost the company offering it at least half of that sale price to adequately support. You need to keep two things in mind

  1. cloud compute costs are real and significant and, thanks to Gen-AI that is over-straining global compute infrastructure, rising year-over-year
  2. the development talent needed to maintain and secure your solution (and despite claims, Gen-AI can’t do either, especially since it typically makes your solution less secure) is not cheap either

So if you intend to have 10,000 users hitting the app daily and doing at least one compute-intensive task (and LLM queries are compute-intensive, at least 20X as compute intensive as a classic Google or Lucene search, and possibly 200X depending on what’s being asked), your provider’s cloud costs will be in the six figures — which means the 95% discount isn’t even covering their hosting costs and they are digging themselves into a deeper grave just by signing you!

How You Know Your Education System Is Broken!

Only 40% of employees say they’d be fine NEVER using AI again! (As per a recent Section AI survey in the Wall Street Journal of 5,000 white collar workers, as reported in a recent post by Stephen Klein who also noted that the majority of employees say it only saves them 2 hours or less per week. Furthermore, he also mentioned a Workday study that reported every 10 hours “saved” by AI resulted in 4 hours being lost due to required error corrections, flawed output revision, and necessary verifications, which means there aren’t much savings at all. [Specifically, for an average employee to actually save 10 hours, they’d have to save almost 16 hours, which would take them two months to achieve!])

Gen-AI is failing 94% of the time. It’s causing serious cognitive apathy and decreasing our IQs far beyond what Twitter achieved on its introduction (where it reduced our collective attention spans to that of a goldfish). It’s direct and indirect costs to run 8 hours a day are often more than to just hire another person (due to compute requirements that are 20X to 200X that of Google for a basic query, and the extreme amount of energy and water [for cooling] required on grids that are already stressed and ecosystems where fresh water is running out).

Chat-GPT. Claude. Grok. Rufus. Gemini. Meta. DeepSeek. Perplexity. Co-pilot. Poe. Le Chat. They’re all over applied due to over promises when they all have fundamental issues (like hallucinations) that cannot be trained out (as the issues are a result of their core design and programming), limited data sets (and now that AIs are being used to generate additional training data, performance is getting worse), limited guidance, and no guardrails.

There’s always been a time and a place for proper AI, but it’s not now, it’s not everywhere the investors losing Billions on Open AI and competitors are telling you, and it’s not the “AI” they are pushing.

Every time a new advancement in tech comes along, we always forget how long it takes to get from prototype to safe for unmonitored regular industrial and home use, be it hardware or software. With AI, it’s always been about two decades between a new algorithm being invented, and a production ready system with known performance, limits, and guardrails being ready for the mass market. In other words, this tech shouldn’t even be out of the research labs yet! We definitely shouldn’t have every major consultancy trying to push it as the cure-all for every problem throughout your entire enterprise. (Or new start-ups claiming they can offer you AI Employees!)

How many more examples of (silicon) snake oil do we need before we accept there is no panacea for all your ailments — be they physical, mental, or industrial — abandon this current iteration of Gen-AI, and go back to the targeted, mature, solutions that were finally ready for prime time (as we finally had enough processing power, data, and research behind us to deploy them with confidence)?

And even though the technology might work as much as 12% of the time, as per a PwC study that found that 12% of 4,454 CEOs surveyed reported both revenue gains and cost reductions, that’s not much of a validation of the technology — especially since those gains and cost reductions could have nothing to do with AI at all (and the pilot success of 6% from a recent McKinsey is a much more reliable metric here).

If you want real success, find a (A)RPA solution that works, lie its AI and buy it while you wait another decade for this technology to mature to the point its reliable, guarded, and safe for mass market adoption and widespread application. (Or wait for an AI-enabled SaS provider to come along who will do the 24/7/365 human monitoring required for you and make its software is usable and safe through this monitoring. Because all the current generation of LLM[-powered Agentic AI] tech is doing is increasing the need for human monitoring, not decreasing it.)

Tomorrow is International Women’s Day.

So prepare for a massive onslaught of posts by companies large and small, from far and wide, that will lavish heaps of praise on their female (identifying) employees and all the hard work they do … and then prepare to hear absolutely nothing about how great these female employees are for the next year!

Right now, there is a lot of pushback in the US against DEI, and rightfully so since the whole point of DEI — equal opportunity and equity in treatment of all individuals from an employment perspective (future, present, and past) — has been replaced with objective outcome measures that result in the first person who checks the right mix of race-religion-gender (identifying) boxes being hired, and not the first person who qualifies for the job, which not only results in poorer organizational performance but resentment and backlash when qualified candidates are discriminated against because they don’t check certain boxes (and this includes discrimination against more qualified female applicants who would be rejected in place of a disabled male Asian Zoroastrian because that checks 3 boxes on the DEI bingo card).

But there isn’t nearly as much pushback against virtue signalling for accepted causes, or, even worse, basic decency. And this is a shame, because
* you don’t recognize your female employees by publicly lavishing praise on them one day a year and then completely ignoring them the other 364 days,
* you don’t respect your female employees by paying them less than their male counterparts because “that’s just how it works”, and
* you definitely don’t honour your female employees by claiming they aren’t suitable for C-Suite positions because they want more family time or you expect them to take a career break to raise the next generation.

Instead
* you recognize your female employees by acknowleding them when they do something significant — no one wants lip service,
* you respect your female employees by paying them as much as you’d pay a man for the same job — especially when these female employees are probably more qualified, and
* you honour your female employees by recognizing that they are probably more capable of a C-Suite job than you are! (Remember, they regularly juggle work life and family management — which typically includes their work schedule, their partner’s schedule, and the schedules of 2 to 3 active kids — when you struggle to schedule your own meetings and make your tee time.)

In other words, if all you are going to do is annual virtue signalling, please don’t. It’s disrespectful and I personally can’t wait for the day the next #metoo movement in the corporate world calls out this hypocrisy.

Last year I penned a long post after IWD asking what you are doing TODAY to help women. Of course there were NO RESPONSES from any of the companies in our space who did multiple women’s day posts and ads, and in the next month where I scrolled LinkedIn feeds daily for at least 15 minutes looking to see if any of these same corporate feeds recognized a female employee, I came across three posts from three companies doing so — compared to the well over 100 posts from over 100 companies claiming to celebrate women on IWD.

I think our resident unwoke/uncancellable anti-virtue signalling crusader Jason Busch needs to take up this cause too! True equality for all! (And no lip service!)

Without Human Smarts, There Will Be No (Usable) AI!

And I’m so happy I’m not the only one pushing this theory. Mr. Stephen Klein recently published a great post on The Age of Pretend.

In the post he notes that:

“Everyone assumes AI’s biggest bottleneck is compute. … That assumption is wrong. The real bottleneck … is architecture, specifically, a design decision made in 1945. … The real constraint: the von Neumann bottleneck. Modern computers separate memory and processing. Data has to move back and forth between them. For most software, that’s fine.
For AI, it’s catastrophic.

Some numbers the industry rarely highlights:

  • Accessing off-chip memory consumes ~200× more energy than the computation itself
  • Roughly 80% of Google TPU energy goes to electrical connections, not math
  • A 70-billion-parameter model moves ~140 GB of data just to generate one token”

LET THAT SINK IN. Us old timers remember “640K out to be enough for anyone”! The Apollo Guidance Computer — you know, the one that was installed on each Apollo Command Module and Lunar Module in the Apollo Missions, had 2K Core RAM Memory and a 36K ROM. Even today, unless you have an iPhone 17, your phone probably only has 128 GB of storage. That means, even with the processing power of your phone (that dwarfs most computers us old timers have ever owned), you can only process ONE token. (Now do you understand why the data center [energy] demands for your Gen-AI chat-bots are destroying the planet? Anyway, we digress …)

This means that (Gen-)AI has hit a wall. Computer Architecture supports massive compute at scale, massive storage at scale, but not massive transfers at scale.

So what does this mean?

Do you remember the days of RAM drives? Not only did it speed things up, but it kept your machine cooler because, as Stephen noted, less energy accessing data in RAM than on disk.

And do you remember the fun of Assembly? (Okay, that’s sarcasm!) Once you learned to maximize register usage (i.e. re-sequencing processing so that you minimized reads from, and writes to, memory), your code got faster still (and machines stayed cooler longer, which was obvious by the lack of noisy fans spinning up).

We’ve known about this problem for decades. (Eight decades to be exact!) It’s too bad today’s students don’t study the basics and understand it’s not strength that determines computational speed and energy requirements, it’s data scale — whether the data fits in memory or not, whether “significant” chunks fit in the onboard GPU memory or not. (And specifically, can you scale the data down enough for the efficiency you require?)

But this is still the key point in Stephen’s article:
“The next major improvements will likely come from smarter algorithms.”

We might need brute force to detect patterns we can’t (yet) see, but the only way to truly advance is to understand those patterns and code optimal, light-weight algorithms that exploit fundamental rules to allow us to process data quickly and efficiently.

Until we figure that out. You’ll never have usable AI (and definitely never have REAL AI as not only will it never be intelligent, but it will never, ever, get anywhere close).