Tech Layoffs are on the rise. People are training their very replacements.

“They’ve been screen recording every senior engineer’s coding sessions since October. Building training data for the offshore team.”

“All the L5s and L6s who’ve been building the knowledge extraction systems. 1,200 engineers who documented their own workflows into training datasets”

“One L7 told me he spent his final two weeks creating detailed prompt libraries and workflow documentation. Thought he was being helpful for the transition. Turns out he was literally training the AI agent that replaced his entire org. The contractors offshore are using his exact prompts …”

Then the C-Suite off-shore it to Pakistan. (What happened to Made in the USA?)

We will wake up in Q3 and realize our economy is cratered. No matter what Layoff Tracker you use, the tech industry has lost 50K jobs in 2026. No new jobs are available because of AI right-sizing.

Profits are at an all-time high and rising, so the layoffs appear to be just to squeeze more juice from the lemon.

The thing I cannot wrap my head around: our economy is 66% of GDP. If layoffs are increasing and no (real) new hiring is occurring, what do you think happens to that 66%?

Entry-level jobs are in decline so college grads are not finding jobs either. People are unemployed longer than six months.

The employees who are remain post-layoff are told: if you can’t 10x productivity with Claude and Cursor, you’re already dead. They will burn out employees in no time, then fire them. Am I missing something?

Meanwhile, these 2 examples of AI burning companies: (1) at Amazon: “Amazon’s own AI coding agent Kiro reportedly “decided” the fastest way to fix a config error was to delete the entire production environment. Gone. A 6-hour outage. 6.3 million orders lost.”   and (2) at McKinsey: “Red-teamers used AI to gain full access to McKinsey’s chatbot in two hours. CodeWall accessed 46.5 million plaintext chat messages about strategy, mergers and acquisitions, and client engagements. The agent also obtained 728,000 confidential client data files, 57,000 user accounts, and 95 system prompts controlling the AI’s behavior.”

I am not entirely convinced that every document, file, prompt, data set entered into an LLM is walled off in the pay window. I believe it is all fed back into the Sentience and available for the right prompt. Think how many times Meta, Google, Amazon, et al ever stepped over the privacy line? Every. Single. Time.

There there are the surveys and studies about how enterprises are fed up with AI since the hype outpaces the reality. That may be due to (1) lack of available data due to silos and proprietary software; (2) lack of employee training on AI; and (3) employee fear of AI. But if the hype isn’t real, why are all these layoffs happening?

And again where does this go?

We will have another generation of college grads with huge loans and no way to pay them off.

When you consider how much advertising gambling does – and the Robinhood app is an addictive gambling app as well – maybe the whole point is to keep the populace broke. Can’t afford bullets if you are broke right?

But then how does the economy run?

When Amazon is laying off between 16,000 and 40,000. Oracle is laying off 20-30K. These employees had email, SMS, DID, tools, SaaS – so now the SaaS providers and the CSPs have to turn down at least 36K seats, at multiple providers. That will affect SF, Zendesk, Wiz, Microsoft, Atlassian. These companies are also announcing layoffs. And around and around we go. But the revenue will decline for all SaaS and monthly recurring revenue.

And if we are just using Lovable and REPLIT, why does anyone need to pay for CRM or any portal? That is outer edge thinking there.

Many developers are not going to be able to transition into AI operators. It is a different skill set. So what do they do?

I just don’t see a great short-term future here. What am I missing?

 

 

I am fascinated with Private Equity. I organized startup events, helped at others, mentored some startups in Tampa Bay for ten years. Investors were part of the startup environment. I have watched PE buy up MSPs, SaaS, ISPs, houses and more. So I follow the news. This popped up on LI: a little bit about the PE Report, then a post about Gong, a PE backed SaaS company.

 

Here’s another interesting fact from Bain & Company’s recent Global Private Equity Report 2026 (released around February 2026): The private equity industry is sitting on a massive backlog of 32,000 unsold companies valued at $3.8 trillion in unsold assets. This “exit bottleneck” has contributed to the prolonged liquidity issues, with average holding periods for buyout funds now stretching to around seven years (up from 5–6 years between 2010 and 2021).

We see this in MSP vendors and in TSBs.

Private equity’s “dry spell” refers to a prolonged slump in deal exits and investor payouts, with distributions at just 14% of net asset value in 2025—marking four straight years of decline, per Bain’s report. It’s worse in duration than post-2008 due to high interest rates, tariff uncertainties, and $3.8T in unsold assets. Ramifications: Longer asset holds (avg. 7 years), tougher fundraising (down 16%), pressure for 12% annual EBITDA growth, potential job impacts in PE-backed firms, and broader economic drag from stalled company sales. Still, PE offers unique diversification.

Comments on this LI post were also interesting:

When private equity freezes, it’s not sentiment… it’s funding cost. Higher rates compress exit multiples, lenders tighten terms, and leverage math stops working. Dry powder isn’t dry… It’s expensive. Capital only moves when the spread justifies the risk.

Comment 2:

In 2008 it was credit collapse. Today it’s exit paralysis. PE exit value fell ~30–40% YoY, IPO window shut, and $2T+ in dry powder sits idle. This isn’t a liquidity crisis it’s a valuation mismatch between sellers stuck in 2021 and buyers in 2026.  [The Gong case below is a perfect example.]

So PE money is tied up longer with smaller returns, but they still have money to invest – and they are in less risky areas with assets like data centers and fiber in the ground.

One case of valuations versus exit explained on LinkedIn: The Gong case:

• Gong was valued at $7.25B in 2021.
• They’re now worth $4.5B in secondary markets
• That’s a 38% drop while revenue tripled!

So why haven’t they gone public? Because they can’t…

Here’s what nobody’s saying out loud: Gong should have IPO’d 2 years ago.

Category leader, Gartner Magic Quadrant, 4,000+ customers, massive ARR – Every metric screams “go public”.

Except one – their valuation. The 2021 problem: $100M ARR, VCs paid $7.25B. That’s a 72x revenue multiple. Even the most generous public market SaaS multiples in 2021 were 20-30x.

They priced for perfection and got it – for about 6 months.

The 2025 problem: they’re at $300M ARR now and secondary markets value them at $4.5B – That’s a 15x multiple.  If they IPO at that valuation, it’s a 38% down-round from their last raise. No founder wants that headline.

But here’s the real issue – their moat disappeared.  In 2021, conversation intelligence was Gong’s proprietary tech. In 2025, it’s a feature in every CRM:
→ HubSpot has it
→ Salesforce has it
→ Microsoft has it
→ ZoomInfo has it

Market data shows 40% increase in “Gong alternative” evaluations in 2024-2025. When your core product becomes a commodity feature, your 72x multiple becomes a 15x multiple.

The CEO’s tell: March 2025 – “An IPO is very interesting but not the most important thing. We are focusing on building amazing products”

Translation: We can’t go public at current valuations and we’re hoping the market recovers.

“We’re nearly profitable and still have plenty of cash from our 2021 round”

Translation: We don’t need to raise, which is good because we can’t raise at a higher valuation.

“We almost haven’t touched it”

Translation: We’ve been very careful with cash because we knew this was coming.

This is what happens when you build a point solution that becomes table stakes

Conversation intelligence was revolutionary in 2018. By 2025, it’s expected functionality in any sales tool.

Gong tried to expand – forecasting, deal management, revenue orchestration. But they’re competing with Clari (who just merged with Salesloft), Salesforce (who just acquired Momentum), and HubSpot.

All of whom have deeper integrations and broader platforms

The trap they’re in: too big to sell – who’s buying a $4.5B conversation intelligence company when every CRM already has it built in?

Can’t IPO – market won’t reward a down-round from 2021’s peak

Can’t raise – no VC is writing a check at a higher valuation when secondary markets say you’re worth 38% less.

So they wait – And hope that AI hype brings back 2021 multiples.

Great comment about Gong: The real headline isn’t “why no IPO.” It’s “did they become the system of record for revenue decisions, or just the best UI for call snippets.”

 

There was a SaaS sell off in the stock market this month.

“Software stocks have undergone the largest nonrecessionary 12-month drawdown in over 30 years, shedding more than $1.3 trillion in market value.” – Prof Galloway

Everyone thinks they are going to vice code their own personalized software. Yeah good luck with that for so many reasons. Unless you are using a localized LLM, your data will be in a public LLM. Tech companies do not value data privacy! When they go broke, like 23andme, someone will buy the assets and have all your data!!!

I wonder why anyone would use SF. A CRM at its heart is a UI to provide you insights into your customer data. All of your data can sit in an SQL database and Lovable can provide you with a nice dashboard and portal to it. Why pay $199 per user? Now your data isn’t sitting in a software company’s silo, you have it.

The biggest hurdle to AI projects is the data problem. Most data is not accessible. It sits in Salesforce, Stripe, a PSA, Quickbooks, any number of SaaS apps, Gmail, Office365 — your business doesn’t have your data, many other companies do. So you will pay for them to have your data and then pay them for AI agents so you can have insights into your data. Does any of that make sense?

For $150K upfront in hardware costs and $3k per month in colocation, you can own your data and the LLM.

While we are discussing AI, I have had to sit through a 100 pitches and demos on AI Agents in the last year. Very few are impressive. Most of my customers are meh on the idea, since they don’t want to spend money on emerging tech without concrete outcomes. They will wait till it is baked and ready to eat, which will be a while yet.

From an AI class I am in:  ‘85% of the workforce does not have a valuable AI use case. That’s what [Section School] sees in almost every enterprise we work with. Companies deploy ChatGPT or Copilot, hold a few training sessions, and leave the workforce to figure it out for themselves. Almost no one does.’

We don’t train people anymore (we aren’t hiring either). The technology has outpaced most users’ ability to assimilate it. If you want the best results, you have to train for user adoption.

 

During the last month I have been to 3 conferences. So many conversations. A majority of the conversations were around buying/selling providers and hiring. I follow a number of recruiters on LinkedIn and can honestly say that hiring is broken in the US. Job announcements for jobs that don’t exist since now job openings are like publicity. Multiple interviews. Upload video replies (discrimination much?). Do a small project. Have companies lost their collective mind? No one is going to jump through hoops except maybe the desperate. Do you think this is how you hire the best?

As an aside, your hiring process affects your brand in a large way. One ghosted applicants tells 10 people, posts about it on social – it is a mark against your company to the next candidate. Just saying.

Then there is the Gen Z/new college graduates who are entering a declining job market. Here’s the big problem: We don’t train people any more. We expect someone to come in and produce. It used to be just sales positions were like that, but now it is every position.

That is a lousy expectation. It also means that there is nothing unique about your company if an outsider can come in and just get to work.

This is also why your Culture sucks! Your Culture doesn’t consist of words or mantras, it is made up by the actions – who you hire, who you promote, who you praise, how you manage. To maintain Culture, you must  have an on-boarding process.

In fact, the researchers noted that when “senior leaders changed how they led—how they ran meetings, gave feedback, made decisions and responded to challenge—trust scores rose by an average of 26%.” [Forbes]

At places like Verizon and Lumen where they have let go tens of thousands of employees, there has been a massive brain drain. This isn’t good either. AI is speeding up the brain drain too.

Many businesses do not have written procedures and processes. Workflows are not documented. CRM systems have spotty data. That is the makings of a problem. You can’t scale this way. You can’t have consistency this way. Your churn will go up.

Train your people. Then Cross-train your people. Mentor them. Coach them up. Why?

“When organizations invested in system-level improvements like manager coaching and conflict resolution, “burnout scores decreased by 22% and perceptions of fairness and leadership care increased significantly.” [Forbes]

Zig Ziglar emphasized that investing in employee training is essential, stating, “The only thing worse than training an employee and having them leave is not training them and having them stay“. He believed in nurturing, respecting, and training employees to foster long-term loyalty and high performance, viewing it as a crucial investment rather than a cost. He famously said, “You can have anything you want in life if you help enough people get what they want.”

Now for the commercial:  RAD-INFO INC is presenting a Workshop for marketing people who want to be coached up AND they will meet with a small group of peers. For 6 weeks, we will meet on Fridays to discuss marketing topics like Branding, Fundamentals, email and messaging. There will be 4 sessions of one on one coaching to enrich the experience. There are only 5 seats available. Details and Registration are HERE!

 

NCTA has a new impact campaign out to let people, pundits and politicians that the cable industry is impactful. I never thought the cableco’s would be talking about how they save a consumer money! The new campaign says cableco’s save consumers up to $1000 per year in cellular bills. This and TV they buy wholesale and bundle. Granted, Large Cable is building out 5G, CBRS and wi-fi offload to reduce the wholesale costs of mobility, but how long can they maintain thw low price lead?

Consumers do not get device discounts from cable – and that seems to be the bundle for the MNO’s: free phones.

Cisco just rolled out a new points-based partner program. I know many vendors, pundits and channel execs are going to talking about how this is the new model. (I have already seen many column inches about it in the last two months).

Here’s the thing, Cisco has very few peers when it comes to their partner program. Microsoft is really the only one that is comparable. HPE thinks it is, but with all their M&A, the program is too confusing to keep up with. Dell has gone through a period of getting rid of its M&A assets. VMWare is a mess.

The main reason Cisco and Microsoft can make these big changes is because their certifications are a billion dollar industry — and yours isn’t!

They have Brand. They create demand. They create business for their partners. More than 80% of their business comes from partners. They like their channel.

It is interesting because MSP M&A is not just heating up, it is an inferno of investor dollars. Yet the number of MSPs has actually shrunk from a high of 62K to about 44K now. While 85% of that 44K are too small to be involved in M&A. (I talk to investors often; they all want the MSPs in the $10-20M range in revenue and $2M in EBIDTA.)

At the same time that hundreds of vendors are looking to the channel, MSPs are stretched thin due to the share width of the IT space – from print to mobility to cybersec to managed services. For an MSP with 10 employees, that will be too much. It also doesn’t leave room for a dedicated sales rep.

All the more reason that the MSPs and VARs – especially the smaller ones – will embrace this, since they need to get points to maintain status, support and discounts!

I know smaller service providers (vendors) ($5M – 40M) are going to be looking to change their program too. DON’T.

Unless you provide leads to your partners, the change will not help you. In 2024, reports indicated an average of over $6 to $9 in partner value for every $1 of Microsoft revenue, depending on the specific study.  Do you make your partners $120 for every UCaaS seat sold?

Most of your issues with your partner program are foundational: no marketing, no strategy, no focus, no brand, no demand.

The biggest part of Cisco’s partner program is the certifications. And those certs are portable. And those certs are recognized by Buyers. It is a billion dollar industry, yet Cisco certification holders make an average of $94,840 a year!

Partners are invested in the Cisco program. Their business revolves around it. They are Cisco first. It is similar with Microsoft partners. TAs and other programs can work that way but usually do not.